建物における避難時間と群集流動の解析 by hitoshi watanabe

一

―

-1-―

―

￨―

￨￨￨￨￨￨￨￨￨￨￨￨￨￨￨￨￨￨￨￨￨￨￨

￨￨￨￨￨￨

― ■ 十■ 二上二_上 __L二塁二上 _l

1+￨￨十寸十卜￨十 H十月一 ―― ￨

IJ一

=‖

「

vl - oot

Anぃnaリダ,42強 cυ よ n寸ヽ nで、防いο a柚く hο υ Cいけけ

十 CD"ds l,、 bollclty,s ら

ぃつス暖ンはの責ガd牡瀬さり、え訴し了尉■に射♭ 雌、、 ‖ 、 lL― 臓燻ハ蚤 Qゴし ■ してつメ期次t鶴 D。ンっ `7でシい肌辱扇吼LLュ 7びり1っ 71よつ4tの。衆勧孜なのル0 ヽ Cつ η L3外 a編徹極tヽヽ )_ 戦爾1斗。,ヮ毛はゎもじ、｀ヽがるくみ鵬′ ・癒汀颯b秒で 1・

′

CP 96/75

AN ANALY3 0F EVACUAT10N TIMES AND THE MOVEMENT OF CROWDSIN BUILDINOS

修ギ 211つ￨)らり難ヴf榛ヒ鶏飛キ動か解斉

υF― θθ

S J Melinek,BSc,PhD,MhstP and S Booth 艦

ヽ

Data on the movernent of crowds are reviewed. The evacuation of buildings to a protected area represented b5r a staircase is considered and a theory for the estimation of the total evacuation time from buildings is presented.

ツ Building Research

stablishment

Fire Research Station Borehamwood

Herffordshire

AN AIiIALY$S OF EVACUATTON TTMES AND

fiIE

MOVEMENT OF CROWDS IN BIIILDINGSI

bySJMelinekandSBooth Ａ＝ ″

INTRODUCTEON

i澳駕塁鰹器営槻麗tミ_Ъ mce ぬ eltti雷ぶ島鷺お雷ま端出鮮淵 ∬ 1:電躍嵐Ⅷ 観器鮒器艦機 ei 漱 :露 .Thi8 m鱗 m祖 emagenり procedures. The ame for

_ノ

α aCuatiOn

etr response to an alarm. dependS On how alertぬ e occupants are andぬ

Cyわ anottr.For examplヽ

early,mob■ ity and ttiVお

轟留勝電 the response of hOSpltal paients躍諾礎:思欝 ained serVICe personnel. The

cal faCtOrs and by the presence of smoke or

evacuation‖ 7ne COuld also be ttfected b psyChO10」 e eVacuatiOn is orderly and assume tatユ OW hOt gases.Current regttations aim tO ensure ttatぬ en be rates on the usable e2dt rOutes are not ttected W the f"e.The ime for evacuation canぬ calculated from lmOWn data on the moVement of crov7ds.

脚雷鱚 :r:鮒 f憲」り

■e

‐

属ぷ撃題凛1躁器 ::庶:訛官

estimation of evacuaion imeS tO a place of Safeち

″.

In SectiOn m a formula is der市 ed fOr

ime for the tOtal evacuation of tall hildlngs. Evacuation imes calCulated frOm this formula are found to be in reasonable agreement w■

th ObServed Value,「

・て

鑽縦鋼掛舗榊瀞轟撚

あ可多

by the fOuowing measures:

ぬe 〕 faCiitation of smooぬユow Such as町

proviSiOn of barriers8

andぬe avoidance of bOme=necks

ノ

鋤麟

饉 e 1鳳請ぶ ll憲乳雛冒 ∬ 1:選露 :露闊 ξ席 :■ 1登鶴 1:電雷

rapidly or reach and enter a safe area(See SectiOns l and III) 二i)mtttenance

of CO―un■ catlon宙 th the bdlding occupants.A public address system

can be used for ths purpose iV) prOViSion Of emergency lighting。

It is assumed極 t

e fire. the■ OW On StairCaseS and oぬ er exitS Where usable iS not affected Wぬ

e ambiCnt conditions.

The veloCites and f10W rates cOuld be reduCed by any deterioration hぬ In pamdarD eSCape iS likely tO be made more diffiCdt by any reductiOn m宙

slbiliけ due tO

smoke. SECT10N l‐ CR¨

MOVEMENT DATA

棚宙撚盤剛膿鐸 Ч蹴盤山e 掛酵櫨 ∬ 鮒胤総減￡席導Tl篤塙」￡:鰹撤鰹お酷 y翼わbe c理 ∝ d tm years.■ On and pOSSible arlse because Of the SiZe― diStribユ ■

tions in Briね山 ,althouま differenceS l醒十 ,nts Of Japan and Britain. differences in bё ha宙 our between the inhaЫ

cro組 moVement in buildungS COnsists Of three general lnotions8 i) movement along corridors

ii)

movement up and,/or down staircases

navcr.csf +{}o$h {4t*.

These cases、 vlul nOw be considered for a cro、だr of people all moving in one direction. LIore

CuL A re宙 ew dぬe tteow of crowd mⅣ ementお

露

窯

忠

:ぶ

撫

由

en W

∬

be assumed that the data presented in this section are applicable to any croud of people. No clifferentiation between persons is made in terms of sex or age and no allowance is made for the possibility that some of the brilding occupants will be disabled. Provislon needs to be made for such persona, particularly in buildings, such as hospitals in which there is likely to be a large number of disabled people. In the USA about six per cent of the population suffers from a serious disabilityr. There will also be a proportion of pregnant women and other people suffering

It will

りヽ

い1

from temporary disabilities.

The age and sex of individuals in a crowd will make little difference to their velocifles since the velocity is determined by the speed of other people in the crowd. For unimpeded individuals the velocity does vary with age and sex. For example, men under thirty descend stairs about forty per cent quicker on average than wornen of over fiftJy'. there is less dlfference in the speed with which stairs are ascendd. Most of the-data presented in tlis paper are for adults' eg commuters. Some of the data are for school chifdreJ. Tbe flow rates for school children appear to be less than for afults hrt notmuch less. Baggage has beenfotrnd to have little effect ouwalkingvelocityl.

It will be aszumed that the naximun flow rate througb escape routes is propordonal to the widthIt has been found that tbls nrle is approxlmately correct forwidths of one to tlqee metres6. Most exlt widths fall within this range. lbe smallest recommended orit widtbS is 0.?6 m.

l. H

Motion along a corridor

value --t"-t, as showtr in Table 1. Thisflow and of velocitSr variation people. I showe_the Figure of rlsh large can be exceeded for a rate witb crowd densigr for corridors of 1.1 m to 3 m width2. It can be seen that there are three states whlch are determined by the density. At low densities (under one person per square metre) the speed of individuals is equal to their free walking velocity. The flow rate is a maximum when the crowd densit5r is between one and five persons per square metre. This state is that most likely to occur in an evacuation and is the case usually considered. For crovrd densities greater than five persons per square metre there is congestion and the flow rate decreases as the crowd density increaseg. The normal cE)acity of corridors is about 1.5 people

1・

饉

ｒ２

中， ∽ Ｅ

０

ヽ、

ヽ

＞。。 Φ 。 ∽

: : !

´/~ k__´ ヽ

′ ′

、、

Ｃ。

．

、

● ・０１１

ｏＯ

１

ワ０

｀｀

0・

Ｅ

ヽ

￨ :

一 ● しヽ，１０一

．

、ヽ

、

ヽ

● ヽ

ヽ、、

＝

ｕ

Density-Person m-2

6 oassages

Table l

Meagured rates of flow

ofpeqile (fron reference

Rate of flow, Persons Per second per metre width

British data

8-1.4

1.0

0.9

1.1-2.7

1.4

o.9-1.6

1.4

o.9-1.7

1.o-1。 5

0。

French data Amerlcan data Cinema cmstnrction

0。

9‐ 2.1

North American Transit Authorities

1.5

0.9

■.1

Paris metro Lodon TransPort Board

1.7

1.0

■.3

1.5

1.0

■.1

￨、

升

ゞ

¨′

by TOgaWa3 ~0・ 8 V=Voρ

The fo110Vmg forinula was deriVed experlmentany

○ vJhere V iS tte crowd WalHng vdociサ v。 ls a constant(1・

3ms~1)

ρ is the densittr in persOns per Square metre

The■ OW rate,N is given by

め

N=ρ V

ぬ erefOre

N=Voρ

・2

_1 _1 . =

This fOrmula giveS Values of N of about l.3 to l.7 persons m s .

rdendiesmChare耐凛監 w鶴翼② ヽ

Wiぶ棚

yanme 。 hrgeぬe tOtal aov7お叫わ

輩∬ 1脂 I強l響鑽選 lμttS on the inside Of the bend. On the Outside the spec 棚 : 淵 Ъ

＜´

鮨

lis ume change in the width Of Stepe HOWever.

[Fe=猛 :驚1」臨瓢蹴i乱 :

tsed. FOr larger doWnward gradients the speed

for a dOWnward gradient Of approxlmЦ y seven

degrees.

庶判蹴霊』

_dぬ

糧器 e pers∬ 鶴 FT思 l∬ 雷

鵠e 鴛織量 lncreases tte

The caSe of Staircases is very similar to that tt COrridors.As the croWd Wヤ densiサ r bendS hぬ e

∬:::L脚 :f瑞 :■肌電皿:重re:∴ 五器ぶ牌爆

:f・

爾撻]鶯難猟

rhere possュble.

穏 11器 ∴導葛晶等毬脚l翼11ぎdo― ward travel, the firSt 12 m. 14‐ 6チ in口 _ foで

Stance廊

S配 .me shdy6 r∝ om‐

:こ fOr each 6 m of vertiCal distance eiまt per cent "be"av∝

Tahle

Movement on gtalrcases

Source

Galbreath20

Direction

Densitye people -9

Mean velocit5l Vl (along line of slope)

- "-1

Rate of量 ow・ N'

::[sonsfl)s l

0.8

0.55

1.1

0.8

0.9

1.8

0.7

1.25

2.2

0.6

1.3

2.6

0.5

1.3

3.4

0.4

1.35

4.1

0。

1.25

4.4

0.25

1.1

5.4

0.5

1.3

05 1.94

0.51

１

０５

１

１５

0。

ヽL

ヽｒ

TogawaS

贅鑢聖椰・ "

2。

Ｗ ¨

2。

0。

ｔｉｔｔ ● ■ 一一１．

The time taken for a crowd to move frorn one end of a stalrcase to the other is given bV;the mean velocity. Table 2 shows that the mean velocity along the line of slope is about 0.5 m s-^. Measuiements by Paulsl0 confirm this conclusion. Pauls found the ma$.mgm velocity to be

about0.Sms-I. Takingthevelocltyalongthelineofslopetobe0.Sms-'givesanestimateof is, however,

the ma:<imum time to descend the stairs for an unimpeded srovd. lbis velocity likely to decrease if a large number of flights have to be traversed

iii)

Motion through an

edt

くヽ

For an exit of widtb b (metres), the time T for a crowd of Q people to pass through the opening is given by

Q/Nb

where N is tbe flow rate per metre.

This formula works until tbe e:<it becomes congested. Nis apout 1.7 persons ateragezrg with a ma.:<imum value of about 2 persons m-r s-t.

--1 "-1 oo

ie tSrpe of exit rnost closely studied is the top of a staircase leading downwarde from an open space. Tbis sihration is very similar to a crovd pushing through a congested doorway since those wishing to descend the staircase cannot see the bp of the stairs in front of them dtre to the crowd blocking their sight.

For heavy congestion the crowd density is higher at the top of the staircase than at the bottom. may be e:<pejcted that a smooth flow will not be established until one or two steps from the top of the staircase. The delay is theu mostly due to congestion at the top of the stairsS. In the middte the crovnling is evened out and the speed lncreases towards the bottom. Free flow at the head of the staircase can be assisted by a suitable system of baniersS.

At bigh densities the flow of a crowd can become blocked by a series of arches formed at a doorway or the head of a staircase. People coming from both sides of the exit form a wedge which gradually changes shape into an arch. The arch then collapses, allowing two wakes of people throrgh. The process is then repeated. As a result the evacuation rate as a function of time has the form of a saw tooth wave. The rnore frequently arches are formed the lower the flow rate. The likelihood of arch formation depends on many factorsll including width of exit, the slope of , the approach and crowd density. For exit widths irp to 1.2 m very stable arches can be formedrr. This tSpe of sihration can be very dangerous and can lead to the crorvd trampling over people who have fallen down. The eSistpnce of an arch action becomes conspicuouso when the flow rate reaches 1.8 persons m-ts-^. The frequengy of arch formation is approximately inversely proportional to the square of the exit width'^.

力ヽ

SEC10N n

―

EVACUA

N"PROTECTED AREA

th some E対 sting standards for means of escape are su7n7n,'ised in ais section and compared宙 at an the Occupants of a hilding can meOretica1 lormulae. Ex■ sting standards aim to ensureぬぬ ln an acceptable period. For ids purpose enter a protected area,usuany a staircase,w■ specifica饉 o,s are given for the fo1lowing factors:

1) maXimun dis'・ "ce to a staircase u)=inimum vldtt for corridors,exlts and stalrs.

Esc"e shOuld

pssme w nOrmal mems d話

霧跳Fe 漱 s:鶴蹴葛胤胤Tttl話 cmd亜 dead

墨乱電四 :拠』ギ認智 :盟 FT盤出麗

mrrors h podims whiぬ

p∞ple about the directlons of cor三 dors and edtso Evacuation can be facilitated by a commu‐

躙

盤 " 轟よ謝 r織1棚般電麟蝉憔響警

7ninimlltn time and at the minimum rlsk tO Jttre b〕

The recommended mttmun■ direct distance to a staircase depends on tte occupancy and the fire

presence of Ob8仕 uCtions such as intemal、熙工ls and fi績士lgs. Tコ piCal mx‐ lmum values for shOps and offices are 12 to 18 m whereぬ ere is only one exit and 30 m where

nsk,and i"reSぬ e ぬere are two e2dts5。

Omng to intemal waus and fittingsぬ e acttal travd dlstance wm exceed

the direct distance. The travel distance to a staircase should not exceed l.5 tlmes the maxim― dlrect diS● ・ nce5。￨

｀

監ぶ節冨貯:馴

7譜態ヽ轟識島町1駅 :縄of 翻轟貯器 ■ staircase. British

assumedぬ atぬ ere is a fLow of forty persons per mnute per unit wid■

留盤乱肌T翠ぷ蝋盤 F庶寵ぶ常蝋鳴綱塩∬1よ』冨期 approximatdy 80 persons per metre widul per=

it is assurned ttt any one staircase may be unusable in a fire.

欝鯉題 i膜貯 i:1薩桑 i遊算驀ユ pr the life of occupants after a period of two to m a fire w」 u depend Onぬ e size of the fire when it is discovered。

FOr a s'赫 ctt Of輌dぬ b serving n■

ぬ e sttrcase m a permittd

∞ rs,機

maXim― number Of people,M,who can enter

time,t,isぬ e mmb高

出継肌星:驚￨』議凛rlよ

stalrcase plus the n― ber WhO Can be accOmmく x peOple can be accommodated on the stairway between one stOrey and the next then,assuttg a :種

uniform population density,

M=N'bt+(n-1)S

(1)

It is shOWn inぬ e Appendix that s is given approximately by

s=18b+14b2 where b iS tt Staircase widuLin metreS.

Subsi恒血 g the value Of S in equatiOn(1)and talcing N'=80 perstts m~l min-10t=2.5 min gives the maximum number of people who call enterぬ

be:

e staircase宙ぬ h the acceptable ime to

M=200b+(18b+ 14b2)(n - 1)

(2)

Another formula which can be used is

(3) M=200b+ 50(b -0・ 3)(n-1) buildings not This formula is that represented by Table 2 of the GLC code of practiceo for

exceeding 30 rn in height. 亀ル

of a b2 For staircase widt}s up tO 2 rn equations (2) and (3) are in good agreement' The absence the standfor allowance less term in equation 1f1 implies that flr wide staircase widths it makes ing capacity of the landing (see Appendix)'

For buildings without compartment floors between storeys a draft Department of the Environment

;;"-;a;";;ticel?

gives the maximum number of people who can enter the staircase to be

M=222b

，

∽ ０ ● 属０ヽ０Ｃ００

負Ｏ ρ 口口０

ｏ８晉工ふ島ｚ

餞ｏ魚ｏ０日０Ｑ。ロＯｎ

Ｃ

∞ ．ＨＯ．守

∞ ．ＮＯ．０

ヨ

R88888斜

Ｈ０

∞ ０日

∞ ．守崎． ■

∞ ∞ Ｈ

０．００．ゆ

ＨＨＨ

０． ∞ ● ．０

ＯＨＨ

∞ ．卜一。ト

００Ｈ

∞ ．い０．０

い０

守． ∞ ０．寺

∞ ∞

∞ 。０崎。ト

∞ ．い０．ＮＨ

０．ＯＮ。ＮＨ

鶏

０ゆ

崎ヾ

■・島・１■ １１

１旧■ Ｅ，ＩＶ

００

負０卜と０ヽ

，ヽ

ll){tfNNNNcaNN\fl

6lN6e6tNgl1tN${.NN

錮漏日３３

０

∞

ＯＨ

Ｏ日

■ ■ ■

ＮＨ

卜Ｈ

ＨＮ

子一Ｎ

ゃ１〆凡ソ

．ｏ＞ｏ

颯￡Ｉ選８ｂ翼０

ＨＨ

F{^loa$ro(oFGocDg=

目 ● ｏ口ｏ＞ｏＯＣｍｏｏ痙姜

″ ゝ

ｎｄ

this code permits a populadon hrildings with compartrnent floors between all upper storeys

M=11l b n population, ag shown bY Table For office tnrildings the GLC code of practiceS permits a

y=(f0ob+10)n

code,of

(5)

3 of that (6)

lmrnediate evacuation of two floors only Eqtration (5) and (6) are based on the assumpdon that

wilf be required.

lbe number of people requtrilg-to use a staircase

5: . tottoruiog popufation densities4' Shops 0.14 to 0.50 Persons m-'

ca^u

normally be estimated assumlng the

Offlces and factories 0.10 p"""oo" rn-z' always less tha! tb€ that the population density in shops is usually trrt not "6or" valuesrecommendedfordeslppuryoses.Staircagewidthscalcrrlatedontheaesumptionofa per sqqare rnetre are t'berefore ltkely to be population density ru elqs of O.i+ to O. SO persons adequate in normal clrsumstances' Two strdiesl8rl9

SECTION

IU .

TIME FOR EVACUATION

tme of In this secflon formnlae are derlved for the evacuation values' observed some tlese formulae are cornpared wib

ffilrlings' lte

values given by

value lt evacuation 6ure of a fui[dlng occeeds an acceptable go event in 1he people can which to hrilding the will be lecessary t" p;;;;" places of retuge tn (see Secficr tr) and chosen ls minutes 2! of time .J""pt of a frre. For orample, lf an acceptable storey then no hdlding with ten or more it is assumed tlat a crowa takes 16 s to descend onethe accgltable escaPe time' An acceptable storeys above grornd level can be errasuated wltbin place of retuge wiff ln nost cases be a protected staircase'

In those eases where

ｒ口Ｆ

tle estiuat€d

be fornd by splitting the hrilding lnto cells' The The evacuation time for a partioilar hrilding can This rnodel c"u is calculated o-"rog tl" formulae given tn section I' evacuation time for or disovercrowding to due ".cl or hold.ps works well ln practice provided there are no Ulict<ages nrption of flow3.

Considerasirylemodetofamrrlti-storeyhrllding.Itisassumedt}atallt.hepeoplearewaitfurg ttre gound floor do not reduce the rate of flow at the exit statrs *t"tlt *i O"t people ieaving from the uPPer floors. The following sYmbols

will

be used:

Q" = PoPula6on of floor r 1 arod r \, = staircase widtl between floor r stalrs Nr = rate of flow of people per rnlt width down t'he

t" = tlme for member of unimped crowd to descend population of floor r and above g

= L t=r

one storey'

Minimumtimeforthepoprrlationtoentertheetaircasedownfromfloorr

= (\l i

Ti"l: f:" tail of crowd to reach

e')/tN'br-r) I

ground floor

=fr"

(8)

16 seconds' It is shown in the Appendix that ts is typically about Therefore, from equatiotls (7) and (8) ,I \ r- = ( tr et ) /t u' trr-1) +

\l=r

rt"

(e)

whereT"istheminimumevacuationtimeforthepopulationoffloorrandabove.

Equation(9)glvesnvaluesofT"(r=rtoP.TheminigrumevacuationdmeT.forthewhole T"' U,titAi"g is equaf to the higtrest of these values of

If there are several staircases then b" will be the total width of tle staircases considered. It is usually assumed that where t'here are two or more stalrcages, any one staircase will be unusable in a fire.

If the staircase width or the number of st^ircases increages as one descends thenthe evacuation

time may depend on the order ln which the people are evacuated.. For erample, if \-1>b" then the flow rate on the staircase down from floor r ls Nr br-1 only {vhile there are people waihng on floor r. It tben drops to Nr b1r which is the maximum rate at which people can descend the stairway from the floor above. Ihe evacuation time is therefore minimised lf people from any floor enter the staircase only at the maximum rate at which tley can do so wlthout redtrcing the fl,ow rate from the floors above. It is likely tlat any evacuadon will take place in thig rnqnnsl since the people already descendlng the stalrcase wlll not wish to be hafted.

the populadons and the staircase widths are the game all r, and equadon (9) becomes

for all floors then Q" = q

and

Tr=(n― r+ln/o'り +r ts

lf Qム N'D>ts

(10)

then Tr is a― Ximum when r=1.In this case

Te=nQ/tN'り +ts

rQ/(N'Dく ts then Tr is a maximum when r=n. 任iS Case Ъ =Q/tN=D+n ts evaCⅢ m ime Ъ 宙ぬ Qム Vい住

‖ ぼ糞ぶ

脚

for

(11)

(12)

ShOWn h Figu

2■

お

職

づ稚

Congestion一

n310

颯リ

n35

o/trt'uFigure

Calculated minimurn total evacuatlon times, T" for h.rilding with n floors above ground'level

Congestion occurs when the van of people from a floor, descending the staircase, meets the erowd entering the staircase on tbe floor below as illustrated in Figure 3.

ギ♪

No congestion

Congestion

Flpe 3 Conditlons for congestlon

The time for■ Ю crowd fromぬ e upper floor to descandぬ e statrs toぬ e n00r bdow is ts. nat fOr■e crov7d from the lower■oor to enter ttЮ 8'■ 1'S,aS― lng no congesは on lower dovm. 18Q(N'D v7here Q lS the populaion of the lower■ oor,N'18ぬ e■ OW rate downぬ e staircase per ・・‐lt、減dぬ and b is the w■ dぬ of the staircase down from」

be congestiOn if for any two n∞ rs

(_)

he lower storey. There w」はぬ erefore

Q/N・り >ts

(13)

EOtlon(11)cOrresponds to the case Where■ ere is conges● on and equatlon(12)toぬ e Case

whereぬ ere is no congestion. It can be seen from Figure 2 that when ttere is congesion the 7ninimum evacuaion ime increases rapidly,ほ ■ ■Ю b■ ilding populaion.

Observed evacuatlon imes for eleven office

gs10,20,21 are shown tn Table 3 togeter

述電avdue r卜 F筆響現 1営樹署蝋翼盤1離 landt121as譴 cabdated values. h those cases where tЮ

d綺

1.l persms

,served evacuation imes are greater manぬ e evacuatlon tlme exceededぬ e predicted value the average difference between the observed and predlcted evacua■ on tlmes was abouttwo mutese

ms tt appears tntぬ e evacuatlon times of hlldings are likely to be up to ahttlt two mlnutes longer man the values predicted by equatons(11)and(12).

ndding■ l in Table 3 wasぬ e Commercial Union Headquatrs Bllilding in London.ロヒe evacua‐ tlon ime shown is the average for four evacuaionso During one of these evacuatlonsぬ ere wtts a simulated rescue.The 4 minute extra朝はch this took has been ignored.■ has been assumed 饉 at the populaion of the 25ユ ∞ rs above podium level was 2400.The calculated evacuation ime h Table 3 has been increased byら mnute to anow for the ime requlred to descend the twO storeys fromぬ e podlum to ground f10or levd。 The dlfference between the observed andぬ e predicted evacuatlon times shown in Table 3 is

○

rl霞

h犠

亀輩為

翌器

踏

:L::翼覇

盤轟

輩

器

躍

∵』犠

爾

福

電撃

respond.Loudspeaker communication cm add to the credibiliけ of an alarm.Itis a180 uSem t。 have a systen■ of fire marsha13 0n eaCh floor l″ hose responsibilitr it iS tO See thateveryone leaves.

CONCLUSIONS

1 2

ftg

normal capaclty per unit width for corridors and stairs ls approximately 1.5 persons and 1.1 persons m-r s-' respectively.

m-r s-t

Vdocitles of l.31m s~l along cor五 dors and O.5ms~l on stairs(a10ngぬ e line of the statr8)

can be ass― ed for耐 mpededユ ow。 3

The=ロロdmum population M which can be evacuated to a staircase, assuming a perコ αltted evacuaion ime of 2告 mnutes,is given apprぬ mately by

M=200b+(18bキ 14b2)(n-1) where b is the staircase wtdth in metres n is the number ofstoreys served by ttЮ staircase.

4 The mmimum total evacuation ime Te for a multi― storey httding is given by

T:二

(Σ 〕

Qi1/(Nt br-1)+r ts

where r is the■ oor number(lto n)WttCh gives the ra濾 mum value of Te

Qi iS the population of■ ∞ ri br_1■ s the widぬ of the stalrcase be● 腱 en floors r-l and r

N'is the■ OW rate of people per un■ t width downぬ e stairs 1囁

ts is the time for a member of an unimpeded crtt to descend one storey.

r the pOpultton,Q,and the sね ircase width,b,are the same for each i∞ rぬ en Te iS the larger d Tl and Tn tVhere

Tl=崚 /● :り

+ts

Tn=Q/1N'D+n ts Tl corresponds to congesion onlall i∞ rs and Tn to no conges● one Typical values of Nl and ts are

N'=1.l persons m「

ls~1

ts=16s 5

1fぬ e sttrcase、減dぬ orぬ e number of stalrcases lncrease as one descends then the evacu‐ ation time is罰匝 mised by givlng prlority at landings to people already onぬ e statrcase.

REFERENCES

(_)「 :」ふ1ぷ温協 2

出譜織電

1蹄 "di範

Ass∝ iatl∝

ばUrban Ded"rs

Second report of the 9perational research餞 = m onぬ e capaclけ of f00tways.

Londbn Transport Bbard Research]RepOrt No 95. London, 1958.

Toga■■ .K. Study on ire escapes based on■

le Observation of multittlde currents.

Japanese Btlilding Research lnsitlte RepOrt No 14.Tolwo,1955。 4 Code of basic data for tte desitt of薗

餞dings

Chapter Ⅳ Precautlons against fire.

Bdish S'Ondard Code of Practice CP3.London.1968 and 1971. British Standards hstl籠饉on. 5

COde of practice― Means of escape in case of flre. Greater London Council, London,1974.

6 Fire grading of bttdingso post_war mding Studies No 29.Mは

stw of Wbrks.London,

1952. HM StatiOnery Office. 7

La Pattque?Qudle importance?Auo 18,1971,(珈 hrCh1 1■ 20.

mde tO safety at sports grounds.Home Orice/scottsh Home and Healぬ London,1973. HM Stationery Office.

９０

Predtetschensld,W M and詢皿 Insu・ Al.Personenstr6rne m cebatten.Kohn― Braunfeld, 1971. rO Bt11lding 26 June 1969。 Natlonal Research lding Research. Building Research Note No 80. Ottawa, 1971. 直 Council,Division of]腋

Pauls,JL.Evacuatlon dtt held in the BC Hメ

peschl,IA SZ. Doorstro血 gs― capaciteit 2662

Deparment.

van deuropeningen pameksihl・ tes. Bcuw, 1971.

Farren,D. Student's Notes― Means of escape in case of fire. Britlsh Fire Ser宙

ces

Association,Fire Safeサ Legislative Course, 1974. 13 V7oOlley,EL. Fire protectiOn ll. Means of escape in the case of fire. Bldg TTades」 1973, 167(5019)131-9。 14

Preparation et ex6cution de l'6vacuaton de personnes en cas de shistreo Revue tech. Feu,

1971, 12(105)7-12. 15 Evacuation from hiま

―rise l■ dlding8. Fire prev。

,1974,10525-26.

16 Shorter,G Wo etal. The St Lawrence tttms. Natn.Fire.ProtoAss.Q.,1960,53(0300-16. 17

Means of escape in case of fire. Department of the Environment,Draft]Design Code.

London,1974.

、

18 Norぬ ,MA.The number of people in shops.Building Research Establishment Current Paper CP 10/74. Borehamwood.1974.

19 mrchant,EW. U饉

versity of Edinhgh. Private communlcation,1974.

20 Calbreaぬ .M.■ me of evacuation by stairs in high rise ttdingso Natlonal Research Council of Canada.Di宙 slon of lmiding Research. FiFe Research Note No 8。

Ottawa,1969.

2l 22

Private communlcation. Commerclal Unlon, 1975.

Backer, G W and trrlack, R W. The occasion instant. National Academy of Sciences Dlsaster 15. Wasbington, 1960.

Study

M. A survey of exit facilities in trigh offlce buildings. Canadian Nadonal Research Corncil, Division of Brllding Research. hrilding Research Note No 64. Ottawa, 1968. C'albreatb,

APPE![DD(

Eetimatlo of paranetera In this appendix typical values of t" and s wtll be estiurated where t" = tine for member of unimpeded crowd to descend one storey

=number of people that can be accommodated on the staircase between one storey and the next

Descot 6me fime to descend one storey (igrroring landinga) = d,N' where d is the dletaJrce down tlre stalrE between one floor and the next meagured along line of slope and ignoring landings (-

the

Vr ls the croq/d velocity along the llne of slope

_,)

Gatbreath2S gives the total travel distances for ten offlce brildinge. Assuming distances across the landings of four times the staircase width the average dist^nce up the stairs for these hrildings, measurd along the line of slope and igroring landings, was about six metres per storey. As shown in TaEe 2rVr is typically 0.5 ms-r. It therefore seems reasonable to take a value of d/Y'of, 12 seconds. Ttre staircase

will include landings. As the flow is constant round a coraler we can consider straiglt corridor.

t.he

landing to be a

tlrE'di"tance across Two possible staircase conflgurations are shown ln Figure 4. In each """" the landings (dotted lines) is approximately 4b or more where b is the staircase width. If the distance across the landings is 4b and the walldng velocity ls V then the time taken is 4b/V.

-b*

-tF ￨￨￨￨￨￨￨￨

￨￨￨￨￨￨￨￨

．

兆１

￨￨￨￨￨￨￨

一一一一一〓￨￨￨￨￨￨￨

Figure

Possible gtaircase configurations

At this flow rate the croud The normal capacity of the stairway is about 1.1 person *-1 "-1. velocity acrosa the landings will be near the free walking veloclty (see Figrrre 1). A wandng velocity V across the landings of 1.3 m s-r will therefore be assumed. A typical staircase width is 1.2 m. Assumtng this value of b and a walking velocity V of 1. 3 * s-1 gives a time to cross the landings of about four seconds. lbus, including the time to eross the landings

t" =6f1t+4

time to descend'bne storeY =

= 16 seconds This value of t" is the -time for an individual in a descending crowd to travel one storey, An individual desdending dhe stairs alone can of course travel much more quicldy.

Standing capacitJr

The number of people who can be accommodated m stainrrayszo and landingsl ls about g.5 persons per square metre. lbe horizontal staircase area, excluding landings ls d cos 0 where d is the distance along the line of slope and 0 is the slope. As showu abve' d is typically 6 m, per storey. Ite landing width is likely to be about four times the staircase widtb. d is uzuallyr

about 30o

From these flguree

s=(bdcos0+4b2lp =18b+t4b2

s le the nurnber of peqile that can be accommodated per etorey b ts the etalrcase wldth and p ls the density of peqile per unit aree

where

Ttris equatton assumes a denstt5r of 3.5 peraons --2 cao be accommodated on landlngs and staircases. To avoid psychologtcal discomfort and to minimise tbe rlsk of panic a lowerrdensi$t is deslrable for prolonged aJays. A value of 1.5 persons m-2 has treen recommendedr.

キー ■ 一争覆曇書警懇露晨露群憲静群，一一 ■ 二 ■ ．

史らてこ λミ

■口｀人でバどJ∫ ,一

)丁

7ι

''..

1,1

作秘 F/1(.、

―上、二 1ス＼13、 ■、 3 CT‐ 、 =′

:こ

J、

:i下

￨し

.'. ufl') {r' -i, L.:a t r' i'

レ藤■′暉1

、ヒJ(「 lc tti13

13 ホキ」ことす― 、 1盤二

1｀ 1。

;li'.-r

1,イ

l´

′ ヒ〔 1,t iLく人 L■ ■ ｀ k〔

11・

T4?

llを

:口 k,JI、￨_

11ル

``]す

■【)ア￨[ビ

Zし

:ヽ

1小

ィー(

卜貸→ rふヒ

-'n 7,. ,,1 :l

｀メ柏´11イだ句、 ■Fiつ」17ギセlin tヽ ‐ 民111l

諷

￨「

Jガいてン

)F'「 tlし ti l、￨〔

｀「 F i1 1 し11ょ Z ピリ切・・ヽっ行う t

イ￨￨ボ iに

:ヽ

`, fヽ

ズネず。えしらか泄ι だ喧がいボ起 ,お

0も

｀ぢtらる小

小=l'lt つ崖メィo小晨Ь

｀暑がゎひЭょっ電気が、ゝ徹かな)し何I Ёlた F′

￨〔

j,かう植:毛口}乞す、

ヽっデ足と(艦長てヽた汽ち采 t‐

l~」

ー、ドでシボの Llf虻キたる美」予テ 13す。丁｀ゝし 1/1Nさ

[itし

Lる

(―

ιtt、打多なく1えじぞ■7_η ｀動か引つでりЧいすりヽ」が￨す

￨よ

そk7J｀コ

1」

シし■u,7こりヽ■で:で ,

｀ミの劇友、た面牧 J、だ

lξ

)2角

ば跡と、_"t ひわ

ヽぞたそ￨■ にヒ￨[大た 1三昴っヾたぃ lイン)、曖曜‐ 寺人くη 名￢み勃レ渠義争とうてl■

う '3ひ

3)街魏 Yどう彦うつ｀こ乳フIT

ηび術ろわ■■t.た P

δう、

2 づ携攣ヤ左憚馨卑野瑠』む乃棚直ヵ、ぅ_モ、tら ∩t｀ろ4 (d八 (■ 、 1う。め外、又、だす。 ■ヽざっビ 1颯の 4カ耐ヒヽ｀ ͡ ヽ「大■t、りいの μ十瞬マの110に颯∩句場に関■5加 i橋いうのが均tワ ♭ ]ll猥社宅 1が、印‖ だたて執γ てフに見 ■13颯研敗した人「 _ゴ、力 432めすちヽだかし13-ず、簡 J.ヽキ t」

￨■

`￢

￨ご

1■

t｀

￨」

｀いうした(31しマι聰膨みの外η'じただZヽ多ウすしム、何人力｀、 λ ιl)リバt 了、 _ に %■ 、そづがミてのりヽたて、｀ ■ 拝ちU口、旬効に関 L~71」かか)拳純イして麒マЧ弓す。そに人血ヒ｀｀ツこそ均い 1楓 τ たしらてぁろう1熟作つ、 )■ 0ユししヽてあl) tha 2務 ■れ障のだ災にょ5ヌトし入ヽマ3バヒ衡つぅの手1が、こヽヽ B-3ぬいマも,V「につマば勇,λ 毯以て、￨`でん。 t」

￨

￨・

￨て

￨し

｀ :嘉乳プ焔曽 [徽の寵おるふ ′ 家オ乱｀ヽヒぃっマ∩船子ら小をいぃ夕のた ■ハめ、そ ■を (名

bう

Lttt,の 3た

ヽ ‐久1懲牡吃翼み ■部みしそした。ィ(マう、 `幌

1人

ι ヽ期角氷上にフ″

tこマ多1)盛旧bワ幻雌 tひる｀オ、、 3いう0事 ■唆乞 ,らず鞍い又ヽ1をぃとぃぅl｀ ζ■貝 )l■ 」し。このイ)りご乙~ 暉とたて露っ ― f・ =￨― フ■鶴晰 t7■ でしも易こンの多ハ￢ォ」に■ぃつ■う、す(ll末詳で }らヽ _(1・気がつ(プ後 tl]、こ人陽本たた点ひ,7_ ｀｀ざ号、ザがZ ttkドだ 13で ■ン1断 13ゝ ?為瑠安、のうとでっに、あ3こてて｀不足》tを 77考ィ 1]わいヵす。フこ、 {・

7_―

￨ノ

■卜たみJ

tttaFさ

1)・

^Rもい禾■レマ、、、し bごて小、穫 ≧ ｀￨プ可 Zムと■じめ来

ヽ｀ぜQ形で襴耗ヽマヽ＼、動ガ bてて Ч3aモぅィ、かうスの打つも

ヽ ι Й｀たすフトル%でぢ｀のハけ少t■ る3たさ由がЪです。薇解―

、２ヽジっ

/」

でヒ

しム￨ずすれ襴牢行しド｀一たし火安であつとる里i化したЬ.特｀ヽ tり 4吋政は 43`(42じてたに t頃、ま ttプマヒし乏わ(22 Ctつて゛竿ムの初めに感暑魂 tルしc艦ポて´ 、ヽ｀ ′ ・わη .こ tl、 ■)で lt)｀し難がそこへB々多Ktりヵ昨Ъ陥1鮭ヾ洪ゴにもてこ(ざヒtラトしヾ可ィ´″ てt(7■ . もしプじィ uっ了しょぅ Stて ,7■ て■ 式 L6達家アれ tらく動な行わ｀ ″枢にあbて V痙聾狂麟 'kL tl手 . 釧〕 ■ I・

わ石ッゃすЧ こうに入さu}す。

ヽ臓腸のラク乱囲せ M2し手 '颯

む

｀ V鴻ゆ伝て｀ 1_怨静(める可カム、キ_Lt_いうようにハ班スt_「」 ■てヽオ■U県女3した。チ )しヽ独笥喜 bィ娘選五換機ブ aF匁ラらに1‐ 1つに｀ `ρ カ年ヽ 7レム10 人賀 ttレ 11か ,711畿ぃ●`:準 ■む橘 )やばtるここ』イ ~//tム ■籠徴iい御サ乏あうようもお■)ヽイカDBな ,猥｀お、 'も｀通身てにけっここがで■]●■ 年Lム￨」マン時わ113行動 t ヽ訪δ cl専つて。おもヽろQztttすぃ￨・ヒ■tもヽ直■て｀ TILって 1■

1■

￨「

rヽ

｀人マ′ 文たよブ η鞣つ建ソ均q2薩秘ヽこでヽ、 (ム )豫 10してムぐ ′

c イtマ tlこっoあ考通ぅЪlブリ巧l・暉ヽБ予芙贅嗽Ъ￢:商何藪もヽ｀、たもa■ ヵバーできるじ)IL m乃と0山っっふリワす、著斐多ぃ[￨ざーのて、うフ(」 (1す Ь漱了 lフしイヤ `がコ〕ピ2-7ヒつをげBさてよっ｀ヘヽゝ ↓ カカ礁 t ut人 2ヽ 1 燿2のう ■■ シミ2し +ヒ ■ta■ じ `可えニマー在ヽ年や象の、ムヘヤに市2ぅ xttЪ ヒてtでヽ D とのポすざぅl■ 得らて」しごう。 )っ

￨「

ttλ

:≫

`マ

)、

｀ご①ザシリヽツリー1専、コtン化にこた

I~///受

l~2わ

つ、美再の ‐

外テした抜_zい人υヵ、て,1■ obt,3を行む力I 可ての/t臥てさ。｀ ‐ てフつレにテ、 Qつ 2も7危￨. ずご )を tt h傷たハこい '恥か? tT~ら｀誦ゴ,cィや￨ミンじたを行しつ乏て ]可ヽつしよ● と‐ lЪ 6っハ外ランの)ちみこごど、最初にもレメ ■ ´ い。ノ・∠■ ー又下で >でとぎ寺火(7ヽ 1スカつてこ■ 「＼ 7ワトと｀｀｀ ,て｀うブ(」 FIす 3● Fk_41クでD(りて ′ しらヽ ?)R下てい。午刀'て tノ

1電

`う

‐

similar\e

supply sufficient documents. nc sub-contractors not named ai the Of In order to es\blish the magnitude and tender, failure to understand the t of approxrmate locdl.ion of other leakase :on, and paths, the team \used the .quipr"ii t;Ъ ttaltti:漁 wrngs. especially designed Ind developeO f.i,: if,i, Although the new I take two As part of their research into how people purpose by the prince\,Risborough Labora- years to complete, results are behave tory (see BRE News 33). It was inticipateO _in .a fire, a team of scientisis? Surrey University has devised gu.n", in that the high standard\ana pre-fab;il;i;;

-./and.a agarn the result\were

Pfaying with

撚島

∬

dwellings.

do"rd;;

test individual reactions to a- fire. fhe board consists of an array of Uuttons anA lights representing a fire in a t ot"L fii"

"".r"ni

TttII[臨ご鷺犠ぷ瞥ぽ im 「i:電pf鳳へt¥濡

player is told that he is awoken Auringit-e night in his fifth floor room Uy .ttre siunC

i電

CSTC/BRE paper.

divided i

Survey on information needed for tendering

fire

the form of a fue scenario display"

construction of the test hbusei would make them.relatively .tighter' than most

British

be able

one or two projects which have

of breaking glass'- what would he do? He selects a course of action by pressing one of rhe buttons. A visual dirpluy f.urni then lights up to show the results'of thai choice and the state of the fue, presentinc him with his next choice. fne oi,tcomi oT his sequence of choices is shown ;1h;

board in terms of how successful _ or unsuccessful - his attempt to escape has been: whether he would irave got &

;: This- development forms part of a research contract placed by the Fir; harmed, been injured or even died.

Pre-construction information needs of contractors- will be the subject of a two-year

ng and tendering, difficulties in

彙』騰畷躍啜ey∬ 犠:1 EⅢ l眠ご a.ll info{nation probtems associated

Research Station, aimed at trying

out what people actually do in a

to find

ire

com-

pared with what they are commonty with expected to do. The success of support from the Institute of Buitding. Th any fue ぬe.Ы IⅢ nd sttncations ol~theF i轟 safety system depends survey follows a pilot study in wtricf, for/r not only on the k』 building's construction unO construction projects were Uui F also on the behaviour of the"onients through the estimating and peopte wt o occupy it. Research into human behaviour procedures. i1 fues is difficult because the fire is The problems revealed by the pilot physically complex, and because ,"ly are not new. They concern the ner small.minority of people have had dieci" better communication between ｂ d」 expenence. of the building team through iぶ The game is based on the detailed bills of quantities and specifilati 嘔 WhCh正 :3~c温缶品 analysis of an actual hotel fire in which problems are many, including f Ｔｈｅ」:晰 :11臨譜ヽ several people died, but the situation and survey

to be carried out by BRE

wit

讐撃臨愛曜■轟ぽ貰器￡

弾鞭盤簿

:轟 1難躍ゝヽ滉躙喩酔

Wind tunnel test on nevu' grandstands Models of three-new spectator stands pro_ posed for the Glasgow Rangers football ground at Ibrox park have been tested in tfe BRE Boundary Layer Wind tunnel. Ai the design height the stands will be taller than most of the surrounding buildings and will have to withstand coniiderable-winJ loads. There are no design data for grand_ stands in the Code of practice, th" designers, Thorburn and partneri, "n=a decided that this would have to be obtained from model-scale tests. Studies were made on a l/300 scale model including rhe existin! grandstand and terracing, tfie new standi which will be added in three stages, and ali the surrounding buildings up to a radius of 260 m. The data obtaineA from ttre tests, in particular the cladding loads, will be incorporated into the final designs. ' The designers also wanted to make sure that the new stahds will nor generate areas of high windspeed at ground level which

ノ

じF― クク′

ditκ nt wlり _Speeds On a thh layer Of painting a suspension of pigment in paraffin on the playing area and photo-

sand sprinkled over the playing- area. Except -at the very corners-of-thJ pitch,

the-streaking

caused by the wind. where the maximum wind speed incrlased gr.aphtnS rtow otrectrons were found to chanse by.,25 per cent, might- affect play. The wind flow was wrr glv_e-rncreasingthe phased devetopmeni progressively during the three stages of tf,e shelter over the ptaying development from rnvestigated by studying the effect of area. Wind the incideit wind directions were visualised b! direction to a circulatory pattern.

lossible actions have been greatly simpli_ to a person in a fire situation. Many cases fied. It illustrates the possible intlractions indicate that strange noises and even which can occur between people, fue and odours are frequently misinterpreted. the building, although it does not, of a wisp of smoke may be interpreted asThus the course, reproduce the stresses of a real fire family next door or upstairs cooking or the time scales available to take action. something, and the odd sound of glass _ The Surrey University team, led by smashing perhaps as yet another rumpus Dr David Canter, has studied a number oi outside the pub down the road. actual fires in dwellings, hotels and Once it has been established that there is hosp-itals, aiming to interview

involved

a[ the people a fire, the game demonstrates that a fue Uiigaae number of decisions may have to be taken

occupants, staff,

personnel, neighbours, passers-by, etc. The on inadequate information. A correct team found that people coutd remember choice of escape route at the beginning of much more than might be expected, and the game will lead to a 'clean' escape, but with little apparent distortion in spite of there is no information available to help

the time lag. From the analysis of the it has been found that the ability to recognise immediately that there lb a fue is one of the major factors in determining what will eventually happen interview data

make the choice. If the player engages on some delaying tactic, like getting dressed, a clear escape route may be blocked. The importance of time is thus demonstrated.

tions and conferences, the

prototype

display has generated a great deal of interest within fire brigades and the press. [n its present form the display can be used as an

educational tool to complement films, talks, demonstrations or lists of instruc-

tions by enabling people to participate directly in a fire situation. It has demonstrated that a game is a good toot for complementing the study of behaviour in fires. The. technique is being developed to cover a wide variety of building types, people, fire types and expected actions with several possible scenarios. Eventually it is hoped to link up the game with a computer so that actions by successive players can be recorded and also possibly to incorporate simulations of smoke movement

and other aspâ&#x201A;Źcts

Already shown at a number of exhibi- game.

of real fires into

the

This decision tree summarises, in simplified form, the various actions taken by the pâ&#x201A;Źopre caught in an actuar hoter fire. suppose you had been in the Sth-floor room marked in the ptan: what would you have done? start by selecting one of the g buttons in the first option array and follow your choice through to the final outcome to find out whether you would have escaped unharmed. First array of buttons

Outcome oJ tirst choice

Second array

of buttons

Oulcome of 3econd choico

lhird

array

Fina:

of buttons

outcomo

End of game

fl] El

Ctean

escapemas$stance nmdcd

Assrsred dowrladdet

St,ghr,nlulerperhaps shosk

@j @l

Subsranrral rnlurra

Serere rrlury.pshap3daath

鸞` ・ dr鳥筆

留 Aυ ‖EED STATES DEPARTME‖ T OF

NBS TEGHNICAL NOTE

00MMERCE

PUBLiCAT10N

DEPARTMENT OF COMMERCE National

Bureau Of

Standards

8I8

NATIONAL BUREAU OF STANDARDS

The National Bureau of Standards' was established by an act of Congress March 3, 190t. The Bureau's overall goal is to strengthen and advance the Nation's science and technology and facilitate their effective application for pubtic benefit. To this end, the Bureau conducts

research and provides: (l) a basis for the Nation's physicat measurement system, (2) scientific and technological services for industry and government, (3 ) a technicat basis for equity in trade, and (4) technical services to promote public safety. The Bureau consisti of the Institute for Basic Standards, the Institute for Materials Research, the Institute for Applied Technology, the Institute for Computer Sciences and Technology, and the Office for Information Programs.

THE |NSTITUTE FOR BASIC STANDARDS provides the central

basis within the United States of a complete and consistent system of physical measurement; coordinates that system with measurement systems of other nations; and furnishes essâ&#x201A;Źntial services leading to accurate and uniform physical measurements throughout the Nation's scientific community, industry, and commerce. The Institute consists of a Center for Radiation Research, an Office of Measurement Services and the following divisions: physics Applied Mathematics Electricity Mechanics Nuclear - Heat - optical Sciences" Applied -Radiationr -Quantum Electronics'Electromaglreticsr Time - 3 Laboratory Astrophysig5 3 and Frequency Cryogenics ".

THE INSTITUTE FOR IIIATERIAIS RESEARCH conducts materials research leading to improved methods of measurement, standards, and data on the properties of well-characterized materials needed by industry, commerce, educational institutions, and Government; provides advisory and.,Iesearch services to other Government agencies; and develops, produces, and distributes standard reference materiats. The Institute consists of the Office of Standard Reference Materials and the following divisions:

Analytical chemistry Radiation

Polymers

Physical Chemistry.

Metallurgy

Inorganic Materiats

Reactor

THE INSTITUTE FOR APPLIED TECHNOLOGY provi<les technical services to promote the use of available technology and to facilitate technological innovation in industry and

Government; cboperates witn puUIic and private organizations teading to the development of technological standards (including mandatory safety standards), codes and methods of test: and provides technical advice and services to Gove'rnment agencies upon request. The Institute consists of a Center for Building Technology and the foilowing divisions and offices: Engineering and Product Standards Weights and Measures Invention and Innovation Product Evaluation Technology Electronic Technology Technical Analysis Measurement Engineering structures, Materials, and Life- safety' Building -Environment { Technical Evaluation and Applicatisn. Fire Technology.-

THE INSTITUTE FOR COMPUTER SCIENCES AND TECHNOLOGY conducts research and provides technical services designed to aid Government agencies in improving cost effectiveness in the conduct of their programs through the setection, acquisitibn, and effective

utilization of automatic data processing equipment; and serves as the principal focus within the executivq branch for the developmenr of Federal standards for automatic data processing equipment, techniqires, and computer languages. The Institute consists of the followin! divisions: C.omputer Services

tion Technology.

Systems and Software

Computer Systems Engineering

Informa-

THE OFFICE FOR INFORMATION PROGRAIVIS promotes optimum dissemination and accessibility of scieotific information generated within NBS and other agencies of the Federal Government; promotes the development of the National Standard Reference Data System and a system of information analysis centers dealing with the broader aspects of ,tbe National

Measurement'System; provides appropriate services to ensure that the NBS staff'has optimum acccssibility to the scientific information of the world. The Office consists of the following organizational units: Office of Standard Reference Data Office of Information Activities Office of Technical Publications Library Office- of International Retaiions_

Hcadqusrtcrs ald Iiboratotics at Gaithersburg, Maryland, untcss othcrwisc noted; ---r Washington, D.C. 2t234. I Part of thc Center for Radiation Rcscarch. t Locatcd at Boulder, Colorado 80302. . Part of the Ccnter for Building Tcchnology,

mailint

eddrcss

â&#x2013;

Occupant Behauior in Building Fires

Arthur I. Rubin i

Buililing Environment Division Center for Buildiug TechnologY and

Arthur

Cohen

Technical Analysie Division

Inetitute for Applieil Technology National Bureau of Standards 'Vashington, D.C. 20234

U.S. DEPARTMENT OF COMMERCE, Frederick B. Dent, Secrelory NATIONAL BUREAU OF STANDARDS, Richord W. Robcrtr, Director fssued

tebruory 1974

Nadonal Bureau of Standards Technical Note 818 NaL K、

Stand。 (UoSめ ,TedL

Note 31鶴 28Pages Oreb.1974)

CODEN8 NBTNAE

もヽ、COVERNMENT

PRI-0 0FFICE

WASIINCTON8 1974

For 8ale

Бシthe supe」 ntendent of Documente.usocovemment hdコ嘔o鍮喝 V30LingtonD DK■ 20402 :81め .ndce 60 CenL 40rder by SD Catalog No.C13

CONTENTS、

Page

工NTRODUCT10N

GENERAL BACKGROUND 。。 . . . ● ● ● ● ● ● ● ● ● 2。 1。 Performance Approach . . 。 . . . . . 0 0

３

e . .

。。 ● ●o ● ● ● ● ● ● ● ●0 ● ● ● ● ● ● ● ● ◆0 ● ●o ●o ●

STUDY OB」 ECT工 VES AND METHOD OF APPROACH

● ● ● ● ● ●0 ●o

● ● ● ● ● ● ●o

● ●0 ● ●o ●

. . . . . . .

。o

● ● ●o

o o o o O

o o

● ●o ● ●

４

1 ２４

５

L工 TERATURE

REVIEW

。。.

. . . . . .

● ● ●0

4.1. Popular rAccounts of Fl re Emer.gencies . . 4.2. Non-Behavloral Inforna tion .

4.3. Behavioral 4.4。

● ● ● ● ● ● ●0 ● ● ● ● ●o ● ●

６６６７

Regearch

Other Sources of lnformation . . . . . . . . .

● ● 0 ● ● ● ●0

● ●●●

９

RELATED PROBLEMS . . . . . . 。。 0 ● ● ● 0 5.1. Stress Research . . . . . 。。。 0 0

5.2. Panic Behavlor . 5.3. threat Behavior . . . . . . . . . 5.4. Dlgaster Studlee . . . .

6。

PROBLEM L■ mTAT10NS AND SCOPE 6。 1。

6.2.

● ● ● 0 0 ● o o ● ● 0 ● ● ● ● ● ●o ● ● 0 ● ● ● ● ●0 ● ● ● ● ● ● ● ● ●

９９１０１１

１１

. . . .

Design/Managetnent Probletns e .

Individual Behavioral Problems

４。 ● ● ● ● ● ● ● ● ● ●0 ● ●0 0 0 ● ● ● １１５ . . 。。。. . 0 ●0 0 0 ●0 0 0 ● ● ● ● ● ‐ Psychologica l . . . 。 . . 。 ● ● ● ● ● ● １８

RAT10NALE FOR A RESEARCH APPROACH . . . . 。 . . . ● 0 0 ● ● ● 0 7。 1。 Human Factors Applications for Fire saFety‐ Recommendation3

● 0 ● 0 ● ● ● １９。。 . . 。 o o ２１

BIBL■ OGRAPHY 。。 .

o O o

. .

● ● ●o● 0 0●

0 0

● ● ● ● ● ● ● ● ● ●0 ●O

lli

●●●●

OCCUPANT

SAFEIY

BTIILDING FIRES

Arthur I. Rubin and Arthur

Cohen

Flre oafety

buirdlngs ls important in building design and the formulation of'ncodes ani standards. no"".r"a, an er<amination of the information concernlng the nee{s oi o""up"iis

in fire emergencies (as opposed to ensuri-ng a'ae!re;-;Fstructural integrity for the building) indicates thJt the"sclentlflc rnrorroairon' base is woei.rrry lnadequate. The increasing prorninence of tigt ir"_" buildings havrng EEny occupants lntensifles the need ror uetili ^Fire information about the behavior of-occu.pants during fi.re eurerge;;i;;: researchers have indrcated that it is oiten not feaslbr" io L.r"".r"te buildings because of time constraints. rnstead, designeis use techniques guch as safe areas within buildlngs aia-."qiiri"g people to respond differentially, based on thelr particurar iocation. This approach enrphasi-zes

cornnunications and warning systems to transmit messages. as roud nolses or-brinktng rights should be designed to ,take advantage, of the usual respon."" can actrvery be a part of the Jlre w".rrirrl ,ia"-iy peopte. occupants lnstead of berng ur,wllltttg'.rrctiis."-rti" possibiltty "iJ righarrrg and a human factors approach is taken to suggest some meansis pursued'y'tem of better ur,a"."i"rriing the capabilltles of occupants. Ttrese s]-stems such

Key

l,rords: Disaster research.r high rise building fires; occupant safety. 'r*ooouarro"

A fire breaks out in a buildingl

do the occupants typtcally find out about lt? llhat _is their reaction to the fire emergency? How

How

do they usually escape?

Are present fire safety systems in bulldings responsive

to the occupants. needs?

in anT::pl:::::[e l:li』 :llicin::la::duξ :d :;r:i:s:h::ic:li::i

:: :::::::if::i:ileilii::::[:il safety in bulldings.

nS and prOv■ de informatiOn

lli::ill:￨:rili:::::l:'1:::]:ple

The resel f

i:I:::::::;'8;:::I:iS i:da:::11 the prOblem Was already accOmpl

Cus primarily On the

t ]I:::￨:il::1le:￢

:I:11::f:III:nt

DiscussiOns cOnducted with members of the staff of the NatiOnal Bureau of standards and others familiar with fire research activities indicated that the assumptiOns made i::i::::lili:i:iよ :i:i:::ii::li::lili::::i:::f:[:b::lye::ξ

approplille l::Iodttalla:::le:[ili:ti:::t::。

::I]il:bl:::I::"lul:s::I:htir::ian

:::i::a::Pali:L。 [h:Ii:::::i:lil:il::::::ii::e

eslhe

f:1:::::a:::1::1:ji::・ i:::::tilh:S: ]:::ilifile:a::j::::↓

Safety problems warranting Sepali::::i:ili:i::i:;h::::a:ntli:le::i: :u::11:i:it :::8:: fire

It is lmportant to note at the outset that fire research associated with the behavior of people during emergencles poses a dilenrma for the researchet. It ts not'feasible (or desirable) to artiflclally establish an experimental sttuatlon where subjects are in any real danger. ltre data must therefore be developed as real situations arise with the researcher having virtually no control over the I'experimentrt being conducted. Ihe best that can be done is to make careful and systematic observatlons and measurements of events Judged to be critical. 2.

GENEML BACKGROT'ND

Ttre flre problem is one of stgnificant magnitude and many conflictlng facets. Although this study does not purport to investlgate the entire range of problernsr the following cornments from Ahern and Morgan (1) and Amerlca Burning (2) provide an indlcation of the studyrs scope and ramlficatlons: . More than 121000 persons have been killed in fires in each of the past slx years ln the U.S. (1). Only motor vehicle accidents and falls rank higher arnong the causes of accidental death (2>.

'In the Unlted States, the death rate attributed to flres on a per capita basis is twlce that of Canada, four times that of the United Kingdom, and 6 l/2 tlnes that of Japan (1). . As a n€tlon, the United States spent about $5.5 billion tn 1971 on flre-related actlvities: $2.7 billion on fire protection, $2.8 billton due to ProPerty - losses (1). . In 4gg!g4s!!g (2) one finds that the United states leads all the rnaJor tttatretii-attzeanations in per capita deathe as well as property lbss froo fire. ltre total cost of destructive fire is estimated to be $11.0 blllion Per year. by sorne experts. Both of these estimates are thought to be .rygllgg There are many fire-related problems which become increasingly critical in htgh-rise bulldings3 distance to exists, safe areas wLthln the bullding, fire control and extinguishment, detection and alarm, etc. Ttrese factors ralse a number of questions. Is total evacuatLon wlthin reasonable tirne llmits possible in a high-rise butlding? Hqr doee one deal wlth disabled occupants? Are exite appropriately designed-? Are conventional fire alarm systemrt adequate for high-rise buil.dings? As a result of the fire problem, more governmental and publtc atterltlon ie being given to try to control, and perhaps reduce, the needless personal suffering and human loss from ftre. I$ere ls sox1e-evidence that efforts designed to imProve the exlsting sltu'ation during the past several years have been somewhat effective although ln the case ofrhigh,rise builiings the problemg reoal.n severe. In the words of Ahern and Morgan (l)rof the National Flre protectlon Agsoilatlon (NFPA) rrNo single organization or grouP of organizatlons has been responslbte for the galns. A large part of the credit oust 8o to ProPerty orners and r"r,"g"tr, from corporate heads to small homeowners, who have had'the interest and taken the lnltiatlve to do soroething about fire protection and preventl.on in aieae under their authority. !t Aelde from organizations, other contributlng factors have an impact on flre Protectlon. There are new methods and materiats avaitable to the architects and builders who are responsLble for buildlng construction. lhese heve resulted in improvenents in some instancea whlle adding to the problen in other cases, especlally with tespect to smoke generatlon assoclated wlth some products. Finally, there have been some iuprovements in flre department apparatus, although serl.ous deflciencies exlgt in the abillty to fight high rtse flres.

lFigrrr"" in parentheses lndlcate the titerature references at the end of thls paper.

一ヽ環・一〓一ｔｔ ¨ ■ ． ● ヽお驚薔慧選還饉爵覆畿踊機鴻亀．．

ぎ書〓・ｉ量魔・ヽ一 ≡ ．一 ■ 響

[:::;F::::::i::al::i:i::i:::i::::::::::::::i:lli::hal:tu::I:I:::1:iyh:I:uliintini::::i18 8。

lall::;8f::1::ョ [::id[::ilfi:;:::fi:I::::81::I:￨

However' the buildlng industry today is somewhat different from what it has been in the past. Although many builders appear to be following a I'business as usualr' approach, there is conslderable impetus to speed up the con3truction process to neet the denand for more and better homes. As a result of tlUDrs Operation BREAKTIIROUGH and similar programs, a variety of lnnovations have been introduced. these innovatlons encompass materials, comPonents, bullding subsystems and housing construction technology in general. In some ways, they can be consldered revolutionary in their impact on fire safety research and applicatlons. For example, the slorly accruing empirical data base associated with fire safety may become rapidly obsolete when innovatlve buildings are under evaluation, because tnany major deslgn psrameters w111 differ from those traditionally enployed. If it is not posslble to tely on past experience, innovative design solutions should- be evaluated froro the etandpolnt of fire sefety of the occupants. ' 2.L.'Performance Approach

ltre rate of appllcation of building-related innovations, coupled with the need for more better houslng, has resulted in a sense of frustration among those concerned wlth applying these new technologies to bullding problems. The lnstitutions and processes assocLated with building construction activities appear to be designed to resist change. Ttre traditionat methods of deslgnlng and testing building components and materiats make use of tlne conguming, parametrlc, tria_l and error procedures. After these tests have dernonstrated the feaslblltty of an-innovatiori, building code rnodifications are often necessary, before builders can apply this nell technique. . lAese modlfications further lengthen the tlne span between the development of an innovation and its ultimate large-scale use. As e resutt of this traditional procedure, many domestlc and foreign organizations concerned with buildings have become dissatisfied with the rtstatus-quorr and have taken a new look at the methods currently used by butlding deslgners. ltrls reexamination has reEulted lrr the use of the concept of performance, based upon the requirements of the users. Slnce ultinately the function of butldlngs ls to serve their occupants, performance requirements Eust be defined in terms of user needs and wants. and

As generally used by building research organizatlons ln the United States and abroad, the perforrnance concept begins with a. performance statement whlch has three essential partg requirement, criterlon and test. A requirement is a qualitatlve statement which ldentifies a human need. Ilre crlterion ls a quantltative desctiption o.f a requirement. It provides specific levele for attaining conpliance wlth the lntent of the requirement. lh. !gg.!. lndlcateg a method to assess the degree of compliance with the criterion. lae methods Lnclude analysis, physlcal measurement, snd expert Judpent. the prinary purposes of the performance approach are to point the way for the developnent of netr products and techniques, and to promote impartiality to the acceptance of lnnovationg. Lt was formulated particularly to overcome the constralnts inherent ln the prescriptlve approach whlch largely stifles Lnnovative solutions.

Inherent ln the'perfomance approach ls the use of a systems orientation to the solution of probleos. thle situatlon ls especially true when de consider the llfe safety systen of a buildtng where the primary requirement lg the adequacy of all subsystems in minlmizing tnjury and death due to a fire emergency. The performance approach fot BREAKIHROUGts houstng.recognized this problem and iurposed performgnce requireuents for all of the folloring fire safety characterlstlcs: detectlon, 8larm, exlting, design of passagenays, and control of materials.

!!r. lI. Finger, a former Assistant Secretary for Research and Teehnology for HIID, at a wlth the performance concept in buildings (17) pinpointed several speclfic probleos assoclated with fire safety of occupants. He noted that the standard ASI,I /1119 fl.re test (Fire Test of Bulldtng Constructlon and l,laterials) makes no provision for measurLng the amount of enoke and toxic gasea assoclated wlth the naterlals used ln constructlon. As long es tradltional materials are uaed, the test is valld, but wlth the lntroductlon of nerr materlals such ae flberglass-re.Lnforced polyester, neq, problems arise but the tradittonal test method renbins .ln use desplte Lts evident ahortcooings. Ttris discrepancy becomes especially eig-nificant whenwe consLder that lnnovation.is now being encouraged in the use of chemical and plastic construction naterlals rind that smoke and toxic gases have always been a major oource of inJury and death ln fires. !8. Finger also indicated that generally, flre alarns have not been partLcular_ly effectlve b_ecause the lncidence of drills snd false alarme encourages a feellng of-safety rather than danger anong.occupants. Ee then slimposium deallng

二重鷺層著麟曹塁蓮ｌ心

ｔ，■ 凛沐ゞけ，■ 求會せ事ず一轟導儀ざ

questioned the adequacy of a be1l atarm system as compared uith other possible approachesvisual signals, voice corrnunication. He concluded that more knowledge was needed in ttre design of warning systems.

Mr. A. F. Sampson, Conmissioner, GSA, in a talk recently rnade at the InternaLional Fire Safety in High-Rise Buildings dealt v,ith the design, use and management of buildings as they relate to fire safety: lAs all of you know, the present system of codes, standards, and criteria have been developed over the years and certainly contain many sound requirements. However, we can at1 agree that the present concept of unitized performance requirements (i..e., a door, a beam, the wldth of a corridor) fitting together by a series of go/no-go acceptance requirements lnvolved both safety inadequacies and elements of overdesign. We fa11 down because qte usually do not have a total safety system for the total butldtng. Analysis or attack on this problem ls difflcutt because of the Lack of an identlfied goal (a way of identifying a level of safety in a building) and a way of measuring this goal. In many instances present flre safety concepts and requirements are grossly conceived and grossly applied in a manner that provides an overall, though undefined, degree of safety in-a typical building. This results in a sLtuation $here lt ls virtually lmpossible to evaluate the potential of a specific eLement or the inpact of a proposed change. Itris inabtlity to evetuate the potential of change is a major factor in both the plesent problem and in the frequent charges that fire safety professionals are reluctant to accept new deslgn concepts and materials and ale an obstructlon to progress. Building design, use and operational concepts must change rapidly and drarnatically to be responsive to current requirements. Stil1 another maJor inpedirnent is the well known time consuming process reguired to changed any code -and fire regulations are not an exception. Conference on

Ttris presents a dllenrna. Most current fire safety requlrements are based on experience and since building concepts are rapidly changing, much of this experience is of questionable value. Fire safety planning must have a better base on which we can predict the success of a design or the potential risk in an existlng buildini, and from this evaluation, develop the appropriate actions to achleve the desired level of safety. the answer to this dilenrma rests in the development of sound basic concepts and fundamentals that w111 a11oq, the analytical measurement of the 1evel of sif.ty and comparative evaluation of alternatives. lhis will provide truly realistic evaluation of alternatives. thts will provide truly iealistic inputs to the ultimate problern solutlon.t'(20).

Unfortunately these and other user requirements are often described by qualitative statements of the attributes desired frorn buildings. In order for the performance approach to work, these qualitative statements must be translated into quantitatively defined scienttfic engineering statements which fu1fi11 the loeanlng and lntent of the human needs.

STI'DY OBJECTI\NS AND I,IEIUOD OF APPROACII

the objectives of the present program are both long and short range in nature. Xtre long term obJectives of the work are to provide flre safety criteria in performance terms based defensible quantitatlve data base describlng user requirements. Howupon a lcientifieally ever, it is recognlzla ttt"t this ioal is very ambitlous and that a great many lntermediate steps are required. perhaps the iirst step ln.this process is to better understand the many problemj associated with the response-of pLople to iire emergencies. The process of ,rpgt"ifrrg our culrent data base ls a lengthy one and in the meanwhlle, there are fire emergencies occurring daily and it ls necessary to devise interim solutlons. As the flrst etudy of a research ftogr"., one of the prlmary obJectives $as to better understand the probiems and then suggest techniques of coplng with then.

In determining the ttstate-of-the-artrt in the problem arear a number of standard .A literature gearch of both technical and non-technical approaches *"t".rfloyed. publications was made. Interviews erere conducted with researchers as well as adminlstratorg expertd in fire 6afety. Correspondence sras also initlated with several foreign rho "t" reeearchers who are pursuing siurilar goals.

4" LIIEMII,RE

REVIEI.J

The information compiled durlng the literature survey presents a mixture of findings. Eowever, since the rnajor study objective was defined in very general terms, and we consider this work to be largely that of trying to first define the problern, this resutt is not very surprising. lAe source material examined ranged from Itpopularr accounts of flre emergency behavior, stress behavior in laboratory investigations, disaster studies, to theoretical work designed to improve existing conrnunications capabilities in fire fighting actlvitles. Because of the dlversity of source material, it is difflcult to identlfy corcnon themes which can be employed to coherently organize. this section of the report. Consequently, several topics have been ldentifled as being of partlcuLar relevance and wilt be treated separately. Whenever lnterrelationships can be ldentified, they will be explored in some deta i 1.

4.1. Popular Accounts of Flre

Emergencies

A major source of information was thought to be anecCotal and narratl.ve accounts of fires appearing in the popular press, magazines and other publicatlons. Fires are an everyday occurrence in most cornnunities and receive extensive treatment in the neus medl.a for several reasons. they often are spectacular visually and therefore lend themseLves readily to photographs and televiaion coverage. Fear of fire seems to be a unlversal phenomenon and when combined rcith the perceived vulnerability of most homes to fire, it can be readily anticlpated that accounts of fires creâ&#x201A;Źte strong empathic reactions among vl.ewers and readers. In other words, they are ner^rsr.rorthy events from several standpoints.

lAere is no disagreement about the avaLlability of information from t6e news media. The major probLems are the technical adequacy of the information for research purposes and the anticipated payoff for a determined effort to explore these sources. The sheer mass of available information poses a forrnidabLe obstacle because its form does not lend itself readily to compilation. Rather, in most instances it is a case history gf a single event, resulting from an intervielr designed by a reporter seeking a "good storytr. Simple, straightforward and objective accounts of behavior (which might prove useful to researchers)

are ltkely to be considered not nehrsworthy and would therefore not appear in print. Investlgators concerned with fire safety have been very reluctant to use nelrspaPer accounts (6, 22> ln their studies. Interviews conducted with dlsaster study researchers have resulted in similar conclusions about the adequacy of the press information for research purposes. Iheir primary caution Ls concerned with the urajor sourees of bias inherent in collecting information for the mass media: the selection of cbntent as lre1l as format, and the inability to draw conclusions about the representativeness of the dvailable sampLe of behavior.

ttre personal drama .inherent in a major fire also lends itself to extended narratLve treatment, in the form of both non-fiction and fictional accounts of disasters. Ttre account of the Coconut Grove fire (4) ls an especially compelling exarnple of this type of treatment. the author expended conslderable effort in detailing nany of the factors adsociated with the fire and describing the activities of many of these who experienced the fire. I{o$evert despite this evidently exhaustlve description of the fire. the book is not useful. as a scientific source document. Although rnany individual and group behaviors are treated Ln some detail, it is almost impossible to attribute any causat links between environmental (flre related) events and individual or group behavior. lltre inferences nade by Benzaqul.n are based prirnarily on a single fire. lbey are reasonable, but not researchable in any meaningful way since it is extremety difficult to translate the individual circumstances described into generalizable and testabLe hypotheses.

4.2. Non-Behavioral Information lhe s- dy of fire research programs was limited to those which focused prlmarily on the behavior .f people in fire emergency situatlons. Obvioustyr this undertaking is enormous in ^.:self. Only a curgory examination was made of some general reports whlch summarized the current status of the t'fire problemt'. .ltre obJectl.ve was to detâ&#x201A;Źrmlne the 6cope of the problem by obtaining relevant statistics and to develop an insight lntp the type of data available and its applicability fof, use in'-d'esigning Programs which are directed toward studying human behavior. 6

The Federal Fire council produces a summary of federar fire'losses annually, which indicates the number of fires, dollar 1g:::", i"3"ii"" and deaths in fires associared with Federal Government property. The t967-;"";;i-;;;ort (1r) pio'ioes case history studies of rrajor fires and iteti" in some detail the causes and"r"o progress of the fires. the most recent Directory FT: Research (21) presents- a comprehensive overview fire research activities in tfre"funiiea of states. e., volume is revealing. rt demonstrates the lack of irri"* of the subjects incruded in the attention to the human behavioral component of the problem. It is as follows:

- Chemical aspects of fire - Gombustion engineering - Detection of fires (primarily - Fire

hardware)

damage and salvage

- Instrumentatlon and fire . Meteorologl.cal

equipment

- Model studies and scaling - Operations research

laws

and mathematical methods

- Physlcal aspecte of flres - Physiologlcal and psychological aspects of fires - (of 34 studies problems of building occupants)

none mentlon

- Prevention of fires - Propagatlon of fires - Suppression of fires 4.3. . Behavioral

Research

rhe material prevlously covered ln-this be described as being only tangentially related to human behavior in firereport can best Horvever, it is hoped that this inforroation proves "r"rg"rr"i"". a perspective for those few studies which have been -useful in providing directed toward understanding human behavior in fire emergency situations. Behaviorar tesPonses to fires has been treated depth by tero researchers, Dr' J' L. Bryan of the university ofMaryland, Mr. J.inL.considerable pauls of the iutlding use section from the National Research Council of Canada. Iheir work serves as a,ajor Uasfs for the follorlng discussion.

Dr. Bryan has focused his attention on panic behavior (6 and 7), but in contrast to the Disaster studiee of panic (see sec, s.g), ne has concentrsted on fire emergencies. Ttrerefore the research I's covered in this r""tion. iry"r, cr"irs panic is rnanifested in two different !'ays :- the lndividuar response of a person iuring a threat, or a râ&#x201A;Źaction by a group of people causing a crolsd panLc response. He notes that the responses of lndividuals are a withdrawal reaction and a hyperactivetwo extreme behavioral reaction which elght be a prereguisite to runntng from danger. Dr. Bryan conducted a study in the aftermath of the Arundel park which he interviewed eurviving victims of the fire. IIe was concerned Hall fire (6), during with the effectlveness of the illunlnation and urarkings of exits in the building in leadlng people to safety. Approxlmately 867" of the people interviewed did not notlce whether exit lights $ere present and ln working conditlon. Especially interesting is the fact that eight piotessional flrenen and policemen lrere Present, and they weri unable to recal1 wtrettrei they noticed these llghts.

In a study perforned for Dr. Bryan by l.l. L. Cannon (8), an interview was conducted in a departurent store to determlne the effectlveness of exit oarkings. The priBary purpose of the investigatlon.was to determine the degree of publlc awareness of exit locations and of the sensitivlty to danger of fires ln departurent stores. Almost 927" of. the shoppers interviewed rrere not at all concerned about the possibility of a fire. Darley and Iatane (10) have developed an experlmental situatlon wher.e an emergency situatlon ls simulated. r.{hile subJects are involved in conpleting a guestionnaire, smoke ls introduced lnto the test room. If a subject leaves the room and reports the smoke, the .experlment is tenninated. Ttre authors q,ere prlmarity interested in determlnlng the effects of social pressure on the responses of a subject to the simulated eoergehcy. ltrelr results were quite drametic. lAey found that subjects alone in a room reported the srnoke 757" of the tirne. l{hen tno passlve individuals were ln the room with the subJect, only 107" of the subjects reported the presencg of smoke. The authors of the study suggest that the snoke represented an anbiguous threat cue. Under these circumstances, the behavlor of an tndtvidual is readlly influenced by others around him. Therefore, when the other occupante of the roocr do not respond to the smoLe, the subject wlll tend tc interpret the smoke as being nonthreatenLng and will ignore iis possible danger. Gatbreath (13) nade a survey of ten office buildings at least seven stories in helght in order to determLne how tong it pould take occupants to evacuate the buildings in a fl.re emergency. (Data were collected during a series of fire drills.) His study was performed in the context of evaluating the adequacy of some of the f.ire safety requirements of the Natlonal Building Gode of Canada. Els naJor experimental interests concerned the number of occupants as well as the exit and the stair widths in the bulldings. Ee found that the drop ln travel rate eith increasing denslty ln the stainrell posed a major problem durlng evacuatLon whtch took as long as 12 minutes ln'some lnstances. On the basis of hLs lnvestigation he concluded that, trit is not realistic to attempt to provide for evacuatton of high buildings by stairs in a llmlted time.t' He noted that hls findings wete in close €greement. wlth a elnllar etudy performed for the London Transport Board.

J. L. Pauls explored the relationship betveen b-ehavloral requirements and design eolutions in the emergency'-egress'of buildlngs (22). this study, which ls very sinitar ln lts prinary objective to the present lnvestigatiotr, w111 serve as the rnajor reference source for the rdmainder of this -report stnce the treatment of the froblem by Mr. Pauls was the rnost detailed and relevant account found. Ttre maJor focus of attentlon was the hlgh-rlse offlce bulldlng. ltre investigatorrs major concern vras to..r'e;plore how emergency'provisions night be deslgned tf the designer worked logically from clearly stated premLses or principles through to design conclusions..... Adplying this concept to design of emergency provisions, the designer nould have to- clarify the behavioral goals relevant to the occupancy and building characteristics. He would also need theoretical knowledge of some form of tdsting to be able to predict whether an environment would satisfy the behavioral goals.rl Itr. Pauls ldentified three classes of behavioral goals: 1. Traditlonal Rapld Egress from Buildings. -- Within a few mlnutes after a fire danger alaru. 2. Ttre tlluster-Sle.lE:Leqtr-.lg9ngegq. -- Occupants assemble at various polnts ln the building before taking further acJion. 3. Nin-Egrese. -- Most occupants would suffer only mlnimat dlsruptlon of usuil activitl.es because of the localizatlon of emergency conditlon. For example, when an elevator ts temporarlly renoved from servlce for lnspection or repair PurPoge9.

l.tr. Pauls perforned a number of Lnvestigatiorrs durlng fire drllls to determlne whet&er auditory alar,m slgnals ehould be suppleurented by visual signats ln facilitatlng egress frotn bulldlngs. Ee adnlnlstered questlonnaires to occupants, made observations of behavior, and varied th6 ltghting conditlons by tnak{ng the lighte fllcker. Although his findtngs were lndeteminater.the rationale and approach ate intetestlng and wtl1 be treated. ln eoue detalt ln the discusslon sections of this report (22). 〓導甚鷲ふ雄一ゞ遂繹子

4.4. Other Sources of Information rn an effort to augment the lnformatlon which was avallable in the open llterature, a number of discussions were conducted with NBS Fire Research personnel members of the criginal Disaster Research Group from the National Academy oi sci".r"u"and (NAS) who $rere still engaged in similar research sctivities. As a result of these discusiio;" it b"".*" evident 'that little current research was directed toward the probrem of concern.ih" Disaster Research center at ohio state unlverslty has agsfunilaied many of- the functions performed by the disbanded Dlsaster Research Group of the NAS, but their interests are primarily concerned !tlth developing improved connnunicatlons among the organizations concernea wittr combatting fil:": - they employ'a systems orientatlon to bette! understand the necessary interactions of lndividual's and organlzational units (whether government ot private) during the many processes associated wlth fire emergencies (detectlon, control, evacuati.on, et".). ' contact was also nade wlth a number or research organizations Ln foreign countries. These include:

Tte Japan Fire frevention Equipnrent rnspection Association Japanese

Building Research Institute

Natlonal Swedlsh Institute for Buildlng Natlonal Research Councll of

.Research

Canada

Flre Research Station, Eerts, United

Klngdom

Unfortunately, these contacts lfere no more frultful in identifying relevant research - were than the dlscussions. Hcrirever a number of questlonnaires laesignJa by others) were made avallable. lttese provided a starting point for the present study of tehavioral teltPonses during fires. Ttre type of study ls typtfied by a report by the Fire Research Statlon in the United Kingdom (9) which provides sunrnâ&#x201A;Źrry information on flre deaths in 1971. The data are Plesented in tabular form and are of the type available from insurance companies and the Federal Fire Council in the United States. 5.

Directory

Cooperative Buying/

８。８４８５８５８６９２

Lei3b COSby Assistant Art Director

Carl Toner Advertising Manager

Wo C.Klrby

ProductiOn Technician

Mtt Wentworb Senior Secretary

Jody HOmiter

Boad of Directots Chairman

Dcnnis Pnitt Executive Dircctor Gatl Englkh Assistant Dircctor for Publications and Communications Stcvc Shglc Administrative Services Coordinator

In This lssue

．２ ■

Production and Art Director

Facililies managers, campus programmefi; ond arts p,esenlers con benetit from examining this exemplory opproach to

DEPARTMENTS ４７

Do●●etto Hd色にhmldt

Uncovering the Iceberg: The Feasibility of an Arts Facility

Erecutive Director's Comment

距

Tour Schedules Anist/Agency

Reports Lecture Reports School Reports Sound & Lighting Reports Travei Reports

Lm 080ロ

Norrer l*oprrd

Marketing Services Coordinator Rich Mrit Educational Services Coordinator

Managing the Student Work Force

Donnctl. Hritschmidt Computer Services Coordinator

Crrclinc Fler|tmi||8

student employment opportunitiu in lhe camqus union can be a legitimate Part of on institution's total

Resource Serviccs l,ce Spcncc Studcnr Activitic Programmint (lSSNq) bv 9t1664) is Dublishcd ninc tim6 a NECAA, (January/Februar), March. April. ttay, Summcr, Septcmber, Ocaober'

educational pnogmm. by Jack Smith

Noicmbcr. Deccmbcr), cxciu!ircly for

NECAA mcmbcF. CopyriShto l9tl by Na' iional Entenainment and Campus Activitics Association, Editorial, pubitshlnS and advcr' tilinr officcs: 749 Saluda Avmue. Co|[mbis'

霊

疑品 :T船般 ibf出1器

scriDtion ',鋭 information and other (orr6pondcncc: Programming, Bor I l'169. Coiumbra'

sc 292I1.-NECAA full mcmbcship is re'

Advertisers' Index

itricted to institulions of hiSher lcarnirg: 975 of thcir membership fe ir lbr iivc:ubscrip'

rrons (!20 each) lo ProStamming. A5sc'ia(c m€mbctshtp is r€srrictcd to iinn5 $hoie talcol'

COVER NOTE Illusuation by Leigh CosbY

Causes and Prevention of Crowd Disasters ra

″ ′grso″ ″ ψ り α′

棚脇協郷 sysra‖ αr,c appraα

crt rο

1麟

′ Oι cro w′ cO″ ′

0【 0

幽

ヽation31 E■ tenal● ment● nd Campus A● ::、 ities Assocl,tion Box ll■ 0.Coiumbiao SC 19211 Teleplone 303‐ 799_o768

by John J.Fmin

Drodudt, broglam5 s3 5slvice arc dircctly ielated to the field of colleS,iatc crlrlcurrlcular

acii"irie: uJ to f7s of thcit mcmbcrship fc

(i:0

Progrr$mlng each) F.orctional mcmbership is 130 pcr ..r?-iJi.iiir ana f25 for ttudcnts' tlO ol rh? for subscriotions

i*""iitii.

'r* it tot

onc lubsariptign to Progrlmmlng

ffid'ffi ni:ffit*iffii+h*i

し」θ/ァ

・一・

TIさ

IN「 ()Rヽ lAT10NAL

REPORT

と [〕外91 ″ い早し牟￢れレ a￢ ι %η 6 F｀ C135 "ν

Cliaracteristics 'and Seryice R equirements of Pedestrians and Pedestrian Facilities Bv lTE Tcchnical Council Committee 5‐

D. Navin (lt); Stcven C. Protost (A):

ccntrulcd areas sucb as prssSgcla\s. strir$a\s. plazas. sidcnllks. etc.. and lo inrestigate thc {lorr characterislics of pe-

thc aid ol John J. Fruin. matcrial from

A Caution. To a great extenti Committec 5-R has adopted Dr. John J. Frurn's concepts and methodology in thc identiiicutron oI pedestrian levels of scrvice. While it is felt that these are

total plrrnning. design and control. Since the objcctircs of thc committcc did not corer lltis total rienpoint. plannirrg. de-

acknos ledged.

only iocal practitioners could aflirm or denv. There will also be national differ' ences based on the differing reactions in various ethnic groups to the concept of

hc objectives of Commiltee 5-R were to prepare an irtformationrl rcport relating lo lhc characleristics and scrrice requircments (rf thc movt'tnent of pcople in cr)n1

destrians. inclurling spccd. r olume and densitl or its t'quiralent' llotrevcr. these s:ipccts nrust be considcred itt tcrms of

sign and control are discusscd

onlv

brie ff1 .

Two con:mittees hare precede'd the presen(: l' -F / '6J ) and 5-l{- I' ('68 ). Comnrittee ii-F produccd the rt'porl "l'edestrian ( h:rractcristics aud Space Require' mcnts." uhich har nerer been puhlishcd. Honerer. nratcrial from

Vukan R. \'uchic (Il): and Frederick J. Wegmann (A ). l'hc prescnt report !+ati prepared with

rhose book. l'('dcstritn Planning-jlnd t)rilg!. has becn used exlensircly nith his nermission. .\ddirional comments and suggeslions frorn Scott Rutherford of \orthnestern Lnirersity' are gratefully ('ommittce 5-R. as organized in l\larch l9?1. includcd Donald S. Berrv ( I ): Rob-

ert l-. Elol ( Fl: ohn J. Fruin (lVl): Barnard ('. Johnson ( F); Littleton C. IlacDorman (ll ): lrancis P. t). Navin (Il): Stoen ('. l'roros( (A); James E. ,.1

\\'att J r. ( J: ): and Frederick J. \'\'egmann )' Edmund J. Cantitli ( F )

Chairman

is includcd in

the pretcnt, rcport. \lcmbcrs of Comnrittcc ll-lj includctJ [<rchard l. Strickl:tttd II"). ( hairnr:rni \orntln (. llrrrrttt ( F ): I lernran l](ttzow (i\l ): J ohn l'' ('ar allero Jr. (\l ): Ilonuld J. Fishcr (A): Barn:rrd ('. Johrrson ( F); F-ugcnc J. l-essicu ( I ); \\'allt.r S. ll:rinrille Jr.; and Darid \1'.

(ll).

Schoppcrt A report produccd b1 ('onrmittcc 5-R-T

'l'his com' has bccn crtcnsivcl) rtrised.

nrittcc inclurlcd Jonrcs E. \\'att Jr. (F).

(h:rirmnn: Robcrt 1.. Blol(t't: Ralph N. llrcsciit ( \l l: John P' Carallcro 'Jr. ( Il ): John.t. Fruin {ll): Jtck tr|. (irtcnspan (\l): Roe l'. llendrick (F): lhrnard C.

Johnson (I): liu[enc J. l.cssicu (l"): l)onald \\'. l-outzcnheiser (1").; l.ittleton C'. trlacl)orman ( \l )l \\ illirrm \larcorti ( l ); Robert ll. i\lurph.v (r\l ); !'rancis l'.

Wulking

is such u

fundatncntll meuns of

trlvcl thlt it is often trrken for granted rurrd ovcrlookcd Virtuull)' ull ntodcs ol truvcl rec;uire sonte pcdestri:rn trip link' ugcs. Circulrrtion uithin mujor uctivity cuntcrs. n]()dal trilnsl'ers ntlrdc through

tcnnirrul lircilitres und ucccss to tlrbrn

puhlic transportittion systcms are lll ex' lunrplcs ol pcdcstriarr uctivities that requirc the increuscd attention of trallic cnginccrs lnd trunsportation Plunners. This rr:port dcals only rvith the most blsic upproach to ovcrall pcdestriun sys' tcnr plunnirrg' thc collcction und consolidutron ol krrotr n cliaraclerislics con' ccrning both pcdcstrilns and their llcilitics: antl thr.lsc purarrretcrs which nrrght hc callcci.rert'ice r&luiretrrenls of

dcsign critcria. With consolidation of suclt nru!,cri:rl. thc dcvclupnlcnt oI u sys-

tcnrrtic nlcthodologv for

plunning. dcsrcn und control

pcdestrian posstblc;

upplicable throughout the United States. thcre nray bc rcgional differences which

"pcrsonrl spacc." Fruin's work, how-

cvcr, represents a brelkthrough, and the ITE must recopnize this work for its value und utilitv to the profession. 'l'he re lationships Utitity of thc Report. betrrecn the '{dequacy of walkways and the adequacy of roadu'a)'s are clear in

thc urban context. ln the citY.

in-

udcquate sidewalk spuce spills pedestri' uns into the road. and inadequate crossing time at crosswalks affects tramc

movcment as well irs creating safety problent

Plunners and engineers have not al\\'3vs seen themse lves in the role of designcrs oI sidcwalks. They should. how-

evcr. be ntore cognizant

of the basic

nrode ol. trunsport: walking. This is not onll' bccause u alking is busic. but also bccause it constuntly interacts with. and

nrany respects conllicts with, othcr

modcs oI trrnsportation.

Thc udcquucy ol walkrvays, hallways and stairways in und around terminals and trip gdnerutors allecls the entering and cxiting capacity of other traflic. ln'

Points to Be(・ ons:dcrcd in:

PLANNINC

sYSTEヽ 1lELEヽ lEヽ liS o

ohvsicil'

o inedical limitutions. handicups o user mix: scr. ilgc

Human

o physical linrits

. trip purPose

o pedestriun dcnsities o desire lines o impedanccs

surfacc grade

orotection from elements:

. comfort . convenlence

b) other pedestrians

TotalTravel

SYstem

densitics o applicabilitY of demandactuated sYstems o new technology

maintcnance and control for an interesting. Plclsing.

o interference

r terminll

design.

including platforms. r

holding areas. ticket and

Oeitn of. and transition .'tith

ramls and vertical lift nrcchanisms

. entrunce und cxit design o physical obstructions

* ith

vehicle

pedestrian transition on

continuous slst: ms

o cror* ding

bagguge qucues

o intcrlcrence with other oedestriun [lows . pedestriun'vehiclc scparution

o regulutions

o cost etlcctivcness

o efliciency o continuitY

r enforcement

o economic feasibilitY

o hurdn'ure

r flexibility

conncctivitY

eFl竃

Figure

control

cllicicnt sufc cnvi ron nrcnt

o transitions involving other modes

. oedestrian

Total Plan

r r lighiing

o need for continuous flow o pedestrian uork mcasures

dircction

r guidunce

ientncruture. hunridity control

o ethciency

o in motion

o olcasunt visual, :-rudir.l cnvirons

lnterface with a) other travel modes . at rest

IR()L

pcrcL'ptlon rcuctlon o comprehension o salbty

iover und shade

o esthetics

Environment

よ (・ 0｀

. .

o ph1 sicirl sclle and diltlcnstotts . humi.In bodl dinlcnstons o locomotion chlractcristics r ps1 chologicul Preferenccs

alignment width

o relationship lo total triP PuthwuY

()PERAT10｀

DESICヽ

認:1淵 ::Li輌ぃ

o fi nancial considerutlons

o esthetics o social considerutions

o zon:ng

Considerarions

lor

pedestrian

facilitl'

decelopnrenf'

charac-

queuing

n1:::1lil,"i::J::::

::T;.X'flili1;:'*1":o-.1uate spaces. wulkwuvs and escalirtors' He

p"lcstrian circurirtion rlii.ni exiting from aren;rs. srut1iums und uiririqr rt"ii. up u purt. and thcir imp,ct on uy, oorts. for insrancu. ur. l-oit,'"rrr.o lsuvcl it' u tbtul tran.sportrrion plan and a cuuss of, relutcd n,',uu.n,',.nt, plan). given "nj city u on pluce(or. indccd. s)-stcnl ocdcstriirns lcaving thc l.acilitics. f"i--^irpr-.-rhc luitur! ro att;act pir' ment and dcsign ol.p.,r*liiiun lu.iriti.. rrons to u public tr.nrpurtution ,yti.* phcwill have obvious cllccts on propcr u.. due,. in purr.-19 incrlicicnt or ing and design of trre vihicurar tiatric ",rf . uttt"

bc.consicJcred in the context of con' trolli.g nrechurtical m.dcs' Pcoplc cun be controlled through chunnclizution' propcr regulutory mechunisnrs,nd their tnlbttttint' proPcr inl'ornrutionrlsign-

deed, pedestrian capacities and must not only be gi,,-en to '1,ll1Ll:ici.": the consideration te ristics of movement may determine the s1 stcm but ulso to of elcments ti,. characteristics or.transport peuking. to un1 gi".n .o;pr.,. The peaking .t"ru.t.iiitili-ol trarnc their rerurionship ul *nicn

t,ie'hfy

'1'opcruiion ur rujor'crivir)' ..r,"ii is c'ontinviding the prtrfessionar thc sputiui ,,ti"n!.nt.n, .or' cdgc thur wiil hcrp in undcrsruncrrng s;n, ,n rhe eusc of moicnrcnt bcund ictivrtics unl n.,ou.r.nt pedcstrun the naturc of mr;dcs' uctivitl r , t**"n lcal.urcs in providing a<Icquiltc rJcsign pcrJcstrian tircilitics ol' o*rign iii. rvurkrvu:-s Dcsigning for th't nlovcnrcnt. thc inrcr' coniide'riion iiro rn.,si-tu[c for a proper lcvel of ,.'rrici is as imelcfund.mcrrtll thr.. r.i",l,...iii,ii of porrant as dcsigning roudways ro a the humlrn' thc slstcm: ih of ments propcr level ol' servicc. ih. ,1.' whilc this rcport rcrutcs onlv ro chur- ;;;;,rt. 'nd thc .nuir.r'iu,"n,. thc r.irlc. pr,iri."i .u'.,ri.t", rcqurru,'rcnrs. ;L;;;r 'lirrt scrvicc ucrcrisrics -controt pllthwu)"' thc r1e ol' charitctcr ur:'J surlucc 'nd must planning. dcsign tnd physicirl. ltnd topogritphic inrpcdanccs' givcn considcriltion.

exit

facilities:

prowirh buiic knowr-

This report rcprcscnts ir stcp in

"npttl-U i:l::l::f tcirrs. Thc succcssr'ur prarrning unrJ

:m::'-

:: h::

".1,1,.i.;'

s;: :fi :i : ;

"' cr:tl utrlity itnd convctttcnce ol' thc

;:;l

sys'

should arso cvuluate the lc'el of scrvice' muy necd to chcck the cost'elrcc' 1nd,hc oi' ulternative designs within 1]lt-lttt lin.ncial constraints' giwn

controlling pedcstriun

in-'. und

movcmenL may

c.ntror devices und pcdestriun

educ'tion progrltms' Ho''vevcr' pcdestri' urls ure considcrubrl' mtlrc llcxiblc und

unpredictuble in thcir movemcnts than automohilcs or othcr modcs' where dis' ohcl ing ccrtuin rulcs could inrpair thc sultty und wcll'ure ''rl'hunran lilc: i'c" the p.ttcrns ol'pcoplc in the wulk modc ure tnttch nlor€ irrcgulur' [:igurc I sunlmurizcs pcdcstrian facility dlvclopmcnt considcrations' dividcd

'l-uhlc

Suhdivision of Avrragc I'cdcstrir=ii-Sfecd. Standard l)eriation

Avcrage

pcr Fcet per i\leters per i\linule Scdond l\linule

:: ven i■ ・、

pcr i\Ietcrs per Sccond l\l inu.e

fect per l\linute

Feet

Fc-'t

８

０

415

８２

Wo men

２１０

4.0

７３

２

０

Cbmbined

２６０

4.3

７９

８

０

２７０

ilm()ng thc thrcc nlitj()r clcnv:nts of pe' dcstrian systcnrs. Elch ccll ol the frumc' wrlrk prcscnts uspccls and exunrplcs of thc pcdestri;rn.prohlcm that must be rec' ognizcd and considcrcd. It al.so illus'

trirtcs thc intcrrclirtionship of all pedes-

trian lacility dcvelopnrent

considcrations.

Charactcristics

\\'alking Specds. Walking sPeeds are allccted by a variety o[ factors. including:

Table 2. Speeds Crossing a Roadway. Number of ObserYations

Average Speed for All Observations Feet per iVtinute

Men Women Combined

Fcet per

Meters per

Second

Minute

r personal (hunrirn) characteristics o path$a)' characteristics o environntent 1 tnP purPose o trulllc density o personal comfort and safetY. Leuel Pathways. A wide range of wulking speeds muy be found among pe-

602

２３０

７０

２００

６１

:38

２２０

６７

740

destriuns. lvlany of thc dillerences in speed can bc ultributed to personal char' acteristics. such as age und sex. For ex' umplc. one survey (MacDorman) shows r,r'alking speeds

Table 3. Speeds Obscrved, Based on the Horizontal Distance'

Ascending and

Descending.

per Feet per l\linute Second

Feet

l\{cters

per

illinute

Feet per

Nlinute

per l\lcters Per Second Nlinute

Fcet

lnch Riser and

I | 7:-lnch

Tread

2.0

1.6

6-lnch Riser and | 2-lnch Tread

:44

ハＶ

2.4

′

idth Obser■ ations or ll Stal『 sin Newヽ ork City. Table 4.1｀ ′

\\'idth of Stair Center-to Center of Hand Rail

Nominal Number For Pcdestrians nith tew Hand Packages, This \\'as: of Lanes

Centimcters

nchcs

Inudequatt (Scldom l-u o Abreast undcr One-WaY Use)

122

114

49 to Sl

!241o:30

Sari.rfoctory

150

S a t i.s-[a

63 to 64

160‐

Sante <' t o r.r' ( Some Th ree- Lanc Use Whe n Pcdestrians Are Not

uns but

170

691o76

1751o193

TRAF「 IC ENGINEERING/MAY 1976

I nadequat e ( Except

When

Pedcstrians Are Not ExactlY Abrcust ) Adequate (NormullY OnlY Two Abreast. But Some'f hree'Lane Use ) Satis factorv

to 280 feet

per

ctn nornrallr be expected to

ubout 270 fpm (82 mpm) or 4.5 fps (1.4 mpm). Walking speed is increused or decreased b1' the length of the stride and by alterirtion of the ccnter of gravity, which occurs u'hen one takes a forward stance' like leaning into the rr'ind. A linear rela' tionship hus been found benr'een walking spced and puce length. This would indicute that space is un important ele' mcnt in humun locomotion. Pedestrians not onl)' necd suliicicnt spuce lor normal pacing but rlso to sense obstacles ahead, or to avoid conflicts with other pedestri' ans. ln dense crowds. normal human locomotion is signiticantly restricted. forc' ing peuple into an uncomfortable,

shullling gait. Thc usual travel sPeed for persons in

prssagc\r:lvs ut on lsvcl. unobstructed lrcc-llorr rng; srdcrr alks is conservativell estimuted ut 2'10 fpm (73 mpm)

.1.0 fps (1.2 nrpnr) for most design purposcs. This represents the average specd at a "comfortable" design vol'

Exuctl;" Abreast 163

150

on the mix und composition of pedestri-

Descending

Ascending

minute (46 to tl5 meters per minute) or 2.5 to 4.7 feet per second (0.76 to l'43 meters per second) for elderly women, and 250 to 185 fect per minute (76 to I l7 meters per minute) for Young men.t The averuce free'flow wulking specd for any purticular population depends

li漱 ‖ tよ C｀ PCゞ

1●

%:留

l(ハ

,密 l:I(1'I電

ぶヽR

d th● t. for longcr distunces, W● lki● 8 Speed

declines

ume. This uvcragc pcdcstrian specd,cirn bc subdivicled as shown in Tlrhlc l''

The distribution ol' wulking.

spccds

seems to',ltrllow.u' normal puttcrn: thus. the rtnge of spectJs (deiigncd ,b.t- = one standxrd deviution ubout thc average) includes about 6li perccnt ol'ull pcrsons. Womcn wulk more slowlY thun mcn

when crossing

a roadwuy. The datit

shown in Tablc 2 wus rccorded bY DiPietro for a 29-foot (lJ.lim)wide strcct. As expccted. pedcstrian rvalking speed

varies as a function of the tin:c bel'ore the arrival oI the next vchiclc. Rcscurch completed by the Roud Rcsearch Labo-

ratory (England) indicates that

the

nearer the vehicle. the greuter the walking speed. Grades. On PathwuYs with under t0 percent of grade, thcre appears t.o.be verv littie increuse or decreuse in walking spe;d due to sloPe. A controlled e.rperiment of soldiers walking on a variable' grade treadnrill (Wuyne) for the purpose

developing fatiguc lactors 5[er'red

that an increase in positive treadmill

grade. from 5 to l0 percent. decreased walking speed by only I 1.5 percent. and increasing the grade 1tt +20 percent (which iJ not too common) decreased speeds by 25 Percent. ' Slopej of l0 percent or more '*ill affect speed and volume. At 12 percent. speed drops from 260 to 2'10

"".rag. fpm 179 to 7J mpm). or 4.3 to 4.0 fps ( 1.3 to 1.2 mps). The maximum pedestrian volume gocs from about l0 Persons per foot width pcr minute (PFill) to 7'5

PFlvl (33 persons per mcter width per minute IPfvlNll to

]5 PNlM).

However. thesc dccreascs are less than what can be expected just on the basis o[ uge. sex or trallic densitY. 2. Stairs. Movemcnt on statrs ls more

structured und restricted than wulking because of the rcstraints imposed by the stair steps and the need to overcome

gravity in asccnding or to control it safely in descending.

Thc use of stairs in the circulation pattern must be bulanced agirinst the reul or potential needs of the hundicapped' the necd to minimize travel distances and the geometric limitutions of thc sitc' Eiergl'. Thc totill cncrgy expcndcd in ascending sttirs is l0 to l5 times Sreat€r

than thui uscd in wulking a horizontal distance equal to the height oI the statrs' Unlcss spctilicd, 'llow.

ligures given arc for ons'wuy

The encrgy crPendcd in dcscending stairs is only onc-third grcltcr than thut uscd in.walking u h,;rizontul disiunsc equd to the hcight oI thc stltirs.

Speed.l'.rble

served (buscd

I shorvs somc specds obon the horizontul dis-

tancc) in uscending und descending. Ri' scr hcights havc a significunt cllcct on spced. Slightly lowcr riscr hcights tend to producc fastcr pcdestrian spceds. but cxtremc variations rcduce cllicicncy' Thc approxinlutc average spccd to (rilversc a horizontul distunce is 100 l'pm (J0m) or 1.7 fps (0.5 mps) when climbing, and 120 l'mp (37 mPm) or 2.0 fPs (0.6 mps) when descending'

Length. There is no available informa' tion on the ell'ect of thc lcngth oI stairs

both ascending und dcscending. remain relatively normul, down to un uvcrage

pcdcstriun arca occupancy of l0 square fcct (0.9nr'z) and thcn are reduced due to

trallic density -fhe

volume and density curves in Figurcs 2 and J illustrate thc relationship ol volume and average pedestrian arca occupuncy. The up direction is generally uscd for dcsign purposcs because of its

lowcr capacity vulue. Maxintum flow volunres-15 Pcrsons pcr minute per loot of stair width (50

pcrsons per minute per mete r) ascending nnd 20 (66 per meter) descending-werc developed.

The critical areas derived, of 2.9 and 3.2 square fiet (about 0.1m':) cerng vert

to a t'r'o

stair'treud. shoulder'

on the rate of flow, but extremely long

Width. Observations of I I stairs in New York City are shown in Table 4' Voluntes. Voluntes of 30 to 40 persons per 22-inch (56 centimeter) lane per min'

trian zone is about fouf to five treads

stairs evidently slow traffic further.

ute have been repor'.eC. Descending

ratcs are slighly greater than ascending rates.

Observations from a study of a 5-foot wide (1.5m) stuir under 1's1ggd flsw (after o train had discharged) are shown in Table 5. This studY was made at the Union Turnpike station of the IND line of the Nerv York City subway system' Counts were mude only while the stairway wus in use. 3. Speed-Density Rclationships. Figure 2 compares levcl walks and stairs. in relating thc size ol the pedestrian "mod-

ule" (area kept clear of intrusion

most walkers) to wulking spced. Figure J relutes the sume module to volunres. ln both instances, it is appLrrent that stairs have a lesser capacity than equivalent level wulkways (upstairs movement ls

taken us the "worst case"). Leuel ll/alks. Maximum florv volumes of 26 pedestrians per minute per ioot (85 Der meter) of walkrvay havc been meas' ured (Figure 2). The reciprocal of pcdestrian density (square fcet per pcrson) ls used for convenicnce in visualizing relu' tive levels of pedcstrian lrcedom' This curve conlirms that normal human walking specds rcquire signilicunt anlounts oi pcdcstrian area. and th11 slowding causes restricted human locomotion.

This capucity value occurs ncar the criticul region of pedestrian arca occu' pancy. which is shown to bc approxi-

width area. The zero-movem€nt area, of 1.5 and 1.6 feet (about 0. l5mr) is equiva' l€nt to occuPancY o[ one tread.

At l0 squarc

feet (0.9m2), the pedes-

long and 2t/: feel (0.8m) u ide. This gives

sufficient room for reasonably normal stair location. but not enough area to by-pass slower walkers. Using the two shoulder width-spacing criterion for by' passing. lateral spacing would have to expand to 4 or more feet. giving a re' quired area for by-passing of about 20

square feet ( l.llmt) Per Pcrson. 4. Sidewalks. Volunte. A *ide range

of volumes has been reported during counts ranging in length from five minutes to l2 hours. ln most cases. only the total sidewalk width, not the effective

width. was reported. Tablc 5"shows the variations summarized by pedestriuns per foot width Per hour (PFH). The

maximum flow reported was

San

Francisco, with a pre-Christmas count oi 13,138 persons per hour on a 22'foot (6.7m) wide sidewalk, having an effcc' tive \,lidth ol'16.5 fect (5.0m). Efeaiue ,Yidths' RePorted widths uary from l0 to 25 feet (3'0 to 7.6m) lor

totil width of sideu'Lrlk.

The ellective width. which is more meaningful. was reported by only two out of l0 cities in thc survey and are shown in Tablc 7.

The common obstructions rePorted

arc polcs. signs, tlower stands and mail boxes. Supplcmentary studies of volumc

vcrsus density for various sidewtlk widths did not show signficant varia' tions: an 8-foot (2.4m) ellective width hud about the sume rate pcr foot as a

l6'

nrateiy 5 squure feet per person (0.5m'z)' Usc of such a value for design would gire a vcry poor standard of pcdcstrian

foot (4.9m) ellective width.

.Sloirs. Bcr,ause o[ thc limitations imposed by stair tread and riser rcstrictions und considerutions oI pcrsonal safcty. avcrirgc pcrlcstriun urea hls a lcss signil'rc:rnt ellect ott pcdcstriurt spccds on stairs

process is illustrated by the kind of mass

traffic flo*.

5. Arrival Processes and Queuing. Fruin has identified two arrival proc' esscs: bul/< l;,nd intermittent. Thc bulk

than during lcvel walking' Stuir spceds,

I976

Table 5. Observrtion of a S-Foot Stair under Forced Flow.' 'l'imc lnterval ( illinuies )

per jllinute ７０

15 L5

122 102 106

0.8

４６

l.75

Fioot

Pcrsons per

Persons

Volume

Persons per

l\lcter

r*'idth per l\linure \\'idrh pcr l\linure 46

14 i3.6 14:2 9.2 14.5

６８７１

45 47 30

７３ 09 67 tTrirflic in onc dircction, moving up thc stairs at a rate ol ll0

steps Per ntinute.

Table 6. Pedestrian Volumes EfFective VVidth

per Pedestrians per Hour i\leter Pcr Hour

per Pedestrians per Hour i\lcter per Hour

Pedestrians

Foot per

Chicago. CTA

Chicago, CTA Chicago, CTA Seattlc St. Paul

950

289 299 309 396

1,0:0 1,300

137

450 1,000

1,991

Citv

576

1.890

650 608

2,130 1.990

Seattle San Francisco San Francisco San Francisco

Meters

Feet

Nleters

Feet

6.7

:6.5

4.6 4.6

9.0

3.4 5.0 2.7

7.0

2.1

ヽ idth of()bstruction

Efrective vvidth

Feet

[\l eters l.5

5.0 5.5

1.7

60 80

A verage

Table 8.Summary or Levels oF Service Standards forヽ ″aikways. Level of Service

Normal Arerage Tralfic Volume per l\linute 14'alking Reverse Per Foot Per Meter Speed Flow

Average Areg Module per Person Sq. Ft。

3.2

(Or greatcr)

Ｂ

2.8

Ｃ

1.9

Ｄ

Ｅ

0.9

Ｆ

0.5

l.4

(Or leSS) Ｆ

723 (or less) 826

Flow

39 56 17 72 22 Variable up to 12

Ｆ

Ｒ

Ｆ

Ｒ

Ｓ

Rclatively free. minimum of restrictions or inconvenience. Restricted. higher probabilities of conflict and inconvenience'

¨ ・一一Ｓ〓

Severely restricted.

TRAF「 iC ENG:NEERING/MAY 1976

Ｒ

Cross

∼

first-come, first-served priority syst€m

(an example is u tickct line): or as bulk. rvhich is unordered and lacking in any

considered in two \A a) s: those which contain persons *ho stand and wait,' with limited movement rvithin thc qucuing areu (such as at the foot of a motor stair): or those with people who combine waiting with some reasonably free internal circulation through the queuing urca (such as on a sub'*ay platform). No standards are uvailablc for thc design of queuing spcces. but standards can be suggested that are based on hu'

Table 7.EIFective Side■ alk Widths. Total Vゾ idth

Queues can develop in t'*o ways: as lineal or ordered, with the conventional

discipline. Bulk queues can be further

980

304 606

San Francisco San Francisco

minals and oflice buildings. s'here therc are: multiple sources of demand with short-term surges called "micropcaks." These are illustrated in Figure 4. As Fruin notes, "the distinctio' in arrivul proccsscs can be important to the dcsigncr, since the assumption of an 'average' design period. which is not truly r€prescntative of the uctual pattern, may produce an inadequate and inconvenient

faciIity."

Total Width City

exodus that occurs after a sporting cvent. or at a railroud platform when a loldcd passcngcr train discharges. The intcrmittcnt proccss is the common arrival process sc€n ut trtnsPortatlon ter'

R S S S

man bodl' dimensions and

Personal

space prefcrences.

6. Signalized Crossings. At signalized intersections. pedestrian crossing opportunities ure creuted b1 the signal phusing. A high percentage ofpedcstrian violations ma-t' be an indication that the signal does not adequatell reflect pedes' triun necds. (lt mal srmpll' reflect min' imal enforcentent. or local custom, but this must be the judgment of thc engi' ncer/plunncr.) Cood signal timing includes a minir,rum green time, allowing rredestnans to cross thc roadway. I n

nlost cases. the pedestrians usurp prior' itv ovcr right-turning vehicles because the lutter do not have adequate tlme to start the turn before pcdestrians havq startcd to w'ulk with a gre"n signal.

Nlalcr hls studied "de-indiriduu-i tion" ol' pedestrians observing a traffic signul. ll is h1 pothesis is that a pedestrian ucting individualll ma)' conscien-

tiousll observe u DON'T WALK signal indicutitln. bu1 s'hcn thc same pedestrian is in u group observing the samc signirl. und another pcdestrian begins to

cross. the entire group

uill follow. Based

upon u studl' of one sile, lr'layer con'

cludcd Ihut the dc-individuated bchavior

is at a rutc of 9.ll pcrccnt. Individual violalors ilcc()untcd lirr 60 pcrccnt ttf thc violations. Nlcn hrrd il grcutcr pftrpcnsit-v towrrd. bcing.viohtors thun wonren. In it scpafate study. Ntunnrng has concludcd thar childrcn sho*: l highcr proprrrtio4 of observ;lnce thun udults. Tlie rcsults of such studics must bc ircceptcd \,r'lth cirre. ho,*'ever. sincc pcdcstrian disciplinc * ill vary from rcgion to rcgion in thc U.S..

just as it most certainly vuries from country to country.

l\lechanical Stairs. Walks

and

Ramps. Moving stuirs. rvulks und rantps

offer convenicncc to pcdcstrians and tend to movc thcm in an.orderly manner. .Except for variablc-specd moving wulks. they are not currently designed to save time beclusc they tl picully move .tt rates rvhich urc slorver than norntul

walking. Their initial artd maintenunce costs limit their use to special design problems. or to high-service and highconvenience installations. By dcfinition.

these conveyances ure power-driven and arâ&#x201A;Ź arranged like endless belts. with tr+.o moving handrails within closed bulustrades. ln[ornration has not been found to develop pructical cupacities of moving walks as conrpared to rated capacities. but queuing characteristics in advance of

the facility will usually determine vol-

A stationilry walk is usually provided adjaccnt to a moving walk: short ovcrloads will naturully divert to the stationary'walk. Pedestrians entering and exiting movumes.

ing ramps mul crpcriencc some discomfort and dilliculty bcc:rusc of thc abrupt change in planc as thcy stcp [rclnr a lcvcl

surface to un inclined moving surface.

or vice versu. M:rnufacturers are developing moving ramps thut will start out hor-

izontally before climbing and level ollto a horizontal pcisition. ullowing pedestrians to step off.

Rated capacities o[ motor stairs assume that each step is occupied. Peak

five-minute actuul flows were only 63 percent of ratcd cupacity lor stair pperation at 90 feet (27.{m) per ntinute. and only 53 pcrccnt lor stairs operating at 120 fcct (J6.6m) pcr nrinutc. z\s a corollary. it was found that increasing the

of the motor stair from 90 l-eet (27.4m) pcr minutc to 120 leet (J6.6m) specd

per minute (u 33 pcrccnt increase) raised the cupacity of thc stair by only l2 percent (see Scrvice Requiremcnts, Secticn

4, Peaking).

Service Requirements

l. l,evcls of .Scrvicc. [jruin hls dcvcloped lcvcls ol'scrvicc l'or pcdcstriun llow and qucuirrg. sinrilar to those described in lhc Highray Capocittr il{anual lirr vc-

hicular trallic. Thcse estuhlish standards that allow u prcdctcrminution of trilllic chilri.rcteristics resulting l'rom ditlerent allocutions of pedcstriun space. In areus rvhcre somc lrccdom ol' design is allowed, such us shopping malls. high lcvels ol' scrvice can be providcd, resulting

zero.

in improverncnt in thi pedcstrian environmcnt. In restricted areas with cx-

treme linritations. such as comnluter ter-

minals. problcms can be anticipated and perhups ollset by ultelnative designs or

operating procedures. lValkv'a!'s. A levcl o[ servicc standard

lbr

rvalkways and ramps provides a of determining thc quulitative aspccts of respective dcsigns. However. it does not elilninate the need for designer judgment. Dcsigners should care fully examine ull aspccts of prospective walkways, including such trallic charactermeuns

istics as projccted magnitude and duration of pe:rks. us well as all the rumifications of space usc and cost. When

designing lor peak demands ol'short duration. lower levels of servicc standards may be tolcrated to provide the basis for more economical dcsign. Added consideration must be given to the selection o[ design standards near maximum capacity levels. since the critical pedestriun

densitl,

is likely to be exceeded inter-

mittently. Whcn critical density- is

excceded, llow volumcs l'ull below the spcc-

ilied design levcl und pedestrian delay and buckups ure likely to occur. requiring determination of the adequacy of holding or queuing space at the approuches to the critical section. These levels of service standards are based on a range of area occupancies per person. Design volumes are presented us

a ran8e ol pedestrians pcr foot of walkway width per minute (PFIU). lf unidirectional traflic is comprised ol'commuters or workers, then the higher design volumes in a given range may be safely assumed. The lorver rangc ol'design volumes would be recommended if truflic is comprised lurgcly ol shoppers or pcrsons carrying bagguge. or il' the trallic pattern involves cross movements, reverse flows and othcr conflicts. The following conclusions have been

Normal mean walking speed may

uttuiniid at an :rpproximatc pcdcstrian urca occupuncy ol'25 squarc fcet (2.lmr) psr Pcrson. The prohabilitv. o[ crossing conflicts,, rqmuins at 100 pcrccnt virtuully up to un uverxge areu occupancy' ol ubout l5 squure l-eet ( l..1mr) per person. It remains at ubout the .iC-percent lcvel bet'*'cen l0 to 35 square l'ect ( 1.9 to J.2mr) per person. beyond which it drops to

Level ol service stundards for walkways are described below und summarized in Table 8. Pedestrian volume and area relationships are shown in Figure 5. Leuel of Seruice A. Averuge pedestrian area occupancy: 35 squure feet (3.2m'?) per person or g

rea tc r.

Average flow volume:

PMM) or

7 PFM

(23

less.

Sullicient urea is provided for pedes: trians to freely selcct their own walking speed. to by-pass slower pedestrians and

to avoid crossing conflicts rvith others.

Application: Public ouildings or

plazas without severe peaking charaqteristics or space restrictions.

Leuel of Seruice B.

Average pcdestrian area occupancy: 25 to 15 square feet (2.3 to 3.2m2) per person.

Average flow volume: 7 to l0 PFM (21 to ll PMM). Suliicient space is available to select normal walking speed and to by-pass other pcdestrians in prinrarily one-directional llows. Whcre rcvcrse direetion or

pcdcstrian crossing movcments e.\ist. minor conllict will occur. slightly lowering mean pedestrian speeds and potential volumes.

Applicution: Reasonably high-type design for transportation tcrminals and buildings in which recurrent, but not severe. peaks are likely. Leoel of Seruice C. Average pedestrian area occupancy: I5 to 25 square Ieet ( 1.4 to 2.Jm"l per person.

Avcrage flow volume: l0

(ll

to {9 PMM).

15 PFM

Freedom to select individual walking

speed and freely pass othcr pedestrians is restricted. Where pedestrian cross move-

ments ond reverse flows exist. thcre is a high probability of conflict requiring fre-

reached:

o The ;ritical pedestriun area occu'

pancy in peoestrian trallic flow is about 5 sluare fcet (0.5mr) per person: below this area oc:upuncy. wulking speeds fall within the shr llling rangc.

TRAFFIC ENGINEERING/MAY 1976

qucnt udjustmcnt of spced and direction

to uvoid contilct. llclsonubly fluid flow. ０ ¨ ，Ｃ〓 ≧ Ｏ α ∽ ０一Ｏ Σ Ｏ ● ● ０ ∽ 一〇Ｃ一〓一＞＞

80 60

ｏ，一Ｃ〓２ｏＯ ● ０

０ ● ００め一〇Ｃ一メ一０＞ ≫

bul consiclcrrblc l'riction und interaction anrong pcdcstrians. : Applicution: Heavily used transporta-

tion terminuls; public' buildings;

open spuces whcre severc peakinS plus space

250

rcstrictions limit dcsign flexibility. Leuel oJ Seruide D.

200

Avcrage pedestrian area occupancy:

l0 to l5 squure feet (0.9 to l.4mt)

150

person.

Average flow volumc: (49 to 66 PM M ).

100 50 0ど

4.0

4.5

(M) Module in Square F6et per Pedeslrian

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Ｏ一っＣ〓２Ｏ α ＯＣｏ ∽ ０

０００

cross-flow movements. Volumc

，

Ｃ〓２ＯＬのＣＯ φ ０

卜

by-pass

slorver-moving pedestrians. Extreme difficulty is experienced in reversing flgw or

・０２

for the most

ling. Area is insutficient to

００Ｌ２０

４０

Application: Only

Average pedestriar, occupancy; 5 to l0

〓０ ■ ５

Ｏ α

con flicts.

square feet (0.5 to 0.9m'?) per person. Average flow volume: 20 to 25 PFM (66 to 82 PM M ). Virtually all pedestrians have normal walking speeds restricted. requiring frequent change of gait. At lower end of range. forwurd progress is only by shuff-

〓

Majority have normal walking speeds restricted and have difliculty in by-passing slower pedestrians and avoiding conflicts. Pcdestrians in rcverse flow and crossing are severely restrict€d. Multiple

rmportant movement. Leuel of Seruice E.

Figure 2. Speed uersus module.

０ ¨ ● ヽこ

l5 to 20 PFM

crowded public areas wherc forward progress for individuals is not the most

(M) Module in Square Meters per Ped€strian

０ ■ ５

pcr

proaches maximum capacity

\\.ay. Frequent stoppages and

１０

aP-

walkinter-

ruptions.

Application: Only for short peaks in most crowded area. Occurs naturally

(Up)Stairs

Porl Authority Bus T€rmina,

％

mY 〔

0.5

1.0

der S'uγ e Rl pα 1.5

2.0

2.5

35 :

3.0

3.5

(M)MOdulo in Square Meters per pedestrian

Figure 3. Persons uersus ntodule.

nl 賞ほ

4.0

ll 4.5

with bulk arrival traftic pattern. Recom. mended onlf ior sport stadium design and rail transrt facilities. Adequacy of pcdestrian holding areas must be considered.

Leuel of Seruice F.

Average peak area occupancy:

square leet (0.5m'?) per person or less. Averugc flq*' volume: variable up to 2-i PF Nl (t{2 Pr\'l lvl ). All pedestrian *'alking speeds are rcstricted. Forward progress is only by

shullling. Unavoidahle contact with

othcrs ilrc frequent. Reverse or crossing movements are impossible. Traffic flow is sporudic. movenrent is based on those in iront. Represents loss of control, complete breakdown in tralfic flow.

Application: Not recommended.

In designing stuiru ays. increuscd consideration should be given to Stair.r.

the role of human factors becausc of thc

TRAF「 IC ENG:NEER:NG/MAY 1976

greatcr su[ety hirzards and energ]

ex-

penditure rcquircd in stuir locomotion. ln adclition. to thc crcrcise of dcsigncr 'j0dgnrcnt. in' evuluuting' tr;.rtlic patt;rns' ;and pcak ch'uructeristics rccommcnded' in using wulksuls stundards. thc l'ollow-

1200 ００こ Φ

ing Lrctors should be considcrcd in dc-

sign:

８。

Stairs should be rvcll lighted. provided with sullicicnt heud rr.rom. und

０一っＣ〓２

with properly designed rnd maintaincd

ヽ ∽ ＣＯ゛

ings.

Stairs should be locutcd so as to

，

Stairwa!'s should

be otfset

６。。

４０。

ヽのＣＯ ∽

２。

２０。

０

8 l0Noon 2

０

8lONoon24

Hour ol the Day

Nloderatc riser heights reduce human energy expcnditures and increusc tratlic efficiency. Six inches (15 cm) may be a desirable standard.

８。０

Ｏ一

０，

４。

ＯＬ

readily visiblc und identifiable as I means of dircct access to the levcls they are designed to interconnect.

０。。

Ｃ〓２

６。

０

riser and treud contigurations and ruil-

００ ■ っＬ

０。

Hour ot the Oay

Fipre 4. Pedestian door counts (source: Lower Manhattan Plan, Figure 2'34).

l'rom

mainstrearn corridor tratiic to avoid pedestrian conflict. o Clear areas. sutllciently large to allow for queuing pedestrians. should be provided at the top and bottom of all stair-

t.!

wayS.

When u stairway is placed directly

within a corridor. the lowcr c:rpacity of the stairway is the controlling factor in the dcsign of thc section. Level of 5slvice standards for stairways aie described below and summilrized in Tablc 9. Pedestrian volume and area relationships ure shown in Figurc 6. Leuel of Sercice A.

Average pcdestrian urea occupancy:

20 square leet (1.9m!) per person or greater.

Average flow volume:

PMM) or

5 PFM ( l6

less.

Sulficient arca provided to freely select locomotion speed and to by-pass others. No serious difliculties with re. verse traffic flow. Application: Public buildings. plazas with no severe trallic pcaks or space lim.

，

〓０ ■

８０

、ｏ一 ● ヽ６０こＯＬＯ一ョ４０Ｃｔヱ

Commuter Uni-Oirectional

゛

｀卜さ

ＯＣ２０ｎＣ００００

Shoppers

Multi‐

01rectlonal

(M)Modul● in Square Feet per Pedestrian

1.0

1.5

2.0

2.5

3.0

3.5

4.0

(M) Module in Square Meters per Pedestrian

itations. Leoel of Seruice B.

Average pcdestrian arca occupancy:

l5 to

20 square feet (1.4

to l.9m')

per

p€rson.

Average flow volume:5 to 7 PFM (16

2l Pfvll\l). This level of servicc rcpresents a space , about 5 trelds long and -1 to 4 I'cet (0.9 to I .2m ) rvide. Almost all persons can freely select locomotion speed. In lower range of orea occupuncy sonrc difli' to

culties are experienced in passing slower'

Figuoe 5. Leuel of seroice standards

for

walkways uolunre uersus module.

4.5

Table 9. Summary

of l.cvels of Serrice

A verage Area ヽlodu:e per L`vel o「

sellit` Ａ

sl.IFto M8

(Or greater)

:.6

Ｃ

:.1

Ｆ

1.9

Ｅ

Average Traffic

Volume per i\linute Per F'oot Per l\f eter

Perso■

ＢＤ

6 8

8 0.7 5 0.5 4 04

Ｆ

ｓ

Normal Stair Locomotion

Roverse

Speed

Fiow

16 (Or iCSS)

pcak dcmands und no serious spacc limiLeuel

20 26

Ｆ

l0 to

Person.

Ｆ

Ｒ

Ｓ

1ntermittent up to

Ｓ

)evercty restncted.

of Sert,ice C.

Averuge pedestrian area occuPancy: 15 sguare fcet (0.9 to l..lm?) per

Ｆ

Application: Transportation terminals. public buildings wirh rccurrent

tations.

Relatively lree, minimum of restrictions or inconvenience. Restricted, highcr probabiliries of conflict and inconveniencc.

一一一一〓

movlng persons. Revcrsc llow5 1r6ul6

causc nlinor trullic conflicts.

(Or leSS)

Ｒ

Standards for Stairwavs.

Average flow volume: 7 to t0 pFM (23 ro 33 PMM). This level of scrvice represents a space

about 4 to 5 rreads long and J fect (0.9m) wide. Locomorion speeds are slightlv resrricted. Slower-moving pcrsons cannot be passed. Minor reversc tra flic flows cncounter diffi cultics.

Application: Transportution terminals, public buildings wirh recurrent

peak demands and some spacc limita-

tions.

Leuel of Seruice D. Average pedestrian area occupancy: 7 to l0 square feet (0.7 to 0.9m2) pcr per-

，

〓ＯＳ

son, F_￨「

Average flow volume: (33 to 4l PM M ).

_lpl C I B

l0 to l3 pFM

This represents a space about 3 to 4

０一０ヽこ

treads long and 2 ro 3 feet (0.6 to 0.9m) wide. Locomotion speeds are restricted

Ｏ α

[or the majoritl, due ro limired

Ｏ一，Ｃ〓２

open

tread spuce and inubiliry ro b1-pass. Re_ verse flow encounters significant difficulty and trallic conflict.

Ｏ α

Application: More crowded public buildings und transportarion terminals subjectcd to relatively severe peak de-

●

∽ Ｃ００ｏ

(M) Module in Square Fe6t per pedestrian

0.5 1.0 1.52.02.5 (M) Module in Square M€ters per pedestrian

Figure 6. Leuel of seruice standards

for

stairw,ays uolume uersus nrodule.

mands. Lecel of Sercice E.

Avcrage pedestrian occupancy: 4 to 7 square leet (0.4 to 0.7mr) per pcrson. Averuge flow. volume: ll to 17 pFM (43 to 56 PMM).

This repres"rts a

space about

2 to

treads long and 2 llct (0.6m) wide. rhe minimum possible urea for locomotion on stairs, Almclst all have normal speed

reduced because of minimum tread length space and inability ro by-pasq others. lntcrmittent stopprrses occur as

critical pedestrian densitl. is exceeded. Rcvcrsc lloss huve scrious conllicts. Application: Occurs narurally with

bulk arrival traHic pattern that immediutcly excecds available capacity. This is the onlv situurion lor uhich it is recomrnendcd: sports stadiums und transit facilities with large. uncontrolled. shortterm e,rodus of people.

Leuel ol Seruice F. Averuge pedcstrian area occupancy: 4 squarc fect (0..1m ) pcr pcr person or less.

TRAF「 :C ENG:NEERING/MAY 1976

Avertgc florv volume: Variable to l7

PM (56

PiU

rll

This reprcsents u

sfrace

about I to

trelds long und I lcct (0.6m ) *'idc. Complete brcukdewn occucs' in tnrtlic llow: many stoppuges: forwurd progcess de-

in l'ront. Applicution: Not recommendcd.

pends on thosc Queues.

Leuel of Sertice Zone).

Average pedestrian area occupancy: l3 square leet (l.2nr':) per person, or Averagc interperson spacing: (1.2m) or more. Spacc is provided

ior standing

and

area without disturbing others. Applicution: Pussenger concourse areus: baggxge claim ilrcas. Leuel of Sentice I ( Resticted Circulation Zone). Average pedestrian area occupancy: l0 to l3 square tcet (0.9 to 1.2 m'?) per

perion. Average interperson spacing: 3rt to 4

(l.l to l.2m).

without disturbing others. Application: Railroad platforms: pasSenger concourse arcas.

Leoel of Service C (Personat.ContJbrt Zonet.,

Average pedestrian area occupancy:

to l0 square feet (0.7 to 0.9m'z) pcr

per-

son.

Average interperson spacing: J to 37: fcct (0.9 ro l.lm ). Space is provided for standing and restrictcd circulation through the queuing

area by disturbing othcrs. Within the range of personal com[ort.

Application: Ordered-queue ticket-

sclling areas: elevator lobbies. Leuel ol Seruice D ( No-Touch Zone). Average pedestrian urea occupancy: J to 7 squure feet (0.3 to 0.7m'!) pcr pcrson.

Average interperson spacing: 2

feet (0.6 to 0.9m ). Space is provided for standing without personul contuct with others. but cir-

culation through the qucuing arcu

rs se-

verely restricted und l'orrvurd movemcnt is only possible as a group. M

lcss.'

fcet

Sptcc is provided {br stunding. but

uble. Circulutio:: * ithin the que uing ureu is not possiblc. Spuce cen only bc

sustaincd for short pcriods ol'time without physical and psychological discom-

lort. elevutors. Leoel of Seruice F ( the Body Ellipse). Average pedcs(rian area occupancy: 2 square fect (0.2m:) per person or less. Average interperson spacing: Closc

contact r,r'ith surrounding persons. Space is appro.rimutcly equivulent to the areu of the human body. Standing is possible, but close unavoidable contuct with surrounding persons cuuses physical and psychologicul discomfort. No movement is possiblc. ln hrgc crowds, the potentiul for punic exrsts.

Applicution: Not recommcnded.

Space is provided for standing and rcstricted circulution through the queue

Application:

. r\vssxgl; inturpcrson spacing:

(0.6m ).or

Applicution: Recommcnded only for feet

free circulation through the queuing

fect

son.

pcrsonul contuct with othcrs is unuvoid-

A (Free Circulatiotr

more.

Lerels of Seruite E (Touch Zone). Avcrugc pcdcstrian lreu occupancy: 2 to J square lcct (().1 to 0.Jmr) pcr pâ&#x201A;Źr-

otor-stuir

que uing

areas: pedestriun salety islunds: holding ureas ot crossr*alks. Not recommendcd

lbr long-term pcriods ol'waiting.

2. Design of Sidewalks,

Passageways

and Ramps. These facilities should be availuble to all persons. The design should not be such as to preclude use by the handicupped. Therefore. where possible. sidewalks should be constructed to allow the easy passuge of wheelchairs. Such provision will olten be a requirement ol'law. rcgulation and building codes. The height of the sidcwalk should provide sutiicient definition between the

strcct and 11'ts lsndrva| to insure the safet-v of pcdcstrians. The minimum width o[ any lacility

(where not excludcd by larv) should probubly be the width ol' u w heelchair plus a person. with allowances made for cleurance. This design also ullows for the passing of thc slower pedcstrians by the faster. und gives a total walk width of 6 l'eet ( l.8m ): 2.5 t'eet (0.8m ) for u person and 2.5 fect (0.8m) for the wheelchair, plus I foot (0.1m ) for cleurunce for the wheelchair. Thc old standard was a 22inch (56cm) lune. but this was found to be too nilrrow lbr l'ree-fiow pedestrian

use. Some authors suggest 24 inches (6lcm) but others. using lield obscrva-

tions. huve found l0 inchcs (76cnr) to be a morc uccurate description ol'what humans requrre.

ll' the facility is to be enclosed, then the sumc minimum width should apply. Larger fucilities should bc dcsigned so that each anticipatcd liine ol'pedeitrians is 2.5 feet (0.8m) in width.

Thc National Board of Underwriters recomnrcnds a slope of l:10 for rumps

within buildings. The mujor dcp:lrture of rumps lrom thc othcr I'ucilitics in this subgroup is thut hund ruilings should be

provided.at slopes. grcttef thun l:12. Slopes less than l:15 and l:10 appear to have very little inlluence on speed and volumc, rcspcctivcl!'. 3. Design of Stairs. Stuirs require considerubly lcss space than ramps. but they impcde the flow ol people. Stairs may reduce the llow through u passugeway by us much as 50 percent. Thus, a sidcwulk

with one person pcr l0 squarc

lbet

(0.9mr) has a conservative capacit.'" of 20

PFM (66 PM M ).r The "maximum l'orced" flow for stairs is 20 PFM (66 PM lvl ). at conccntrations of one person pcr 3 squurc leet (0.Jmr) of stairs, To maintain a comparable density of persons on the stairs, the capacity of the side walk would have to be approximately l0 PFM (ll PMM). lt is apparcnt. thercfore. that stairs have a smaller capucity than comparable walkways. This means that the stli; width must be widened i[ volume is critical. A good design value for the average stair is l2 PFM (40 PMM) going up. Minor flows opposite the main trattic should be given

u separate tratlic (76cm ).

lane

of 30 inches

The minimum width of stairs for the most part is regulated by codes. but rvhere it is not. the minimum should not bc lcss than ubout 5 fcet ( l.5m ). Other-

wise. stairways should be designated

with 2 7:-foot-width (0.8m)

modules.

und uscd as the control point in a corri-

dor-stair linkage.

Vulues of l9 and 2l PFM (62 and 69 PfUM ) havc been recommended by

Hankin and Wright as design criteria for

the London subwuys. The New York City Transit Authority uses a design capacity of 1.000 PFH (3.281 PMH) for stairs. or 16.7 PFM (55 PMM). and the M assachusetts Bay Transit Authority uses a design value of 20 PFM (66 PNIM

All of these values, with the e.rception of the New York Transit Authority's, occur at the critical rcgion oI stair loco-

motion flow. Pedestriun area

occupancics are about 3 square feet (0.1m) at this volume lcvel, at the borderline of the "touch zone" established in queuing studies. Use of design values for pedes-

triun trutlic flow at this levcl of pedestrian area occupuncy shows little rcgard

lbr the human requircmcnts of stuir locomotion. which ure considered to bc a minimum of 3 stair treads in length and

The design of ranrps must consider the

influencc

of slope on

pcdcstriun vol-

umes. as well us clinrutic and gcornetric construints. Sltlpcs ol'l:6 und l:E uppeur to bc the most frcqucnt uscd outdoors.

This is thc

fow at

which queues will

a, ,he bottom o[ stain.

stat

TRAFFiC ENG:N EER:NG/MAY!976

to lorm

Table 10. Maximum Theoreticar and Nominar rlrororsrair capacities. Nominal Size

Width at Tread

tl:::llヽ li:, 32 1.25 Persons,/

Inches

陥 1,r

0.6

l― _FPM

卜lPM

Per l lilll,「

i\l inutc

０

６８

０

８９

Step ９０

1.0

2 Persons/

27 37

２０

Ste p

Source: Strakosch,

C. R., Vertical Transportation, Elevators

配111:il11:シ

6,700

3,750 5,025

8.000 10,700

6,000 8.025

5、

tX,0

６８８９

and Escalarors, New York City: John Wilcy and Sons. 1967.

Turnstile Capacity. and the elliciency of the inrernal circulation s1'stem. Examples of thc peaking

Capacity

Turnstile

Persons per lllinute

rund direction of flows for rwo majoi building in New York City are shown in

Registering Free Admission A Ticket Collector Cashier Operated Coin-Operated Low Single Coin Slot M ultiple Fare Coin-Operated 7 Feet High Nonregiste ring Low Traffic Controllcr

Figure 4.

40_60 25-35

5. ,\lotor Srairs and l\loving Walks. Nlotor stuirs manufacturcrs rate thc theoretical capacity ol their units on the

:2-:8

busis of speed. assumed occupancy per

25-50

stcp und 100 percent step urilization. Since the lattcr is never obtained. even undcr the heaviest traflic prcssure with use by pcdestriun commuters, nominal

:5-25 10-15

7-Foot High Traffic Controller

40‐ 60

or uctuul dcsign c:rpucit1, has been recommcndcd by thc Trafic Engineering Ilandbook as 80 percenr of manufaCturer's capucity for 90 fpm (28 mgim) motor stlirs. lnd 7_5 pcrccnt ior 120 fpm (37 nrpml moror sruirs. Strakosch iecommends the use of 7-5 percent ol theo-

25-40

Source: Trafic Engineering Handbook,3rd Ed., John E.Bacrwald,Ed.WashingtOn. D.C.: lhstirute of Tralfic Engineers. 1965.

a humun shoulder breadth in width. or approximately 5 to 6 square feet (0.5 to 0.6m'z).

.1. Arrival Processes. Queuing and Peaking. Thc pedcsrrian holding capacity of public spaces is related ro the ap-

proximate limits of human occupancy of conlined spaces: All females: 1.5 square feet (0. l4mr) per person. l\1i.red: 1.8 square leet (0. lTmr). Contuct with others: 2.75 square leet (0.26m'z).

Uncrowded: J.5 square feet (0.33mr).

Dcnse

bulk

queues (escalators or

crosswalks): 5 square leet (0.5m2) avera8e.

Current lirerature gives little recognition to peuking rvithin short periods of

the peuk hour. Often. the peak-flow rate experienced during a live-rninute ratc flow is onc-und-one-half timcs srcilter than thc hourlr' llow rute. Thus. a 1,200person hourlv llow would probablv includc tlow at un l.ti00-pcrson rate during the peak live minutes.

u single pedesrriun group. such as commuters, are likely to havc higher peaking than those which serve multipurpose tflps. Employee charucteristics must be considered in the design of pedestrian facilities for emplol'menr ccnters. The peaking ol the arrivul rate ubour the startine

timc. ior *'hitc-collar *orkcrs. is suci

thut upproximurcly 20 percent of the cmployees enter ubout five minutes early.

The l5-minutc arrival time. from

minutes bclirre to five minutes ufter the oflicial starting time. includes 60 percent of the urriving cnrplovecs. A building thut docs not hilve all persons of thi same cmploymcnt group may have only 20 to 25 pcrccnr of ull urrivuls bcfore lb

A.M. in u singlc l5-minurc period. The dcparture ratc rs even more

peuked. wirh .10 ro 50 pcrcent leaving ar

thc nomini.rl quirring rirnc. Eight-v pcrccnt ol' all dcpurting cmplovees leave within six mrnutcs (earl1' or lote) of the

depcnding on the .sizc

TRAFFiC ENGINF[RIN6/ヽ 4AY!07`

retical capi.rcitv. as sho\+n in -Table 10. Eased on his studies, Fruin was ablc to drlu, the follorving conclusions about the dcsign of motor stairs and moving

Peaking of pedestrian tralfic is deoendent on the t-vpes of pedcsrrians composing the traliic. Faciliries rvhich serve

norninal quittrng tinre. Thcse virlues may be rcduccd to J0 to .10 pcrccnr ut rhc cxii.

[1:117

ヽomina:Capa‐

Table l

ヽ10、 i:num l・ :leo・

Steps

Spge-d

of the building

walks:

Munufirctufr,fs r€coffi fnended capaci_ ties of fiotor .,tarrs and moving rvalks arc based on the mechanical capacity of these units. and are unrcluted to the hunran capabiliries and trut'lic patterns which dctcrmine actual use and capaci it)'.

lJuman ch:.rractcristics. traffic densityi urrd thc prcscncc of bugguge have ail been found to have un elTect on motor stair use. . Computer simulation of the use of nlotor slairs and rnoving walks, based on observed boarding characteristics und trallic dcnrund. produccd a morc mcuningful and logical approach to desrgn.

' Based on the computcr simulation. a clcar queuing space for 98 persons, or about -i00 squilre feet (47mr) is needcd for u 90 lpm (:U..t

mpm

48-inch

ll2crn ) u idc nroror stair ut rhe practical working cupuciry of the unit: if this ca. pucity is exceedcd. a lurger spacc is re-.

(

quircd,

Nloving ramps arc gcncrally uscd on lS-dcgrce slopcs or lcss. ut a rttuximum specd rrf l{Lt tpm (-13 mpml lbr short .dista-nr;cs. und (tn 7- toll-dcSrcc slopes or .lcss with l nturittrum spccd <.rl' ltt0 l'pnr

(55 nrpnr). Thc width.ol' thc witllwrv

dcpcnds on thc sl()pc: thc stccpcr sl()Pcs rcquirc pusscllgcrs to havc i.lcccss to :: least onc handruil. Thc *idth is linrited to .18 inchcs ( l2lcmt for.u slopc grcutcr than 8 dcgrces. or a specd greutcr than 140 fpm. For ir slope ol'5 tr-r I degrces, and speeds less than lJ0 l'prn (4J mprn)

the width may be up to 70 inchcs (l78cm). l-or a slopc of 3 to 5 dcgrecs. the width may be 100 inchcs (254cm). and for u slope of 0 to J dcgrees, the rvidth is unlinlitcd il'the spced is less than

140 fP"n.

6. Trrrnstiles and Revolving Doors. The cupacity of turnstiles depcnds on the type of turnstile. and if and horv money or fares are collected. Thc capacities of various turnstilc des.:ns ure given in Table ll. The runge o'capucity valucs reflects such elenlents as commutcr ver-

noncommutcr trullic. age of the users and other ph.v-sical dillerenecs, The max-

sus

imum value should onlv be used for

groups, such as conrmutcrs. who are familiar with the operation of most of the equipment. Truly- noncommuter. or casual. trallic should:probably- be lssigned the lo*er valuc. Fucilitics uscd lor sporting cvents and other l'orms of entcrtainment should probubly be given un uvcruge between

the two cxtrenles. since the

putrons probubll represcnt u mi,ring of those familiar rvith all thc details ol'such equipment and those unl'amiliar with their op'

eration. Revolving cjoors huve a theoretical capacity ol' 60 pcrsons per minute each way.r The actual ctliciency (buscd on the maximum l5 revolutions pcr mirtutc) is upproxinrutely E0 to 85 pcrccnt. und rcpresents a cupucity ol' ubout 50 pcrsons per minute cuch rvuy. The ntore usuul

obscrved value. for rclatively busy revolving doors. is about 25 pcrsons pcr

minute. which is thc suggcsted lowcr design cupucity, und a maximum of l0 persons per minute.

Eased on conslunl tralnc denand ond

ing

retulutions per innult.

:」

a door hou-

lliblioErephy

,\llrrue hc. J. l; . "Approsqh to Prohahility" Oistri-

ol !lluc ol \\:rl\rng lrrng und I'c(lc\tnitn irtron." llighua.t' Rrsearch Ruttnl J9J, ?. l2l.

hgtrrrn

(

irctrf

.t(f71..

' "1hc l'ctjcslririn ( otrnt. : \l'rshinst0n j I).C.: ..\nrrflcrrn s([rclv ul Pl.rrrnrng Otlierrlr] Relon ,\o. /99. Juilc l(r(tllcnctlrct. lr. (i . rrrrd ll. S. Itrnrnrctcr. "[incrgr llclrrhllrrrtt rtl !\ rrntcrt \\ ltrlc ,\rrcntlrrtg ,rrr,l l)irccrrrlirrs 5trrl\." lrrcri('.ln Jutmol ol Applitl Phvs. lulr,gt. l9 tt llrrr. Pirul C.. "Prrking Ccneratirrn Studies." \\'urhington. D C.: lli,uh*ar Rcrcarch Eourd (lligh*ur Rcsrarch .\h\trlst). ,\prrl 1962. [frtrcr. iU. R.. Ellicien<'.r'o.l llunan Ifocentent, lnd Ld.. Nciv Yurk. liY:\\.8.S.rurtdcrsCo.. 196{r. llro*rr. S. lU.. "[,lrkrrrg Pnees in thc CBD." Srtciu-tnnon'ic Planning Sciezces. Vtrl. 6. No. l. Aprrl 1971. Crerghton. Roqer L.. "Rcport on thc Wulking Trip Survey." (hrcugtr: Chicago r\rcr lrunsportl. tron StutJy. Intcrnul Rcprrrt.

1961.

CTA. "Pcdcstrirn I rullic in [)(twnto*n Chi.

cruu."

hicruo: Chicrgo I rrrrrit ,\uthont1. Nov.

l9_i9.

CTA. "l'cdcstriln Nl()\cnrcnts in the Shopping r\rca ol Dusnto$n Chicrgo,' Chicrgo: Chicugo Transit Authority. Nov. 19f0.

Davidson. Pcrrr. "l)irtnhutiun of Pcoplc l'rum Ruilrord Tcrminals," tvanstun. lll.: North*crtcrn Univcrsity. D€pt. ol Civil Enginccring. r!1.S, Thesis. |

970.

DiPictro. Charlcs. "Pedcstrian (iup /\cc€ptrncc ut !lrdoloek Crussing." NlorgJntusn. !V.Va.: \\'cst Virgrniu Unrvcrsrtlr, Unpubtrshcti Thcsis. l96tl. Lrgun. Crrkmen. "l)cvelopntent ol r l)ountown

i\lodcl." Chrcugo: Chicago Trunrit Author-

Parkrng

. Chicugo ArcJ Trun\portution Study. Nurthscst Chicrgu Corridor Nlodrl Split Project. Rccport No. it1.

{. Julr

197

l'rurn. J. J.. "Dcsrgning lbr Pcdestriuns: ,\ Level ot Srrvici Cunccpt." Erooklrn. N.Y.: Polrtcchnic lnrtituie ol Erookl;n. Ph. D. l)rssertuton, Jun. |

97(). C unt$ oort. J. Th..

"1 rrllir' (icncrutiun ol'r Lurge Dcpnrtnrcnt Storc." Irafic Enginrcrinz + Contnl. pn :ll5-:')-1. Ocl. 1967. Qrrbrccht. Drctne k. 'l hc Brnomrnul i\lodcl ot' Pcdcilnln l-hr*s-lrnplrcutions and Lrperinrenls.' Tralfii' Quorterlv. Jull 1969. (irrbrccht. Dictriel. "Pcdcstrrun Ilorcmcnt:,{

Bibliogruphv." \lonticclkr. lll.: ('ourtcrl,rl Plunning Lrbrurjsr. tirehungc Brbli.rgruphr ll5. Oet. l97l (iurbrccht. Drctrrel. "l rcquencl- Dtslrtbuttons ol Pcdcstriunr in r Rcctrngulur (irrd.' Jownal oJTransport Econonti<s and Polict', Jan.

1970.

Crcv, ,{. L.. ct rl.. "Pcrrplc lnd Do$ntosn." Scu(tlc: Unrvcrsrll 6l \\ lshington. Collcge ol Archi-

l-ittls. ,\rthur l). "('cntrrl Cit! Trnnrp(lrtrtron Prulcct: Llrhirn Trilnsnortiltion Conecpts,' \,\ashingttrn. l).(.: (i.S. l)cprrtnrcnr ol lrun\p()rtuti()[, U

rbrn l\l irss f riln\trortJtron,,1,1 n11n1511'11s1rn. | 97(1. Luucmlrk. 0., "1',,1c* .,\pprurehcr tl ltcrlstriun

ltrplrciltrol\ ltr)nr StuJtcr rl l,ctlcrtrritn Echuvrrrr.' l'rrir. I ruluc: Orgirnrrution lor l:crrnonrrc C(xrpcrrtr)n xnd | )c\ chrflrrcrrt.( rlrrultrtir c (iroup on I rlrt*n,rrt.ru,rrt Rc.carch. I r.rtr\fiortntt()n Slrlcnl\ lrtr NlJ,rtr.\rlrtrl\ ( cnlcr\. .\Pr lt)71) Ilrel)orntun. l-.C.. "r\n In\c\tt{ltr(rn ol ltcdcs. Prohlcnts:

lilnn fravcl

Snccrls

in lhc

lJusrncss [)rstrict rrI

l)rrsntown \!'ilshrnrton. [).(.." \\'arhington. l).(.: Cirtholic Univcrsitv ol r\merica. IVlustcr's Th*rs. Muy" 1967.

i\lrnning. f.. "Pcdsstrian Trallic

Chaructcr-

rstres." Ncw H;rven. Conn.: Yalc Univcrsity. Thcsis,

Muy

1967.

lvlu;-cr. ll.. "Dc.lrrdividuation und Other Psdcstriln Chuructcristics.' New lluvcn. Conn.: Yule Univcrsitl.', Thcsir. l\luy

1967.

Oldcr. S. J.. "\lovcmcnt ol Pcdcstriun un Footrays in Shtrpping Strccts," TralJic Enginteilng t

Corrrol. Aug. l96lt. Oldcr. S. J.. "Thc

Spcer.l.

[)cnsit1. untl

on l-uot$rrs rn

trian Ffow Churactcristics," Trafic

Engincedng.

Junc 1969.

Ness. Nl.A.. ct ul.. "Anulysis of C€nlrrl Business Distrrct Pcrjcstrrun Circululron Putterns." Highvay" Research Recoal 2!l). 1969. Pushkarcv. Buris. und Jcll'rcr lunln. "PcdcstfrJn Truvcl l)em;rnd." Highwoy Renorch Recod J55,

t97t.

Quurnrby. C.A.. "Choiee rrl Trlvel lVlodc lor thc Journcv lo Work." Jownal ol Tnn.rport E<'onomics und Poli<'1', Vol. -1. PP. 17.l-.lll,5"n,. Rcgional Plan Associatron; "t'l ulking ''r6t.Spacc in Citv Ccntcrs: lnterrm Renrn Summarv." \ew York. N.Y,: Rcgional Plun Associatron. Nlay 17. 1971. Rosenhcrg. Ccrhrrd. " lhc tllcct ul thc l.ocltrun ol Trunsir Statrorrs un Buildrng Drnrrrv in Inner City ,\rers." Hish Speed Gruund Transportution Journol. pp lt5-!$9. Ilul l9{:9. Strrkosch. C. R.. Verticol Tran\noilation, Eltdtors an<l Escalators. \c* Yr.rrk. N.Y.: John Wiley and Sons. lnc.. 1967. \l u1nc. E. O.. "Thc Lll'eut ol Trcrilmill Crudc on Pcrlbrmunce Decrenr<nt Using u Titratrun Sched-

ule." Fort Knor. f.1.: U. S.,\rntv l\lcdieul Rs,\pr. {. 1961.

seursh Labtrrutury. Report No. i-I5.

lccturc und Urbln Plunnirrg. l9?0. Crucn. Victor. The lteart ol Ow Citics. l4ew York. N.Y.: Sirnon rntl Schustsr. 196'1. llrus. R. C. C.. und l'. l'. Nlorull. "Pcdcstrirn Circulation Through u Tunncl Net*ork." fra1fic :19. ,\nr. lu67 Ouorterlr, - llunrburu.o. Jtlhn R.. "\\ rlk lo \\ urk Tript in thc C,\TS Arei." CATS Reseorch Nrrs, Vol. tl. fio. lt. t95r.r.

llankrn. B. D., and R. \. \\'nght, "Pedcstriun f-lorr rn 5uhurbs." Operatiuns Rcsearch Quarterly'.

pp.

-x9. 9(: ) 195x.

llrqhtar (upa.irt',!/anual. Wurhingtun. D.C.: lligh*ar Rsseurch Borrrd (Spcctrl RcPort lJ7). 196S.

llucl. L. A.. "l'cdcrtrrrtt l rilvcl Rutcl tn Cctrtral Burincsr Dfslrfcts,' TroJlic EngineeritS. PP. lU'lJ, Jun. l96lt, Inghs. Puul. "A !lultrrnudal Logit !ludcl of Mrrdul Spht fur u Short Journey.' pitpcr prcscntcd ut SJrd Annuul illcctirrg. llightuy Research Board. Wr\hrntton. D.C.. 197J. Kunu'iani. Adib K.. "Lueuttutt i\lutJcl lirr Purkrng

f- ucrf itrcs," Trunslrorluti,rn Engrneering Jounal ol' zl.S('8. Vol. 9l{. Nu. l 1-1. l cb. 197(). Lanrbc. T.r\.. ".Ihc Chorcc ul l'rrking Lutiution by Workers in thc C BD." Trolltc Quartarl.y. pp. J97'

.il t . t969. Lisco.

T. 8.. "Thr: Vuluc rll'C()mnrutcrs'

TruYel

Time-A Study rn Urhan Irlnsportntton.' ChicuSu: Unrrerstt) ot (hrcitgu. Dept. ul tcunomics' Unoublishcd Ph. D. Drsscrtutron. Junc

ljhr* of

Shupprng Strccts.' Cro*thornc. Lngluntl: Roud Rcsqrreh l.aborrtor1. Dcpt. ol Scicntrlie und lndustriul Rcscurch, Laborltory Note No; LN;trOl' SJO. June 1964. Orzeskc. John D.. "llor to Collect Dilla on Spcciul Generators. Chicago /\r€u Trunsporlatlon Study lirr I'rcscntulion rt Origin.Dcstinution Cummittce lvlectings in l\ashingtun; D.C.. 1967. Nuvrn. Fruncrs P.D. und R. J. \lhcclcr. "Pedes. PcdcstnJns

1967.

TRA「 FiC[NGINEERING/MAY !976

υ手 ―′′ ζ

l.つ

Jぃ

ゎ丁

nttR

θロゴ■

:塀

●

lN Pら NIC Sm,Attα ttS

‐

僣黒群黒 :計畠 iご驚 Lげ凝正、斑、れl=1的人ぬ0■ ■4チヱゅ颯疑し Ъ .ら薗マttЪ てこ3づ゛ . 汎曝t′ _ Q番 ■■.ヤ lll itさユ亀付け3藪 ■ ■ 十マクリ、ン ■ )マス｀ ■｀￨しヒ肇塑Qヒぬ駄Jtt tイ〒,■ 、 16 (1) 4」しつヽ .

○

PRFrrr SlruRTrtrns" I*Fi"" Enpncrrv a{gggJr* i!-Eory_r Vnu. ?6r flo) Er 1971r e, 6e-t7. (r* Durcx)

■麟む

s;TurrtEs; ltFIE FTEBUEHTLY E:lt-ltrTEf,i Fg EltrDENtrE ElF THt Ae=cnu's lor.""o=LEHs l'tHrtrH couLB l?EguLT FnEH F PFlNrc srTURTrtrN. Hg T:こ

￨.=1 1‐

TYPE ・

、

￨:￨

LEEE「 :軍 :17VIESTS ::::ETi言 :言 LT冒呂IHT‖ :‖ EFE EI口 NE LIS=N● SCHロロL CH工 LDFEN 工‖ ‖H"T "RS n t/EEV EX「 =TIN● ■ 工ELD HR言 THIS ‖ロロ THEE FESEnFCHEE′ is=Tり RTエロ N, =N′ . THE FFEEUENCY nNコ E‖ T 口 F nnCHIN● ロ FSE"ν EDヽ日‖ TTH工 ‖曇 L=HE THrS Eν ′ PPEEI IIFn‖ RTIcRLLY コ薔□ HE‖ T 5 EXPEF工工‖ ‖● PESCHビ F nPCH工 ‖ 口 Iコ LIPRT r口 . . . F THE 口 PENIN● 口 =工 !工 NCPEnSIN● =F :銀 [F讐言

に二:、・

日Ts

EEttT‖ .ITIIHttcII言「

‐ ・・

1■

:1.:

I.=■

itユ

:■ 、 :摯 ■イニ J: 「輩￨ltヤ ilヽ ■

■二=ギ￨■

‐

‖

=TH

れ

itIIlkIItu:[I::, し n bij Pa‐ ・じ ni・

′

いギ

Dr. ing. f , !..i),2.Pesclrl+ Iottiv, i971 , 為 (z) 6z ot 一

ヽ ‖夕'F 1 1、

・計

卜ヽ４ヽヽ・ヽ

、

口

ノV ト

一、、ヽ

Lリ

｀ルt

ヘ ′

・ｆ ′ 〓ヽ ■ ■

f, Ie<..ttrre r;jv()ll :i.n courl',.,r Itlla('cty ip lruil1:irrgsrt +.'l'uciri::i.r;clt,: !:clcr:i:jtoo.L

. :i

j r:.Ji:ov-.rt

:11■ :lri

FIЮ ″ CA「 iCITY 01・ 300■ OPifII:・ Cr,III Ilユ I!rC SITll∫ 1110:IS= (D001。 3t10mingSCapと

t卜

ヽ一

o{'lreoplc jrr at br:i.ltiirr6.i5 l-tc).\'i!..ltr.y!; lili'rll} tjruc's 6r'e:rtcr t}r:ln itr tlre pl;st. gne orr)..y lr:.r.g to thj.nl: of'tlrr.: croi:,i|.$ 1;llt.herccl in a nroderrt (:irrtnrirr Lh,rlrtrcrs;Lor.rrr;1tcl1gl11gntr lluildi.rrgrot'f.icc blcc:li or sl,orts I'jr:ltl. iivsn ai.rer:rft nr.u trcirr6 s136lgr to climcns,ioris sjrnj.lur Lo tl:or:e of a tli-atre. fr::r1.1 thcge ci,sclt ii j.s irnpor.tarrL to l:no.,r' hori rrui.ci'.l.y tlrc cror'.'(l.5 6.r]'c ilbLo to l.e€ive the btrildin6S in the evcnt of an elner{:encY i.rtrj.lc th.c f.1o..'r pr.opcr.li.es of' g1rs(:srf i:-uids anci p'l.rrstiC'a)ly viscOus mctterials (perhaps trave been exiiaustiveL;r investigartecl.rj.t lrls lrot co far bc:en posliblr: it lrag not even beerr aLte:rnpteci) to esteb.l.isi, tli<-:laTi goverrrirrg tlre Pt:tjtiace of pcol)Le ip large ntmbcrs tlnotrgit <loors irrxl nat'l'or,r l)erssa6es. Tirj.s 5.c l:robo.bIy &lso due to the fac'., ttrat there is a sPirituaj. r'esistanc:e to tire seplration of hrrman behaviour froin the hulrnn yri1l r:nil to strb;ecting it to n:-'tttral lalrs. '['he nunber of Jreople sta.r'ing j.tr an object suctr as cr 1::r'1.;c noC.ern sports staclium '.ltith this increase hsrs rnulti1,LJer1 by a frtctor.of ten in the last fc:'; dece.Jes. i' the nunLer. of people the d.ange'r of crol,/deci conditions I'nd in special- cases that of er parri.c s:ituatibn hac alrio bcen increasucl, Ln the case of a lnnic can one rely on ,nalr s reasorl? Unf'ortutte,.'ue1.y 'rhis tloes not aP}c€tr to be thc ca.se ariJ. af.ter discussj.on i;ith a srrf'ety l;pecialist -'r.t is noreover clea} to tlle author that a parric gitiration is often I'irich rnorc serious th:'.rr one-cottld ev.rr irrrag;irre. ?ake for e>:om1,}e a sc[oo1 in r','hiclr thcre are over a thousatrd chilclrel v..ho 5:rve to get outsicle throrrglr a cioor oftc'n uot rriort: than 1.5 to 2 m nirle. :'jhat .;.'oulo h::ppen in the case of a fjre'i Tt surely happens to arrybocl'y sometir,res thaL *.1cn lca-ving a cincta.'.1. or thecrtr.'e lre bcccrto;-rs s(.ueez,ec[ in t]B cro:sd.rbut this is lhen nrurcly a d"trscl.y cro'.';riecl situation in iJrich ihe }eople remain calur.so tlrat the.y:rr.c able to 5ct out of thc squeeze a.6;ain. Onc shoulci try to ima'ine xhsrt r.;oul<I happen jf ittsteacl of ar crush of peoPle a i'anj'c r:'ouLd[ be ct.ca,,;ed e.nci crll urc l)cor)l{} u'otrlC lrtrai.n to get to tlre e>:it rrt thg clms noncnt in l.ear of tbeir']i1'e. For.trurirtcl.y 1;:rnio sitrt::t:ions arise rarely encl the dis€rsters r,,hi.clr re$ul t ther.cf'r'oiu hi.rl')l.e n cven r.ol'e re.rcl..y. If a clisastcr occur$ as & resplt of fr pir[5-c rp)xicrorr.: cirriur,lii.cs !t).'c inun,;'iljrrteI.y lccorclecl: getrcr:il1y it is Latet uror.rght that thc rlis,aster couLd h:rve heen prcventeclri.f in tlcsigning tlre 5uiltli1g sorrre thotrillrt lru.l buctt civet'l to'.;he 1;ossibi.l.it.v of a patti.c. Tht:s irr Lirna (gA, lrun<h.ccls of yreo;ri.e ii'oulti not. have bec'n ct'ushcd to ciceth if tlte docrr of.' thc st:,r1 :itrnt hull o; ctrcii ottt''ilrr.'l ' in (:iijci::;o th': livcs <,t 5'lZ t)eoltls cor,tlci t13.\.1: lccp .iil\:(.;.i.lrr tlr",; f.it'.'i,t tlrr: Tr'.il:o:i:: ifhuater jrr'i ?Oirj('tlr': t'i.Ls o1'cliscilai';ctlrrcrtt;ltt'lt"'ilo<'t'sha'ittutl)r;rJtll'trtltlcc(ltoitfrlrcti'ottc'ft;lr'r rror.1r31 r.irt.c (.s ir r.r.rltglt o1' 1r:rpi.c. 1 .!lir':ittg thc r'irc rrt tht: iiirri;t:ltcnttir ut ViC,p:r:r:i.rr 1i,ii1 lrrrp.jrutl:i a;1's111iull.l Lj.r!:i iroulri trot ltr.v(: bCr-rtl bttt'trL itli'lCr.if t.hcr ,Jl1e nulrrlrcr

￨○

￨

●

l),.;ol)]'(,

a{'

Llrtr tlcl<lr'::

jtt :ltt c,l.clrocl ftrrlnirtj'on,tittl..i b]crckini! lrll:l ttoi; lrucr.lIli,.. ..,;cJ{'(:d

rt;r tltu t: r:i L.

Itl:tr;;llct:ttfctuttr]tlro.tthl.,ll,,r..rr:^viottrol.lrccrl.llcj'rral,snicis;rttr1lr.ecl!.ct:tb]c irr.ltltct.e;,;9r.rlnCtluot'eno(-llt:ln(:(:tcclstlll.lj.:;lratltcot..yf.clr.it./rtrilrvrl..rtitSlrpossibl'e ,;tiLltr.rr.tt stagilrl; ar djsast+r an,i ;1c,gr:1ltin6 c: inrllcssi.l.r} tj.on is tltri.ti: lro}i:5si":ritif'j'e' rip'r'ovjtrt;t':i-oti is t'lrrrc {r coir})l'cl',Iy (:tL:juirlt:i.c..J: tltc car,e f.'a jnrpossjb.!.c tcr brirrg tlri-s lroblern llrlars'to a r.lrritu j.sr 1.xg! in r*ss one. rt js rrljr'l'c to r:'3rtr'l peoplc €ts rtcop]'e jn ftrl'l posefL'ort Cll ir.s lotrg solut.lonras l'lacaP:tc:ity' Can ont': not: i'init scnrc other anJ mcntlrl 1''lr'ysir;a'1 o1'the1r secsiott trrra]'pltuttotirutttltr...;j.tltr.l}t:!clr|,lrcrol.ttliirofl)uoIl.!c{.lceinci:1>atrtccirnbuaf)j)roabeetr e>:presrri:ry r:;e]]. be possible,as has tlti; tlrat cclttsit<ier's auilror Tlrr: clrt.:d? irt butt]:ers" ' 'Lire f'or':rr of a prololiitlotr in his [hr]'sic "li'rcirirrg cecl by h:i.ru j.n are bcixtvilrg like ir 6;l'atlulr'l- nlass r".hcn a in leople 1:arric eltcitpil'tg as fol.loi'rs: I'lor:ring out of a br'ltll:et'r ' j's not a btrn":er'br'r'b lrhen lteoFle €''11'I a hr'lI I,tatr.rr.a1ly peopl.e af€, oo'v 6rainr cio not convey thcrl at tlrtl ciesirea rtc lonScr. usc tlre-ir l'e&son itnd, tl:e:i:'feet s'peerl.endj.nthetlegirer].],irectj.o:l'(tSj.si|heci!$cin'tv(:ryc.r'o'.r1.!'erlorpa:ric sit.ualiiorrrtheyareno.!.o116-r:rc1i:trtctl'cr)ofthctotelc'oe;:ingdt'finit'f)l€tt';s'

Solne6elreralc}te.rrcter.is.i;lcsofanassofpcoirieirrltetrlccou]d.be'6ivcn t'c disch:rrde of a grarrukrr n.rss frorl coi.u):rxi.<.i:n...ri,r justify're r.:oul.d rvhicrr pani-c sj.tuation cert be il:istinguicrheil ,r.r*llel.s arxl b.y ulriclr l,co,):lc in a cr.o',,rc1eii u5.thr:especttotheirbclri:viout'i''tolt:thatinnorililsitriatiOtrs'Ttrrrplnj'c as t'uj'ci'J'y sts poseible' If'a larp builciing iba of out get to i:ants cver..-vborly to *n racctu'u"g-tiolrr of 1'"cbu:irci'gr^ulii:r thr, i.n ar.e pcople oi nnr:lbcr 'r:arrl; Plcnclar.tltetr;<itl;.Pcopi.eurepttsireisoclosctogetiier:tlr:rttlr,..lylraveno -r'i'ot'e. effcct tnet'e alor:5 i"it' tire strel'ti' to ltetve t:uci rriovetaent of freedoiu closer tttey tlrc to orte a*otirer"tlta vrill tire' be encottnt::r'ei.nrrrrely tlr::t t'e j.:r trr,-, tresiretl ilir,.r-:ctior oI'rnovencnt. Tlris less a,rc tlrcy a're to o'cn out inic iu lr'*rsctl c''rll l:love olrly if h": citn 6;et tlurt fitct by'uhe ezplai-trocl be cirn titet hc ci'r1 u'cil'iz:e'"lr': f'ricti-ou f'o::'cc bctan incli'el1 st,srrlt'I.i.*g 吐 'o.';'i"ict:,so Tirc {'act t'itrti; ti llclicc ccrl'iorr is {'or.r't'o;'tt1gj'ot't' floor tlrg irtrtl u,cen thc sol-cs 午 ori ilri.:i l,!r(:nour.:non. lilrt': ci'oiiil is cierrsel'"'r birricti is ci.oi.rJ a b:.rck irold to :rb]-e 一 u**ble to get star-Letl. iloir'ovc'rrr 一 tht tirirt so 1:.copl.tr in ii irr.C to;ct,iier. 学 ]rfr:sliecr i.ts croi';tl ir.trJ is tirtr:: nble !o dcr'eloir j's irSai.rrs.L.Llre 1t:*rr:i15 co}..1on t}rc ltolici) fnl1 lioxer. lo itl tt lヽ :iniC j]_]こ iゞビ at thiら C()llCllt,ヽ ioll tll:tt P::。・ :"t'grrti:15 irr t'l:il; rffty,い

■ ″

llr .・

ol' (i.clli;Q cr-it',';t1 !tltvt: lr,'rti t! vu!'Y lllivcrl:t''

llill｀

cl・

1・

1｀

:I・ l・

:11● Vヽ・ 1:ty〔 llly .:ざ〕ヽ

cct On

be f;hOrn lttt・ t01・ , ｀ Ot:」 nti:111,fh i Cil,〔 tS Wil■ 】

ウ■ 1,11、 , di:〕 chitriit, 1｀ tttO.

, ~ ー

1l1105(: ●blc

tO pュ 1:s thC

l'ollo:l'irrg trclllll(t tli1vu ev '^r'w trr l'tln.

tttu rjtr,;sLjor: of

ItoT' rittj.cl:)y

b:r1) rl'ttil'clit16

ot' st;r''iitltl carl be

evactraterl

itrtllef.,vt.,lltolle.nc|llgrcencyis;t}rc:r.t:(.or.t:oI.1lr.e.,rtitIrpoi.tt.,'nccr.Itisobvj.ous jn n'i.clr t)re irrrli.vj.{urrrs irl'c rurll'rirr6 at a $'acing of co'':rrr l,,joprc a Llrrt

(~,

⊂〕

Ｐ

ol's.:vcraltrtr:tl'csbchirrioncutrotlrerisablct':li'ovcvel-yt'ttickly'buttltnt stcp-by-stc:p' Itr oth':r jn a cn-rrirlcd sho|iri.tr6 street onc ctln aclvunce on1'y uj'tlt irrcruitl;i116 dcnsity of the croYrrjof clecrcasr:s sperr-d vl).!:irr3 tlre i.;orrLs: sllcetl,tlre passtr6e rat.e tlrrotrglr i;a1}:irl6 ].o'...ci. the of t,.ic,,,..| In l,coi,].e (},5.o.t;. dec'cases u'til at n:i)::inurnratte'r'ihic'it defi.rriLu a to i.ncz..iases tlrr: rioor rate bcc.rt'js zero'as s'ot'tn in t'e !'e hi5'est possible dennj.ty the Pas:j*Ee t'ttt ti"""ri.6ations in Japonrbut in this f'clrr. hi,.vc been ".trrthtnnar..,icu'arly itr the street or under comParable cirexclusively f.'or situations prevailing of t'e p&ssage is con{.inr:c[ to tire risin6 1ra't stucy thc ,rat so cuustarnces, rninicase o{.msz.i:iirrnn ctensity and. hence s is it situations panic rn rate, fDont of a d'oor only th"': peoirr crush a In speed. llaLl:in;; - trnurr or even zero pler.;irohavereacheo.apositionirrlinev;iththr:cioorarestSletoruna.;ay free space)rr':hicir reduces t5e c*oi';diness' (erssucrinS tir.lt they a'e then in a tile presstlre e:lflried' by the throu6ir rnainly aclvgnce j.n nass a cro.'rdecl Peoiile pcoPle in tlrc ro",li belrirr'i' of an arc h (n.utuar i...e,l.- _--.,,,,r {!-,,r {a6i-n In vii:ii' of the Jcr trall:iug specd.rtlLtt'otmation on the passage thro*gh an.:tit ging i' in gru forn of a bo'r) f,n" ntiiiieci (t.'ig.z).Jrrstli},.ethei*rcltalroveaeoorr.,ay,tiledoneofac}rtt:.chort}e in the vi1ic5 tle st'oucs 5ave be'en ($ig.1),in b'i'i5e stotle a. of arch 'Iacecl the l-oadrpeoi:re iroving to.,'rards suupcrt to c.ble tlrus are anrl sh:rpe of an arcrr to';ral'1'acting as a d'ome over the vaII from rutrtring arch an foi'irt an exit can door.,ablutosup..ortt}repl.es$urefrcrrntheremirirritrglsass.Asaresu].tt}te e):itc:rnbecnl:ir:corlpletcJybloc!:r:tl'foi'aslrorttinre':'rlrichisagreat5'rnped'ipe.ple (:rig't*) nerrt to tirt: r.ii;ri-ci piistjage of anil cgrtt such in arclr cicl:1.:rrtis on nratr;; f.:ctors The charrce or. 1>e<lp'e I.o:.ii:itl' this pltenot':cnon is ir very corr!'1:lcatcd not be pr*dictci tlreoreticall.yrbr':uatt::ie bc obtained orrl.y lry experitrt]Dt$. iio...;cvcr' catr t."'sttlts t,ttr:trtit&tj.vc oot) BIld r.ril1 g;et tircrtrar1'r.re1;s tr chattce tirat 1lt:o1l'e otre ttritr5 is cerL::in: tlrcn: is tlt;rt tlre chattce of arch fo:'rnetion is cvicient iri It j.rrbo arclt. trn selves crc:ri;8I.{:ljtil\):rr.Gl1:rtcir:il[)o.sctttu.1.e;,J,-:1.].eolrl'e'$ittccrii-i:lr:r..tirletloctl. tiurrr:'.itlr a snllrll tli,: *r.clr irls to rrc n:lrJc u''y ir f.*r.gcr'r-u,rlre:'o1''eo1,lg

Ｆ

-l;

tlr'rt')l'r'lL''r'L r'"

---..

(lrl(.i- l't uiLl ll.tio L(! b].r::rr 11i{'l1rr''rt .....,-. lli!l'l'o''? ()t1(:' t';i'til il,^ r..rr.r.rr., tirrtrt r}Oor iriut: r' lr.:g:: li'itir .':,tlri': -"'lrtt:n::u i'r'clr ui'Il' lt:'vt: il ----.-"-,.: t'it:rL c'i'''!uttu": of' trelrl (iruu'r' jts lr:-vir:3 Lo l'l't'dttcc il t,rt..,,i,ilir;ir jtr virlvl rrf otirur. tltr.: Ljllrt '.611-j1ir:;r.|:ittlt1,e t,.,.rt.u j,rlul.l ;;urr:1r..'Lr.ic h* precaflnot b'iriSing 'i'iri':; ilo"rrr:i'ilrorL tjl;'t':' trr'.rll: *lill corrst''ctiott to trgl:cl tlrr: tlools clt(;u'i'r'tl ";lt';Lircr it is 1'oss:ill1c onl:/ c:'rn on'J i:l)'] vont(:r\ v;ithsuclrclillr{-'rrsi'orrs'c):lrtrrrr$t;;bJ'cttlg}1c:;'lolri;Jlor'nt'':rl'ttsLtr.g$tlltcrI'l'lhicir cri:atecl' the f'tiglrt of irrclt 'is tiynr'rrt5'c ir the ezit srortlltl be blocl<ccl' "'i'rcircvur pcopl.cisclol;6ctltt1l.(}lr1et!.rr.:dyriailic:lrcliitirsb,.:etrbl.o!:cn\u}1lco1r1e$l.cth: rr 1.:.csir ci;,.nltitic arc!r is l.or.'rtoci. trnltitrt]|,re(i'rr:rti.' outgicle ab].e to rttn tecthr" oI. tirrc t,en trrkes thr-: {'or.n of "rir"rr functic' a as rate evacrration rate occttrrth.;Qot'er) is t1e passlrge tiynamio frecuent:1..y less t5e 'rchDs the ntrnrlrer d peop).e nees.ruller r,lic]'th,thc rloor t}re srrurl}er (rig.5). The of rilnauric gree.tcr thc frc'1uc:i(:y of form[tion thc archr'..:'ce an forn to decl stallle arcires,r.;iriclr lorrer tlre plrsstige r'ate. J,s reg:*<Is

arches atrrl tlte naylreregard'eclasaspt]cialcs.$eoftl;-neinicotres,itwil.lbefourrd'that i-f tl'e rrrch str.uott*.,:r:.;ir'i.s!r i-s essential i...c*r, ress or nore a of charrce - the jncrcesinS <'roor-.;;irl,r. 'Jri.t, a snfc rij.rlti'r of vrjth is to be str.br-crd.ecreases r.rilL b.l;co:.c ?,ut'o. rt is vety ilnr''cr'uatrt ar.clr str:-.Le o. of chancc tlrc <tocr 're thatthiscri-Licalliiathshoul'dbr:l:tlouna:iacctlratel;laspossib)'e'sincer:'ith foi,tn:rt.iorr,tirougit sna}I,nay ]eacl ar.ch of ch,,itrce tlie <]'ocr a sl.ightly sr,ra13.er be fotrnecl inrr'r::c1ia'terl)'nfter the to h:'il>etrs errc:h stltrle a if to a ciisariter

Peol)lehavestelr.ter]'tol.unout.Itj.snotpossillletod.eter:nirlo-thjscriby â&#x201A;Ź>:pcr5't;tr:nts rrit'h ste*I balLs eitltert irccurac'ytnot cotrPl's-te tical lridth'.'rith j'ti thr': vici'ni'ty of thcse critiertl l'utcly etrcotttrtcred bccause stLble archcs

(trreoreti'clt'ly

extena. ovc). a cal vrioths a'd 3,. e:r]rel.iurent'as to str'l11e a'ch is possible' Ui:'enc that cr'''jttre to ot'.j.e' itr infirrit.:1y lo116) fornted'' tirc rrresr in i'rhich stebLe'arches are in orlly conducte'l be can rj-ments

rorrg

t-i-rrre

I.rornr:lhictrther.ljotlroftlrettoor.(or.tlru,ui.6t!roftlreotrttetirrtlreo:iseof of' cxtr*1:oJ'r'tccfitt utr:i'c' t'e ch'l.ce catr trells) stci:1. cxpcr.irne'ts'nit' 'e s:airl'I' verY is arch formation E x qe

ri'.r!l

-'1'r!U-l

:ccl'f Lc:

Thccxl:erir.rentalsbutlies.;.,5.t1t1,...o1:1ci.;er.ccg::r.ietlout..',iththevolutrt:try '.rith thr: r.c':r'Liv:r.y sn.rll rrtrnbi:r' o{' 1'eop1c strr,ie^tS. 1!O of corl*boratio,

itr.;csrlcce.g$eltyt<lfor.rnnC:Jtlsedci.r.cultofi,lr.:str.r0:ttofpeopl.esoesto :"o!' 1:llis l-\lll'ilojit:!

tlr.' ilrro' fo:' r [";; ir;ittr'tl;":s' proilrrec: c.O'.';ri:il11; irt f'r'ottt' clf | otl L'ri tt t !''ir 6"'rr')l'i ltTrnt::"1 t':' tttoJ i t'"t l.\:i t,(;l'l-cor::1-i i'.,i,i:i lr-ll-.!. "tl tlru iu ii1;. ii.

-5Â¨

ir"' ortt:

:;lr:l"'';n

/ 1'

otlr| ig c:'lo::ed '::liilu '''ir": "' 'rl rr''' ol i'lrc cr'Irtl':t tn irid"lr' i'! t c'f c't rr---''" ":":-.; ' tioc'rs ') ,r.irc: Lslr lrll; th'.t.'; '.ir'!(rrs rrr:trrrll ftrrrDr d(io'!.' r'irtjP';t:tivcl'y' .t dool't :rtri ,,.,.i t1r:t'tlfll """ l1'1;':i'-lltlr'ilrl: rlrri'::'-l:tlr'lI1I.:^:";:',.,r.,.,r-.,,,r.r.j.r,a1. rlli fri'iths: ':"- rr''rlrl used o!i orc, oul,:..:i.l.h tlrrce d'oor ("trl.i''''d' (tu' llrr: otirui. r.,,'., tr(:rc e oI' trcop'l ,llrt: r;reuliulultl(:tlt!J crf tltr': f'l'orr

ion i s not poso'",'o"" ll.J'"'\Y:13;:'''"'visurr lbsc'vr^t nt otr f ilra' uilm cr:inecr't:r'i'rir: r;lror.c ir j.t urris accj.derl. to rc'cor.. eibl.crso t,rit: r€ts,tittr.J2lrad'afixctil.)o$it:ioll€[l'(icontitit..ous}yf.i}rllcdtlrecyjttl.oor.Clrncr&

' i;ii';',,1i*' i

of r:::;r: T;x;"" 're crush 1,cop)c::.::::'::il:

orong:^ti'!re as possible'

itvtasncccligi'ry,invie.*oftit.:srr.ttlltrtttttberofpt,oplc,toc{rn-,1-l.trci€rnitl.l.o'.., those stuntriing of 2.2 ttt' lrrrseagG d.oor (r,,li:.6) to herp r!^-:.-1 xr,'irrrt 're

.i*lo';:';::"""::'::::'Jl;:'l:": jrfi:;::*:n::'"::"::J",:':

lte<lbyasignal(utr61e)'so:l$not'torn

iury' . --r- of a4 thc' *.rrr:r exl cX'cririrant the - I'rior to the s'Lrirt

rrttr..er1"!;

toore 1>lacetl as tiglrtly

gctlrirr.aslro::si.bleitr,.lrc|1.r1ssa1!er.Atthestctrtingsignaltlreylravetopush incr''cafor;..rrd.torttrrcistheclool.astrardaslrossi.b].eanr].tortrnit\..ayasquiclllyas thcn follo* the circuit <loorr:.ntl ti"ra porsible orrcc tlicry arc rpass&6;c:r 'rr.otr1;h of 1'reople in crttslr thc': re.ioirr teci to

tiic

fel-l or

stttnt-

Iiitlral..l.r]oor.r.rj.tltlrsithi:ppcrn<rclteryfreciucntlY+,hatpeoplcfe].lorstttnt. clocr" -^-.,.in the rlvv" to cattse passirrs 'chrottslt l'Irt^i are likel-y ;t:;:" are q.' tSey they blec' blec. after passirrg'chrottglt thc.t i: :-"".:t;:i:#';:ttill; the't ai;cin'so no e'h"ule'e to 6ct up the cLoor' 'r,he flo'r' tlrrotrgh f'all.ctr i'rottl<i ha've eff'ects <1oi;'r*'i.c' s.clve!.se1.y be'incr the cl'oo' rTit'kr nor:e people to far.r $o'r3b.c1.y 5ot stuck that very f.re,,'.lniJ.y t'uic'J"1;"'so t}ret rt suf'cj:t":::1 cgain ttnable t'o gc:t i'ree 'appc',.,cl *r. irrrr or.{.oot nri:.i*s jnterru-.lteil. Thj.s trstrc.).}y lrappeneil .[:Ire exl.,erimcrrt tv:rs also

in thlt

cg;se

to be discour,lltctrtlrc,dool.tliisl']oc};etlbyt,lrcforinlrtironof:rtt!..rclr.S,j.r.c*irr.5tlg5.rr5}clao-s rrn<I trre expcr'5r're't'C'o c.ee$ lrrost i.rr r:i.tr:r.Liorr to a ~~ ・′ ・げ‐ ncr l'V・ ■ ・ lossibLct::.""::''','l'",J],1;: ':rz:irr.tro*sirr thisr stutl;r trcit'her lt tlle lla,tlra lrar; in しintlea,it T:3S tiutretl,it is fOI.1■ ea,11(ヽ

al｀ l

r tO b

of b].00kピ :[;e3 01う

thiS Stutly t10■ こ 00rS

tO assOrt thC OPPOSi・

tC.

:":,t;t'I""i'l']"'

3tateri(■ lts abOヽ C On tl10 basiS C a■ l P。 3Sib■

a= a rCStl].t Of l)riagi.l13 a)・

inCc1 91 4か。熟'Ot・ ■ i‖ prOsSiOllS ι

中

r■ OI′ G● acity Or Circtit I(Fi3o6)was inStll .l・

it a aoor、 riath Of 105 1:l tllё 1'iCiCnt,30 tllttt 30■

Of P● OP]C Was led tllrOush tlC litlCll]Onょ 10 tr tllご Stror.‖ (:t tly

ci rcuit,(11,Fi36).

∴5 ●1・ (1さ tll t tilt・て｀だヽ〕

t tl10 utll lttiol1 0f tllじュ‖、 l・ ctltlcCa L・

・〕C:lu].t3 0r llleュ

la lゝ ot tll・ヽrllttll t,1 001■

'i til・・ 0ヽ Vtt ill rrC,nt, Or tllc dO`)r ilこ・::Ot t (〕 tiht,rt.,3()tililt, thC ■

(:■

11・

tillt rrtS ill

'1｀

ヽ nfl■ y3Ctl Witll ':tlfi l)0 こ

じ ¨

iCiCtlt acCtlじ。Cy`

/2 1・

:卜 '・

1ヽ

ムt

:I・

I「

,ト

´ 1・

2 :n tllo

レ∂ :bttl'1・

or ulc a"・ r cOu]a bo m.缶 VL・

く ,OntlitiOlli ■ert: ::lttC:ヽ゛〔し1ゝ :

Cri・ 1ヽ :lCi

V′

dOor WOu]d bC・

Cil・

ty,30 tltriし

丁し ■:3.y

Cuit I,lThi Cn

Crush Or l,じ ol'〕・O itil.I「

vc l)OeA Ob:lcrV eti. Lr`〕 l10S ll :ュ 1ly stal)1● こ

vi31la1 01)301｀

1)C ttt:r.

∵ incd fOr a roir]y10ng thlじ・ ‖ ■ 10■

.Ciじ nt i う

`=3 3110rt・

l)O COnoluaCa

i.11 1・ rOlit

i∬ li二 11

rccOras t}1・ t in Cnn cnCttlrtl PAniC "ittFttiOn tl10 'tiOn3 and tllO ri]r、 b10CkC・ d lly arCll i=01TILtiOn.

・ricicllt tO a j.dth Or C086 m was acri.nite].y insu〔 Tllc dOOr ▼ LS Very f・ rotuC・ 1lt alla vot、 rCh l'O rillt.tiCIn W〔 peoPle tO Iだ tS 3・・ fOrtrftS it l)。 33il)].C tO gOt Out Or tllo al with tl10 CrCa.toSt Oi tO b)tli SCOntinuOrte 0250S the C・ Xl10riliOrlt ht:こ・OCOras that the paSSagO rato O,the human Strcam ■ u■ a be lt rolloIIs frOn tl10 rilliゝ 00r wiatll(Fig。 7).It Sh ar runctiOn ofこ illじ oor is a■ OCCtlr― tl口 ,Ough ul。こ LSttrOa Only betveell tll● ,assage Can bO mc〔〕 rith not。こ ,hOWOVer,that thiS ratO Or rcanCe or atLllじ orOus SituatiOns,aSEOCiatOa with a VOrv Stable arcll,SO that l・

(_♪

ceF・ tain

a00r wi.atl15 011C h3s tO reckOn With the posSibility Or a zero hi passage ll there In riじ .6 the zOne i.n t′

OCk。こ by a stab10 arCll・ hi133● ratO,ir tl■ e c100r iS b■ ・iagillι iS illl10atOa by hatじ tlle dOOr b)b■ is a ribk Or l1locl=■ 60 01・ ST

『

ｎ口ｒｌｌ

`)i00r Wiで

ttll ●

毬 Xt」正 Л 1霊I∫ ∬

liS∫

ぶ』点1' 鷲‖ i∫

iber Of FCOP■ e.

・0｀こing

1・

Wiat1l of 10う n thC Cl´

=翠

in r]Ont of thO aoCrs iS alsO reau

ricicnt sPcOa or return f■

cea because Ofi tlle insu[・

0■

,30 that a lliJ10r Fas― e haa been

sage ratO raS E10aSttrea than if O Stlf`ricicnt]y ]arge nulill)Or 01' 1)00P■

Tl10rcfOro,it iS t110■ crowaeti to〔 Ctl10r in frO:1 0r tlle a00re the FlleaSu■ ilヽ g POintS f resPOn,s t tllat tl10 dorrect SFCCa c01ヽ

l。

2 an■

lle

il翼

wil■ have tO be carriCa Ollt with a 13rger ntt」 it a c10。

fl｀

lε

:lt prollab■ e

r a00r Fiatlls of

0.30 m in rig.7.

lltO tllis l)llonOli10nen Can be ctlinca rroF. thO FrOtucnCi・ 8. It i better i.nSiこ ht i・ rccOl'as ●ncl itlaiCated ilt Fig。 ircll f・。rl■ 1lti_on 01)titinea rroln tho ri11■ 。こ iasing iS tllO greater thO Om tlliS that tlle fretitletlcy Or bl・ CS becOine lvi■ l be seon fl・ 1。

al■ Or is tllC aoor wiatll.

nlore st'b]_C,as tlley al・ noteri tllat What

ヽ TC

C100r tlle arc:■ OreOVじ 1・ ,With a narroW01・ :「こ■■Cr ntllibOr Of pcO,le. I t Sl1011■ f・ Ormca by a sl■

asStl:■

じa

tO bo thO ]_ilij.t t1 0rDr

ヽイiatl■

at l.Ollich :10 b]・

Cl lDO

Ocl〔 in3

inci.■ ごS With tl10 break ill tl10 atiOn T〔 13 to bO O>l'CCtO`i C 'Ves in rigs。 7 SO bC COnclutica frOn the cll■ arch rrcqucnCy curVC. It ilay a■ 1ly aCrO,■ n SSa80 rこ tte is VllStll■ a 8 that at a tio(pr Wiatll or O.6 n tllo l｀ ■ i' i't'':ir f i'rt": r":i'crt't' :lli t・ 1 13 vi.*.; oi' f.i,:r \'.':.';,' "". 11;'i1 CrTo iぃ・ ′ ■ ltfi Ol':、 :: Lr":iit't 。r li(ヽ ()Plc

duC tO arcll fO■

1■

1｀

f`〕

1・

:: 1'tt1 :l:r'l'irt;; '.1 j Lh .'rL.l. rilrrlt' "ri'iiilr: iil'cll'::l' tln.r to tit..: r'l;,'tx'tnlc ( ti'l:l ['-i)i'i'rrl:irtl;) ‐ 7¨

■ll(.ll a a)1lniliC art'lL iS fO〕

■Crl

／／た

W urou♪

ぬ

ha3 1'00n brokじ

ll tlp

J口

鋼 ●

i3め

tl〕 1, Cl・

07a Of・

l)(,()1′

1(: itヽ

.o''o'r'

:Ti;:;l "i:il";:""":;;"";,'.u '"

よV

r●

3晩 ■ Cい 1

。こしOr

1)10 t()1)よこ

r'rruヽ ′

0^ヽ

,1,ヽ

つ

:':3 tilroll【

ilユ

lllllb:ι :::llr:rc.1,′ t3

crva dy 、 .^^11″ 、 ′ 、fintl brピ tlk`10Wl1 0r i'l

l1r-ina:::^":I:::::::'r

trrunir;trl'chi't-i:!iiro"''ni'lr1"i';'9trl'ittlicz'tt'at:t:i['t'o':tttrcfi]'rnrccor'('lii'

I'o;l!t$i-n-l'-9i:::l':r':!:r1l!:ii:!rr-"-*9-l'-"r'!ii-ll4-!J''t'r'j Inor.jurtocrtittrlr.:i1gr;1|pitl.i:;orltobr.;t;rotiebet'.,1.;ctrtlrcneasttl.ctlrcsu}ts, tlre plrenc'!rlr'.'''[r' xere ce'ive,r,r;iricrr charircterj'zc

dinr...:nsiorrlcsc

'a'am._:ter.ri

Tl10 ChitractCri"tiC P:lrttiOt(:1｀ N =F101rr l'ate in l)OOPl・

3 :11'C:

e/Sec Or ba■

]3/SCC

01' a porSOn Or l,a].1 a = Intan dう ■Ino tC・ Ot D = aoOr wうこしl1 0r dialitOtOr 01'h01)1)er Out■ 1｀

6 = accelcratiOn dtle tO grav■

giililき μ 五君 :翼 l二藍

liltil。

t_1

ユ

= cOnst

111ぶ視二誕』 t重

:il塁 urlt∬ N

★

livca

= 00nst ■■S中

「

∈￨

ＬＦＩ

l thc rOSu■ tS OF stlldies With StOCl bれ Ln bO CalCu■ atCa fro・〔 0とう Tlle l・ actOr.ノ tC・・ bO rOuna that/7-= ・ 2 ana O。 8 m it wil■ r a00r wiatlls or l。 ana peol)1●・I「 t■ inoこ ,but, uO Or/1・ Wil・ be olヽ :hel｀ Vtt■ 1.う n a SliChtly hi〔 or tL こ001・ Tittl1 01・ aoor are cll■ a as alrCa■ y:nentiOl10■ ,thO rcsultS Of lleasurCillCnt rOr tlliS Tiath of・ idtllS e TFitll tlle ractor/Xa thils Ca■ 110r d.00rツ nOt as reliab10 as rOr sm・ rr● こ tO th, l ba■ lS Can be tranSfじ tea a■ l rcsu]・ tS Or rleasurelnotlts tiitli Ste● coI】 ].e tllrOttこ h aC,Ors. Tlll_lS,it f Passing Or a croTa Qr I、 KPOnt:nti じC times as どrOat at an C・ throtlsh a こ00r i_S thr・ Oar tlle aoor th〔 l a rOundCd Wall ■ 2。 3 ti les faStCr wit〕

at the d00r.

ba■ lふ anこ 1,00Pl e

■ Or thO rOSu].tS rOr steO■ frOm a cOml,ari=つ C― It rol■ a1 0r O.023' tililes the accC]・ C has a ■10Ve:itllt POtCllti‐ tllat a croTC1 0r pcol)■ all inclination Or the lrinan bOdy⊃ rュ tiOn aue to gravity,Thi.S COrrcsPOnis t riCtiOn be in l・ elatiOn tO tllC VOrtical. Tll0 1・。131.ゞ PasSac Ll■ CXtellt・こ O ll■ y tO a Silt〔 ■00r oこビユ be tltiliZ。 ■eanint, POSiti n Or tllc bOay. :ι in■ y on the ■

o・ s

ィ〕

l・

acPonds

sloPCS tOlだ

11｀

こS

thO CXit Pl・

tholl atltOlllati Cally i・こravit〔

、tiOllil].

roctiOn of t工

0:10tcfヽ a llicll ratO Or PASSて

in a lar〔 :Or 。こ

fith thO airocti011 0r action or thc an310 ｀

tlon ill tl10 ai.¨ OIICtitS O〔まCCC]Cl・ tι

rOrcos arla rOcCiVO lartior COlll「

こVCl・

(iiricllii:i;; 5':r 1' 6.u'.;u1.('vc'lrrtitr,l; l'tr !runlir!r'li l''clicl)!, i :;::truri b;' Litu'J.ii.;;. -8-

i .:rs )

.'Ji

r': l-'i

:) lr.y !)r'

-)' :r;;' J''

l" S'

、

t':il:clj! !,{'cii ol)ell ort'!l"'irrtl$o 1. tr]I crit tloor's g):trttl'd 2.,ttrclregiiott}rltrunot;tct,uorgtr.ir.uotrtlrctrtaine'jctti)cl.out.Ji}. J.tlrctllirj.nt,5ct|,1ler.otttcr;;i1.i1t'[[1lru(.t:t.i'blys}<r1,cto'.il.t.ci!'tit,Jci:it. r'ritir ei:lr:!'licrto/ iro]'t3 or b'cnkin6 lrr? 1tl'orr:icioil :;ltou'[cl cloors . l;. A1I ezit

:i,

JocP's

tz'Liu

possitrS'e ' doors sirou'ltt be as lticic as

*.$ti*{'{'g{r*t

Stu0+ntcn Rondrr the offer 'uy the "Delf'tsche stutlents' e:<pcriment lrith 1!O C-'yuar the orSr:nlze to (nerft iituient Union)' use olt)'y ' so thnt the iLlugtrai'utertraL for fil'm's the lltos have rcl.easucl

The sttrtiy nas nacle possible

b1'

tionsor.igirratecrclusivc3yfror.irfi]nsiiotstlrl:enil.ithcrunerii}. '!9.?' e>:perinrcrr't on lotir liCyrterbcr the of iruilurrary |I issneci tt03 have

-:)''

!ls!r:tit: 1) ant ltittr.lit;r o[' 1r''o1,le (2) ;ri'.r;;;i.lti; tlrrorrl;ir lr r.e:iLrict j.orr ir:J il frttt/jLi.rrtt c,{' tlttrli:i.L.y (l'"rl;c'tr5 l'"" "') (VCr.l,i.C;rl lj.rru j.rr (:,.;titr.c ti.rrriir.:;;t,.,(i cr:rt.jt,,itrn <iurt:;:it..7rLlt:rl; on th,: c>:trcme :'!.i'lrt, rrxr:ti.tatrt.t cicn:;i.ty (1r:.nio) ) I':i.3.2. iii:tu;,pJ.c 6{' byj.ri6irr;; at a'r out{'lo'.'.' *tr.jJ'tl of 9Uo Iri6..l . !lorrr(,:l ir(,uidllct at tlir,ir:sr15.1 l:ln ).on;rdiv!sj-otrL: ovrjr tirc $r.:'rti 2'o9 ttt 3.o1gr11! n lt:i.lltra::cr'!}''-',i to iri;l'il)i:!-\ 6J y,C. to 12 i"D' i'i.6.1s. ii::u:nllLe oi'rtt.clr l'orrncci a:j a ct'o!.r.I of l,coplt: afc ltirs:iinS tlrrough a door of 0.8'5 m :rrtit one of 4, .2 llr aidth I'ig.5. Vrrr.:li.tion in speed. of tirr: u:ass I'ltri'rirt5 out ite it f'ttnction of titnc ,'nt- = ti-me of troutr1e("t'ec passirl; tiirougir Tn = time betl'lten tr:ro cl.ytlali-c :il'chco ' tL = tinra beiore C.i'trcrrnic arch j.s bro!:en up ir:ii;.1 .

l\.-

l'iE;.

!.:l.o..r {,1.}.j('1[ (

E:<1,et

i:nent,-rl. la.i,otrt

(t) strort circuit

(t) rirrn carilera (i;cs) (z) tii.t.n' cair,r-'r'a (t:os) Q) !'ilrn cirmei'a (tti.i)

(t'r)

l,ong

"i1'gsrit

(4-) l:eesuring door

(ti) Uevice for sign,',r11i.rrg the bcgi.nning anrl encl of an cjlerirrent . (A) iiuS.pers to assis;t tl.oc,: stuu$ling Zaa1- = 11ill-1.: T:.&tr - rrtair.s: Oilllciil.; = rnoilj.I'ied. ccnstlucticn: 6atrg = passagsj IIof = yat'd; 0arder:bbc = closl:r'oont l'iS.7. Pn:,:uag;g r-at.e of ir cl'o'.';'.,I of 1>,-'oir!.c throu;;h cocrs l.ituribcr of p.:ople Pel' $ecoti,i.vct'stts tlocr tzi.citir (crn) r-isl: of 1'ormi'.tiotr 01' $titb-]-c arcir #+H+ troublefrce pl>.sssgts o lneasured. Pci:tts ilig.8. rrer.guency of 1or.i::atj.orr of clJ'rr,::nic erch $hen ir Cl'oiitl. of people ar'.i J)(rs* sing throui;h the tioors Irretinenc.T 9f 4r.ch i'o:'t.retion (nu,,rl;er ler n'Lntit'r.:) vcrstts cloor riirlth (crn) Fig.9. llynanric at.cir i'orrnaiion i';ht.rrt ir ct'oi?,.i oi ut',olrlc iirc 1::issilrg firroul5h cl<tors: extrscts l'r:o.1 a fi.lnt rr-:corct Lirron;1ir cicors o{' tirc ss$t?'wiJ't.h !'ig.1o. lietc oi' prrsl;age o1' :r crot';ci of L'. t !:'.r .'o]': i i't'c,:'r f,q'1i tt1 boLi.cr::r) lrtrt clii,'i'e]'cr:t $li..t:c,:; oi' i'r.'i1.1 :'' t')rl' ilt : ! t: t' :;it:t1ic I c i i't::.t.1..': :' '.i

l,-,i r'.r

Vcrt.iclr.l. irt't'o.'rli ilrrl ie.lrte tlit'r:.:t ir:tr <'i' !jtt'.:rurr ol' ;rcop-tt, (it/lt"c) vr.']':itlij t'lo(,1' -10-

!. 't

_― ―~=■ ■ r'il

′ ぁ′ 一 ●一 ¨ ● つ ¨ ´ 工 ′■ ・Ｅ ● ｏ ● 〓し一つ・ ′ 〓 ′ し一，〓一。〓 ¨ 〓● Ｃ一Ｃ〓，一Ｃり． ´ 。つ ″ 奮 ¨ ¨ ●。Ｎン一Ｃ ●一 ″ 一′ 〓ヽ ′一 ● メ● ¨ ● ″ ´ 一一。・′ コ０■ｏ。つつ。言ｏ一＞一 ■ ｏ ■ ′ ｏ ■ ン ′ 〓〓 ″ ′ ぃｏ〓一。一 ψ〓〓一 ● ｏＥ■ ´ ｏ ● 二一 ¨ 〓〓一 ■ ■ ●¨ 〓っ ″ 〓り２〓ま〓〓， ● ′り、・一一〓

・００Ｃ ● ＞〓Ｏｏこ。。。３・３Ｃ ● ０いＯＣｏ＞一ｏ３

Ｃ〓ｏ・〓工仁 ● ¨ Ｃタ ¨ ″ ヱ ● 一 ● つｏｒ〓ＣＣつ・ ● 。ｃＯ一ＬｏＬと。０ぃ ● ″ ゛０ｏ一０一一 ■ ｏｒ ′ 。Ｏ ¨ い０Ｃ。Ｃ，り一 ● ■ ´ Ｃ＞一Ｃ〓。Ｃ。 ● Ｅ ● ＣＯ〓Ｏつ。 ● ● 〓ｏ。。＞ぃた ¨ ″ 〓０一 ● ′ 。。 ● っ一一Ｏ〓Ｎ〓Ｊ一 ′ ■ 〓ｏ〓 ● つ〓０ｏｒ ′ 一。 ′ 〓一Ｃ． ● 。一〓＞ ● 〓 ● 壼。。Ｌ〓。ｏｏ ● ョ二〓一 ´ ¨

ｃ ● ＞Ｃ ● ０一 ● ００ｏｏゝ ´ ゛

０００

● ＯＣｏ一

Ｃ ● ＞倒Ｅ０一００こ０一ち０。＞０。０。０こ

Ｌ０００ＣＯ ● Ｃ。Ｅ００

● 一。ＣＯＯＣｔＳｐ ● Ｃヽ０＞

００Ｃ ● ＞ｏ〓。Ｃコ

０一０こ一。一つ

(t.tucd) 9taqlct!P al.{lrc$l

・．メｉＣ・・一Ｃｏメ゛ｏ，一一っ υ ● ｏＣっ ¨〓つ ● Ｃ。。Ｃ´ 一ｏＣ ● ＣＯ。一レ０Ｃ一．０一Ｎっ ″ ｏ一リＣＣ。二ゥ ● 口一０一〓 ● 。 ● 〓 ′ ３一０〓 ′ 一 ′ 一 ● 口り０〓一〓０コ０っＣｏ０ ● ′ 〓 ● ● 一 ● 〓０ＣＯ ¨ 一．一 ● ´ ａ〇 ● 。Ｃ。 ● αＯ ● 一 ● ¨ ´ ０Ｃ ●一Ｆョ〓メン ● ● ″ ′ っつ ′ っ、 ■ 引二 ● ● ∵ υ ■．Ｊ０ ´ 一つ３ ¨ 一

ヽし、．。．。． ● ヽ０一 ● 一 υ 一〇

p!.ylausccot

aql aa^ ao{uoo.

t. oo g t-

ＣＯ ¨ ０ ● ＞

・ａｃ ¨

０〓Ｃ ● ＞

手メ〓０一

〓ＯＦ０メｏ

ＪｏＯ ● Ｃ一一＞ ¨ ρ ００ ● ０ｏ０。 ∽

ｏｏ

一一ｏメ工一ｏ ¨ ｔＬン。

． ● ｏ０一ｏ。Ｅ

いｏ一．ｃＣ ● ．α ■ ０一 ∽ ・ＯＥ

・００り ´

∽ 一〓０．ｃ

Ｏマ．Ｏｃ ● 一Ｆ

・〓０ ● ｏ一

００

０一０こ一ＯＣ ● Ｅｏ。 ¨ い

一ＯｏｏこＥ

・富〓〓０ｏ一Ｎ ● Ｃ゛ＣＣ

・Ｆ３

００ン

ｏつｏ

り０一＞

０〓一０ ¨ ¨ 一 ¨ ｏ ′ ＣＮ一Ｏ。千ＣＥっ ¨ 一Ｅ ● 一。。。０Ｌ α Ｏ

一

ＣＯＣＣ０ア

り００一〓工りり。。っ〓。。Ｎ一ＮＣ

０〇一０ｏハ。ＯＬＣ＞。〓ぃ一０〓ｏ。０ ● 。

Ｃ〓６ｏこ．。劇Ｃ中 ● ．一Ｃ〓Ｌめ〓 ¨ ＞．一め０．０ ● ● Ｃ。Ｌａフｔ一。ｏイ、ｃ。ぅ ● ０Ｏ ● Ｃぅ Φ ０，。 ● ヽ ● こＣ０一。０Ｅ００Ｏ０ ∝ 。め

・ ● ０ＣｏＯＯＣＥＣ ●

Ｃ．一 ● 一 ● ■ 〓っ一 ● ＥＣ一。

ｏｏつ

。Ｃ一０一〓〓〓。一一Ｌつ

一ｏ

● Ｃ〓。一一。０一 ¨ 。た ● Ｎ ′ 〓〓Ｃ・涎ｏ ● 口０ｒ。つヽ ¨ ０〓ｃ〓 ● ・ ′ ＞〓っ一 ■ 一一一．コ，Ｃ〓一゛

ｏＣ一Ｏつ一￡Ｃ ● ｏ０〓〓 ●メ ¨ Ｃ¨ 。。ＣｏＣ二

Ｃ． ¨ 一工Ｃし ■ 〓つ ● 〓 ● 。一Ｎ ■ ■ 。ＦＮ声

● ●一ｏっつ ● ０つ。Ｃつ ● ０一 ′ ｏ３，ｏ宝＞嗜

． ■ 一ｏつｒ

●

Ｃ一一Ｃ ● 口一． ● ｏ

・００Ｃ ● ＞〓０。二。ｏ。〓一ＤＣ ● ０いＯＣ ● ，一ｏ３ ● Ｏｃ ● ＞Ｃ ● ０一ｏ ● ０。０＞一

′ ｔ ● 一つＣ一３０ＯつＣ ′ Ｃ ´ ● 。一Ｅ ● じＣ工Ｅり一．ＥつＣョｃ ● Ｎ０一 ′ ● こＥＣメ ● τ ＥＣ３〓 ● ｏコ〓。〓つ ¨ α ｏ一 ● ¨ Ｏシ０ ■ ● ａ゛゛Ｃ ● ¨ ● ● メ。メ。〓Ｅ０。。０ｏ〓 ■ Ｅ ¨ ェっ。。メ． ● 。。Ｆ ● Ｎｃ ¨

一ｐ・５。

〓 ● ● ′ 一

。。ン ″ ¨ 。ｏと ′ ＣＯＣＣコメ

・〓０ ● ｏ一ｏ３ ● Ｃｏ ● ｃ一ｃＯ ∽ ｇ。Ｅ ● 一

Ｃ

ｏ〓

ＣｏＬｏ

。．一 ■ ０一 ● ｏ０一ｏン。ｒ＞ ● ¨ Ｃ

０ｒ

一

ヽ

００

一〓０

Ｃ ● ００た ″ 。一Ｃ

〓ｏ一ぃ〓ＣｏＯｒ。。。〓メ ¨

冨Ｅ．Ｃ

´ ｏｏン

〓

・・一ぃ二一ｏつ ∞ ヒｏ ´ 。ンｏｏＯＥＣ〓 ● 〓コＯ一ぃＥ〓コｏ

ヽ．、Ｆ菫Ｉ．。Ｌ一 ● 一 ●

つＥ。

ち

=■

1凛凛

I::鷲 11lF:蒙

０Ｃ ● ′ 一０ｏｏｏ〓ａｃｏｏｏｍ．００Ｎ一０ ∽ Ｃい〓０一リン

Ｃ ●

・

ＣＯｒ〓００ｘＯ

ＯＣＯＣ ● ０〓ＣＯ〓 ● ゛。。＞。〓〓

ＥＯマ．ＯｃＯ一Ｆ一〇０一０ｏ６ｏＯｏ＞ｏメ一０〓ＯｏつｏＯ

〓い ● Ｃ

．ＯＣ ● 一ＣＥユ ¨ ． ●

Ｎ ¨ ・Ｌ〓０，

．００Ｃｔフｏ一０コ０ ● Ｄ σ ● ＯＣ一。ＥＯＬ

００ ● α 。こ。ごヽＣ ● ● Ｃ ● ＞。し工００００ ● ｏ

一

Ｏ・・・メ０ｏＣ ¨ ０ ● ０ｏＯ．Ｃ一Ｃつ０ｏ一ｏ ● ０一ａ０工〓ＣＯ一ｏＣ ● ｏ α つｏＯＣＣ暉Ｏ一 ● ．Ｅ。゛。 α ＮＣ一ｏ０一。。Ｃ・ｃ〓Ｃ一一っｏ。一Ｏ００〓０゛二 ● ¨ 〓 ∽ ｃ〓０口０二０ｏ一０〓０一一Ｃ００。 ∽ ¨ ＣＣ。ｏｏ一Ｏ０。一ｏ ∽ ３一。Ｎｏ一００〓０上ｃＸｏ。 ∞ 。

ｉＣＣ ¨ ０工

¨ ａＯＯＬ

０ ● 聖 ¨ ¨ ０ ∽ ＦぃＣ３。Ｏ ■ 。〓５〓一 ¨ ニ

高論、

ＩＣ ● ´ Ｐ

一Ｏｏｏこ

つ０ｏ● 。ｒ０メン０。０〓Ｎ● ゥゥっ０

００００〓ぃｏ ● ３〓

Ｃ ● ＞０こ ● フ

Ｏ ¨

〓Ｃ〓ヒ ¨ υ ´ Ｃ

＞ ∞ Ｃ

００一っニ

● Ｃ０ ¨ ０つＣ０ ● 工一〓メ０一一。Ｎ。。０Ｎっｏ一。０

一〓０一ｏ・ｏ

ＯＣ。一．Ｃ一。一ｏＣ一 ∽ 一。＞００ α ｏ０Ｌ〓。ヒ０＞一コ

∽ 一コぃ。Ｎｏ一つｏＣＣＯ α Ｏｏ， ∽ ● 一。 ● ｃｔ ″ Ｌ ′ ・

・ＣＮ一ＣＣ ● Ｏ゛〓ｏＣｏ〓。一ＮＦぃｏＣ。 ● 一Ｏ一Ｃ。Ｃ。 ∽ ，Ｃ。０メＥ０ ● Ｃ０。一＞。日ｏ０Ｏ． ∽ 。ｏＣ ¨ ０じ。ｏ一０い０Ｃ０Ｌ一Ｃ ● ｏ一事Ｃ ∽ ヒ。ｏｏｏ一ｏＣン〓＞Ｏ〓００Ｐ ● 一っウ一りっつ〓 ′ ¨ ´

００

~ 1.丁す‐~~

¨ ｏ工一一０一ｏ ● α ゛０一。工一ｏＣ ′ α 。。一ｏＬ一０一〓００。＞一〓っ

・一 ψ Ｌ工Ｏ一Ｃ一一 ● 工ン ´ 一 ¨ Ｏｏ ● ｏン一。Ｃｏ，０一００．工０ＣＯ・ｏＯ ● ＣＣ・一Ｃ一 ● 。 ● 〓。Ｃｍ。Ｏ〓コＣＬｏ〓０。ヒ０つｏＣｃ＞ ● 。〓一っっ ■ ヽ一・ｏ

ＬＯＯＣ ● ン一ＯＣ．００。っ．一 ¨ ｏ α ＣＯ ● ＣＯンＣ二

八．二一， ● ¨ 〓二ＣＯＣＯ ¨ ′ Ｃ ● ＯＣＣン鋤Ｏｏ一０つｏ

一０ ′ ≧ ´ っっ一イ

０ｏ・Ｏ０たヱ ′ ¨ ´ ０〓 ● ０ｏ一ヽ０ｏＣ０ｏＯ ¨ ｒＣ ● ′ ＣＣＣＯｏ ´ ，つＣＮＦｒＣ一Ｏ一ｒ ● ＣＣ ● ● 。。一 ´ ′ ゛一０〓 ″ ● Ｃｏ〓一Ｃｃ〓 ∞ 一ｏ０ＬｏつＯＯ〓 ● Ｃ，ウ ● Ｏ ● ¨ ０ ∽ 一ＣＣ０。っＣ〇ｏ一．Ｃ一０一゛ ¨ Ｏ ● ● ● Ｌ〓一 ′ ア ¨ 一ｏ

Ｃ

っりつｏＣ ● 一ｏＣ一〓

ｏＯ ¨ ′ ００つＣョンっ

Ｅ

つりヽ

ｃ

。Ｃつ

コ５Ｅｃ＞ ● ＯＬ２ ●

Ｏｏｏ一〒一．Ｃｏ

ｏ０メ一 ● ● 一Ｃ

つい一ｏｏ一 ¨ Ｃ〓ＣＯＣＯつ。。３。。ン

ｏＯＣＬＦい０一〇Ｏｏア

ＣＯ。Ｃ ● ｏｏ工

。 ● つ。た ∞ ｏ ″ 仁ｏメ〓￠一Ｏ ● 〓ａＣコＣ一ＯＣ０ｏｏ ● 。つメＯｏＣＣＣンっ ● ｏちつい・

′ ― ―ニト‐ ・‐――――― ―

'TTttTttT~~‐ ヨ9,:編 E婁瑾菫 :菫多ぜ言軍0」雇ξ挙三七瑾二重予 _

ｒ ● ンｃｏｃＣ ′ ＣＯ ● 〓。Ｌ ′ 一０〓ＣＣ＞一ｒＣ

パヾおこ

一卜一一・

ｃ〓

・一

０＞

● ｏＥ

一

に一

Ｏつ

０

Ｆ．″ 〓 ¨ Ｌ ● Ｌ目

こつ０００

っつ υ メＬｏ ∽

Ｅ壺ｏ３ｏ０

〓 ● ● ● 〓一０一〓

〓り ● ゛０ ● ３

つ一ｒＣ００一 ∽

¨ 一ＣＯコｏ

「

r・ ―

―

rr‐

¨ ｏＣ一一一〇

■ 、タコｃｒ

こ．二 ● Ｃ一

● ´● 〓 ■ Ｅ ● ．ＮＯい ● ． ¨

一〓 ●一 ¨ っＯＦＣ一ｏ ● ＞０ ● ０〓

″ 〓一ｏ ″ Ｃ ● ０

一っ。 ¨ 。一一〇

′ ＣＯ ¨ ● Ｏｏ０

● っ一ＮりＣく

口 ‐‐口田暉 e "只 t￢

｀り

ｃｏ ● ＣＯＥＯ ∞ ．Ｏｃ ● ＞。一。 ● ０ＯＣＯ ● 。Ｆ、っＣｔＣＯ。。。つ

コ〓 ∽ 〓。一ＣＣａココ〓ｏｏｃ一Ｅｏｏヽコ ● 一Ｃン一 ● 〓ＯＣａ０ ∽ Ｏｃ〓ヒ０ｏ０００

００一Ｃ ● ＥＣ００ｃ ● 一ＣＯ ●

， ′ 〓 ■ ■ ■

′ ′ Ｃり０７００

● 一０］ ´ ＞

・一 ● ¨ ゛＞〓００

¨ Ｃ００Ｌコｏ０

ｔ ● ■ ¨ Ｑｏ ● ト

ｏ〓Ｆバ

コＬ９ｏ＞

一 ¨ 。こ゛Ｃ ●

，ｏＣ

● 〓ｃ。 ● 一ＣＣ０Ｏ ● Ｏ一Ｌ〓ｔＣＸｏ ● ● ０゛ ¨ ● 一００一ト〓一

´ ２ ∽ Ｃ〓０く

ヽ一一．一椰 ● り、ヽ詳夕ｏ一一Ｃｏ ● り誡ｏｒ ● 〓ｏ・ＯＯ耽ＥａｏＥまｏり゛一一一 ♂ ● ■ 一〓〓・一っ一Ｅりｃ軍Ｏｃ、。メＣ。３礎Ｃ ¨ 。 ´ ０誤メ〓一０農ｃコ０ｅコ＞〓０ ≒ ′ ０ ● 。翼〓一ＥＣＯ ¨ Ｎ

ｒ、Ｎ． ¨ ｃ ● ＞

００３

〓ゥコ ′ 一 ′ Ｃ ● ０

Ｃ ●

゛Ｅａ０一Ｃ０ｏ ● 。一ｒ〓面ｃ ● ｏ一メ ● 。Ｎ一Ｘ僣．ｏ ● ¨ ．Ｃ ● ぉｏ、こＣ一 ≧ ． ´ 〓Ｃ〓Ｃ一りつも０Ｃつ一 ´ っっｏ〓０一エＯＮョ二 ´ コ ● ● ０ ´ 一ｏ３一 ● ● ０ ■ 。一一Ｏっりｍ０一０。一Ｃ冨００ ● ● Ｅ ● ¨ ● Ｌ ● ゛ヱＬｏ。ｃ。Ｎ ●

Ｃ０００３一コ

ＣＯ ● ＦＯ

¨ 。こぃ００Ｆ〓ｏヽ ● ＯＣ ● ＞一〇３ ● ｏｃ ● ＞ＣｏＯ一〇〇０。。＞

，０３っ ● 〓ｃりヽ。Ｃ．一一っプｏＣ ● Ｅ ′

Ｌｏ ● ｏ ≧ ′ 一０。ＣＣＯ ● ＣＯＥ

００Ｃ

一コ

りっ。っ。綺

修輝碑輩装荘轟轟一螺

Ｉ一聾一誕］艶異ｃ ¨ 押華難・０

３〓ｃ＞ｃ。０

０一ｏ一０ ● 一一一一

っＯＯっつｏＣ゛〇ｏ¨ つ３一Ｎ。。Ｏ〓一．ＣＣ。コ一。 ● ＬＯ。 ¨ ｏ ¨ 一０一 ■ 一ＣぃＣＥ。〓 ● 。＞ヒｃ゛０ｏＦレｏメ。〓一３，３メリ ● ¨００ ∞● ＞つＬ。ｏＯ。 ■ 一Ｃ２Ｃ。 ● 〓メＮ

誕 ¨ 翼醜瞳一電幹卍妻震

。ついＬ ′ 一０ ′ ¨ ● 一

〓 ● ｃヒ二 ″ 一 ■ ● ０。。ＣＣ

Ｆ

ョ一 ■ ，安〓

ぃい ¨ Ｃ ¨ ｏ一ｏＯＮ

ｏ一一ｏ ● ｒＯン。〓一メ。Ｎ。

一ｏ一〓ｏＣ ¨ ０〓 ● ● ゛ＣＯ

■

。〓一コ

〓

０Ｃ ● 一０〓Ｃ ● ＯＣＯｒｒＯ〓一

コ００

０

ｃｏ。

シｏ一ｏ二一一ｏ゛〓ｏＥ。Ｃ ● も〓

、Ｆぃ０＞ ● 〓り ∽ 一 ¨ 。Ｅ

噴一

Ｌ

夢許

轟ｒ．Ｃ ● ＣＥ．．。メメ０Ｃ ● 汗Ｃメンａ ● Ｅ ¨塾。

ぃ０

０コ

頼．疑

Ｃ ● ´ Ｃ ● Ｃ

・００

０ ■

′ 一Ｃ一ｏ一

ｏ。一。一一。一Ｃ００ ● Ｃｏ ¨ ｏメＯｏＣ０つ

ｃｏ ●

ぎ

炉鍾 ¨ 羹一発欝苺 ≒ 押疑鍾．ギ詢期

勁Ｐ，一円０に〓 ´ い一に〓コに一ｏヽ＞ ● 一旧

っ υ ｒコリつ０一００一０一

ｃＦ ′ 〓ｏＥぢＥ

０。ＣＣ＞Ｃ ● ● ０００

¨ ′ 〓ョ０〓Ｃ ● 一Ｎｏ鋤Ｃく

● Ｃ ● ● を ´ り一っっ ′ ¨ ＞Ｃ〇一

．〓 ● 一〓一０Ｃ ● ｏ鋤鋤レ ′ ● 一Ｃっ ● ¨ 〓 ′ ″ ０一＞

Ｃ ● ´ ］ ● ● 一 α Ｏ ′ ＞Ｃ００

一〓 ● 鋤コＣ〓ＣＣＯコ０

ヽ

． ′ 一 ● ● 一 σ ＞

Ｌ

Ｅ

ｃ ● ＞ｃＯｅｃ ● ＥＯ ∞ ．Ｏｃ ● ＞０一一 ● ＯＬＯ

′ 一

ＣＯ ● 一 Φ ＥじっＯ τ ｃ００

，

一 ■ ・

０００００こＯＥｃ ● ∽ ＬＯＥ

，

Ｃ００Ｃ ●

ｃ００〇一５一 ● こヽ一〇ＯＥＬｏ＞００ｃ ● ２ ● すｏＯｃ ● ，ｃ ● ０３ ● ０ｏ。＞

ｏこ一っ一．Ｃ，

Ｎ ″ 一一っ一〓 ′ 一 ● ３ ● 鋤 ″ 一ｏＯ一 ■ つ。Ｃｏ ● 一一一ｏ ● ′ っ一つ〓Ｃ ● 一Ｎり〓 ′ ヽつ・ ¨ υ ¨ ′ つ勁〓つｒＣ ● ′

Ｅりｏコ

ａＣ

′ 〓 ● まり・ ′ つ一り ¨ Ｃ

ｏキ ′

勁りコ一０つつりｏ０工〓ｏ ■ ０一０一

ＬＯン ● 一一 ● υ ０ ●

Ｆ一 ● ● Ｃ

りまＯＮ

Ｃｏ

りＯ

声ュｏＣＥｃつ ′ ｃ ● Ｅ０ ● Ｅ一ヨ

っ〓 ′ ¨ ｏ鋤．Ｃっ３〓〓０ ¨ ＥｏｏＯ＞Ｃ ■

●，Ｏ ′ Ｎ二ヽり、Ｆ０一ＣＯ一Ｏ ∞● Ｆ〓り０・一Ｏｏうａ一００Ｘ ●まＯ ′ ● 〓一一ｏ ● 〓こＥＯ〓ｏＯＮ００Ｃ． ´ Ｏ一ｏＣＥｍｃ ´ メＯメ ¨ ｏｃコｃ ● コ＞一 ●０ｃ ′ Ｏ０つつｏ一ｏＥ一

．

´ 一

Ｅｃコっつっ一Ｅ゛Ｃ ● ０〓 ″ 〓Ｊ

つ３２ご一０二っ一鋤 ● ，エ

● 一〇０ヒコＯＮ

〓ヨ０〓一 ● 一〓 ● 〓 ′ Ｃｏ

可０一一二 ●一

３ ∞

ｃＯＥｏ ● ∞ｃ鋤 ● Ｃ０２ ■ ¨ ■ つ一 ● ¨ ＯＣ二３一Ｏ一工ＣヽＣ２．〓ｏ ′ 〓ｏ

ｏ〓Ｆｏ ● Ｎ炒Ｃ ● ＥＥ〓 ¨ ｏ ∞ ｃ２メＣ ● 一ｏ一 ● 鋤ｏ ′ ● ■ 工 ¨ ｏ ● ｃｏりっっ、 ´ ′ 、む一００ ¨ り¨ 〓〒つつ０．つＥ〓工〓ｏ一 ● ．ＣＯ ′０ｃ〓 ● 一り一、夕０一０ＯメＮｏ ● ０ ● 摯一０つ〓一￡Ｆ ●〓凛一ｏ￠聟 ″ 〓Ｅ一ゲ● リ一り０ ′ 一．〓ｏ勁ョ工 ● ンＣＣＯ，ＣＣ〓ウ ¨ ● 〓っｏ工３ ● ０

ンコＣっＣ一ピ ′ ｏ ¨ ｏちｒ〓Ｃ．Ｏ一 ● ろ一一つｏ ¨ ■ 働ＣＥ ′ ＣＣン〓り ′ ● 鋤一一 ● ′ 〓

，り

つ

¨ 〓 ″ 〓 τ ● Ｅ豊〓

二りヽ ¨ ｅｃコｏａ ● ．ａｏｃ ● ＞２ｃ〓ｏく

ｏ ■ ¨ ´ ０ ●

夕０ ● 一工０ ¨ ０〓Ｃ ¨ Ｃｏｏ一ＣっｒｔｏＯｘ ●

一ｏ夕 ¨ Ｅ一Ｆこｃン

ＬＣしＣ

ＯＣ ● ＣＯＥＯ００＞Ｃ ● ＞一ｏ３０ぃｏ一ｏ ¨ ￡

・０つ一つ一工

ａｃｒｃ一

′ ¨ 一つ００ＣＣ ● ≧ ′

ｏ

Ｆ

一 ● ３０切

３っＯｏｏＣ一 ● 一〓

● 一 ″ ¨ っ ¨ ヽｏ一 ″ 〓 ″ ¨ ＯｍＣ ¨ ｏＣ ′ ′ ´ ′ り￡一ｏゥンつつＣＣヽ

つ〓Ｃ

ｏ一 ´ Ｃ＞り一０り０ｒ ● 一一ｏ〓囀一Ｃａ ¨ ｘ ● Ｃ＞一０つつ一ｒＣ

ち〓ｃ〓＞〓島ｏ工ち〓 ● ｃン一一３一つョ一 ■ ＞ｏｃこ

． ′ 一Ｃ一０一ヽっ ■ り３、らりりむ ■ ● コｃつっ ● メ ● 鋤 ● Ｆ夕ｃ ● Ｅり ″ ちう〓

唐ｏ ● Ｃ一・〓ｏＣり一０ｒ〓ｔ

．０っっ〓Ｃｔ ● 一゛０

．り ● 一〓〓 ¨ ● 〓Ｃ ● つＣＯｒｒＯこ一、 ″ 笙Ｃンつ〓 υ ′ 一 ¨ ● Ｅ

０一Ｃ ● 一０一＞

″ 〓ＬｏＯＣａＣＯ〓 ¨ ヨｏつ＞ｏつ００工ｏ０しっ０つ

０〓 υ Ｃゥエ ¨ ′ ＣＯ

つＥＯ．ＣＯＯ α 〓ｘ０ ● ｒ ● ０〓つ

ｃＰｏｃ ● Ｅお ¨ Ｅｏヨａｘ３ ● ¨ ．ｃ〓

ａＣｏ一 ● ０つ一 ● 一Ｃりつ ′ Ｃ ● 助 ● メイＣＣＣ０一 ● つ

０一

ａｏ

● ¨ ＯａＣ

０ ´ ｏＮｏＮＥＯ

一 ● ＣＯＣ ● ＥＯ

Ｃ ● ∞ ＣＣ一一りヽ ´ コ一ＯリＣ０ ∽

一Ｃ ●

つつ０〓〓

０ ● ｏ

し ′ ＣＯＣ０

０

Ｏｏュ ¨ 一Ｃ〇一

．ＣＯ〓コ一゛

● ０００〓ョ ¨

・〓ＯＥｃｏ ∞

● 一 ′ ｏＣ ● 一０一ン３０００

´ ● ´ ヒ０一 ● ０コＯＮ〓 ● 一ＣＣａ

〓・〓 ´． υ 一ＮＬｏ、ＯＣＣ´ つ ● Ｏ０．Ｃ， ¨ｏ工〓警 ● ｏＣ工一ぃおｏｔ０〓 ´ ● ｏｏｏ゛ョＣｏ

こ ∽ヽｏＣョ ℃ ● ＮＮ

二コ〓 ● ０Ｏｏ ∞ ０一００〓，゛Ｃ一〇メ一 ´

Ｎｏ〓

０

ｏｏＮ

一Ｃｎ ● 一〓０一Ｎ〓ＯＣｏＣ

● ＮＯＣ一〓０〇一一ｏＮ。Ｃｏ

二。Ｌ。一０ ≧ ¨ 〓ン一ＣｏＳ

Ｃり

ＣＯ

０一Ｃ ●

ｏそ ´ Ｃ ● ＣＣコ

主

０ＸＯ ¨ ００〓０

Ｅ

Ｃ００

・ ●

Ｃ ● Ｃ ● ″ ＬＯ

ｏヽｏ・Ｃり

００

，Ｃ ′ ＣＯ ● Ｃ二 ● 、Ｆ一 ● 一ｏＬＯっＣ

０い ● ● Ｅ ● Ｏｃ ● ＥＯ゛〓っ００Ｃ ● ＞ α ０〇一Ｌｏ＞ ∽ ０一０一一 ● Ｃ ∽ ゆ

り。ＸＯ一０工

０ Φ ＯＣ ● ＞ｏ〓ＯＣコい

Ｃ一 ● ３＞ ■ 一９ ● ３ｏ ● 〓ゅｏ０ＥＣこＥＯ０ α のｃＯ事ｔ一ＣＣ ● 一０Ｃゝｃ一Ｏン ● ● ＯこＣ０ ≧，ｔｏヽＣ３Ｏ ●，００Ｃ０ ● コ ∞ Ｃ＞Ｑ００６一 ¨ ＝＝ ¨ ｃＣコ一・ト

ｏｏ

ｏ一Ｃ〓００一ｏ．ｏＣＯＣＬ一ＯＮ＞

ＣＯｏメ ● 一ＣＣ〓〓こｏン

一りヽ多

:像

、こ ≧ ５０ ●

・０

一卜一 ¨ ・

た０・ ¨

豊

二０ン ● 〓一〓りＮＯ

Ｃ ●

ン ′ コｏ工

´ ３

．゛ ¨ Ｃ ¨ ０一〓

Ｏ一つ００

Ｏ

Ｃ〓〓Ｊ

¨ ｏ一Ｌ

っ〓工。 ● 鋤 ● つＯ一ｏＥＣ゛Ｃｏ〓Ｏ

ン ● ＮＥ・ ● Ｅ ¨ ０Ｃい＞ｏ〓 ● ＥＩ出Ｏ ∞ ｇ α ゛ｏむＥ゛ｃ

．ＣＯＣｃＯ

〓

ｏＯ〓メ

´ ¨ 〓 ● ＞・ ″ Ｏ〓ｏ０Ｃ０一Ｃ０〇こ一Ｃ ● Ｏ０Ｃコ＞ ¨ И

１一０工０

ｒ ● ′ ｃｏＥ

＞〓０

虫こ

ヨ一っりｏＣ ¨ ● 工ＣＣＯ卜 ′ 〓Ｃりつ〓〓り一、一〓 ¨ ＣＯｒＣ

工ｏり〓ｏ一ａＣＯｒｔ一０ ∞ 一ｃ０っＮＣ ● ● ０ ● 一 ′ 一

・二Ｏｏ ● ・．〓Ｌ〓ｏヒ ′ ｏｏ＞＞ ¨ ｏ ∞ なＣ０ ∞ ¨ 〓ｏヨ ¨ ● Ｃ一ｏｏ ● ● 一 ∽ ｏ ● Ｃ ● Ｏｏ ● ≧ 一ｏ ¨ ０ ● りヽｏ一

● ● 一〓０一〇３ｏ〓ｏ゛〓ヒ ● Ｃ、ＯＣ ● ０ｏっＣＣ ● ´ ′ ゛ＣＯ〓

ワ＾． ■ 一ヒＯＬ￠０ｃｏ゛ＣｏＦ ¨ ＥヨンＣＯ０〓 ¨ ３ ● 〓

ｏシ

．Ｃ ● 〓 ● 工ｏ〓ｏ ● Ｃ〓ヒＯ一〇３０

００ンｏこ

ニ〓ｏＯｃ

Ｌ ● ＯＣＯ

一ｏ＞こ ´ ｏ ● ∞ ３〓ｃｏコ ′ 〓ヒ ● Ｃ ● ＾つつう・メ〓〓 ■ りＣ一〓， ′ 〓こ０ ● 〓 ¨ ｏ，

Ｃ一

りＣ０一

Ｃ０＾００Ｃ一〓０ ● 工ＣＣ ● ● Ｃ＞Ｏ ′ ＣＣりり

￡

〓 ● ● ¨ ´ ¨ ＣＯ

．・ＯＣ ¨ ● Ｎ・ ● ００ＣＣ ′ ｒ０レ ● ００つＣｏＣつＯＮｏ ● エ

言つＯｏ

ぺい．０

， ● Ｃ０り０ＬＣヽ０＞０ＣＣ，一Ｏ０ＣこＥ・メ〓０ｏ０００工 ● 〓 ¨ ｏ．８ｏ工０ ′ 一〓〓Ｏ０ヒＣ ● ゛ウｏＣｏ０ｏ＾Ｏ＞Ｏ〓コ０二一Ｃ ● ヽ ● ０ｏｏ０うｏｏＯＣｏＬ一ｃレ ¨ υ りｔ一一〓ｏｏ０ｏ０ｏ，ｃ〓中一

一０工Ｃ＞ｏ〓ＣＯコ σ ＯＬ０ ● ｏ一ｏ ● ′ り０・一 ● ・５０

０

Ｃ

´ ０

´ ＣＯ ¨ ＮＣ

・ＣＯＣ ● ● ＞Ｃ ¨ ０ ● た ´ ０・ ¨ ．００一ｏ〓こ０゛一一〓Ｃヒ ∽ α ＣＯ、Ｃ〓一Ｃコョン。ＯＣ ● Ｌ０ｏ０ＣＱ０

一０一〓

・ｏ ¨

・ｏｍａＯ

ｃ一ＯｏｏｏＬ一ｏ３ＣＯｏＣＯ一Ｃ一ｃ５０一０ ● ．〓Ｃｏ ● ゛ン ■ に一 ● ３Ｃ ● 、つ０ｏ一０一０〓゛ ● ●

ン一囀＞

Ｃ・ａ００ｏヽ一０一 ∽ 一〓つｏｃｏ ● 一一ｏ・ヽ ´ 句 ● Ｃ０Ｏ ¨ Ｌ ● ¨ Ｏ゛０ ● 一 ● Ｌ ● ０一０一００メしＣ ● メｏＣ． ¨ 〓 ● ＞ヒ

・り一ｒ

舗 ≒ 許 ¨ ¨ 一唱踵 ¨ 薔識璽篭聾躍，

● メｏＮ

〓一 ′ ＣＣ

．一

Ｏ

ｏｏメ

Ｌコ ● 一ｏ

● 〓 ● コ〓 ′ ● 一Ｃ ● α 一０お一０メｏち

Ｃ一至０

¨ ｏこ

ＯＣ〓一０ゅＱＯ００Ｎ００Ｃ〇一

。。Ｎ ′ ● メＥ

Ｏ ∝ Ｃ．ｏＯ ● ｃＣ ● ３＞ ● ｇＯＮＯ ● ｃ〓 ● 工ン ¨ ０コ。０一。ｏ。一

００Ｚ

∽ ＯＺ

ロエト

Ｏｏ＞

０００〇０〇一。一ＯＣＣＣＣＸ〓 ● Ｃ０一ｏ０一０〓〓一

０９ｏ ● ｏｏ、コＥＥＥ０ ● ｏｃ０ｏｏ ● ｏＥＥＥＯ〓こ一こＥ ¨ Ｎ０ ●

Ｃ ● ＯＱＯＯ

、ｏ＞０〇一 ●

ＣＯＣ ● ０〇一ＣＣ ●

Ｃ〇一 ∽

ゆ０

Ｃ ● ＣｔＯ α ヽ ● ″ ０こＣ ● ＞

く

ＣｏＯ一０

００＞０Ｏ α 一０こ＾ｍ

，

こ一 ∽ ００

~~~ヽ

/~~~~~丁

、、

′

ヽ、

￨

ヽ

ユ

￨

'■

ｏ０００一 ´ ● た一Ｃい． ¨ Ｃ ´ ｏ ¨ 一ｏ ● ０ ′ ヽ〓ｏ・一〓田

ＣＯＣ ●

０．一嘔

一ＯＣこ〓Ｃ一０一Ｃ ● ＞遭ＹＦＯ ● υ ¨ 、一

０ ● ● ● ´ 一。０ ● ＣＯｏ一 ● 一。Ｎ．Ｃ ¨ ０一メ

● ●

〓メ一 ● Ｏｏ

．Ｅ．一

● 一

〇一一Ｅ

雨 ‐

α ＣＯ ¨ Ｃ一〓０ｏ ¨ Ｃ ● 二ｏｏ＞ ● ｏＯ００

００

００Ｃ ●

３

ｏと ′ 一Ｆ

。０

０

０．０こ ●

Ｃｏ

ｃ・一ｏｏｏ。一３ ● ０ｏ。ｒメ゛一０

，

―

゛〓０ ′ ｃ ● こ

・一Ｏ α ´ ０

０７。 ● つ０つ０一ＯＯｏｒぉ卜コ０一

Ｅｃｃ

〓い一ｏ

¨ ´ ＣＯｒＣＯＣＬ００Ｃ＞Ｌ。 ¨ Ｃ一 ¨ ＯＣ

′ 〓

′ ・

0-‐

‐

_口

０一 ¨ ０．゛

土・

￨;;二

夕Ｃ＞ ● ０ ¨ 一Ｃン０ ¨ ０〒由．一ｏ ≧ ″ ０一ｏ一０ ¨

● 一〓００ ● ゛

・コｏ一一￡ ¨ ０ｏｏ＞ｏ〓コ ″ ¨ ｏ一つ０００コｏ一０一一〇

―

ぽ

〓ヒｏい一ぃｏＣ ¨ ´ Ｃ〓〇０一一ＥＮ一ＥＮ ● ．一ＥぃＣ

ｏ

０ ∞ ．Ｏ

￠〓コ

Ｃ

ｏ工

ンＣｏ ¨ ０一一二

ＣＯＮ。 ¨ ♂ 、

薇

ｃＣ〓い

つｏＮ。二一３ ¨ 。Ｅ

Ｃ。ＦＥＯ α Ｘｏ

聖

Ｏ一ｏ〓０

¨ ＥＥｏＥ

● ３５Ｆ

一ｏこ０，

′ ．８

騨 ―彎

Ｏｏ

メＣ。ァ〓 ′ 。ェ

Ｃｏつつ、〓ＮＥ ● ■ ¨ ● ＯＥ υ 〓Ｆ。Ｃ ● 匈ｏ一０・・ ¨ ¨ ＞

■ ´ Ｃ ● 〓 ● ● 一０ｏた ′ ｏ。Ｌ０一〓Ｃ ● ゛ ● 一一。 ′ ´ ｒり０

Ｃ ● ０

Ｃ ●

０ ● 〓い

一０一〓吾 ¨ ０

ｏＥ

ｏｒ

い ● ● 一ｏＮ ∞ Ｃ

．ＯＦ上こｏコＣ〓〓０。 ■ ● ０

一

。 ●

一 α り一 ′

5ざ

０工 ´ Ｌ

＞

― ―

ｏＯ ¨

―

ｏ ● ３．メｏｏ３ｏ ∞ メ一一ｏ ● Ｎ一ｏＣＣ０〓 ′ 一．Ｃｏコつ ¨ 〓こ＞

０〓

へＣ．０

● ＣＦ゛〓Ｏｏ．ＣＯｎＥＯＣ一 ● 一Ｃ ● ｏＣ一 ● 一ｏ〓

。′

，ｃｏ３コ０二，０ ¨ ″ ヒロ．［Ｅン

′ Ｌ。 ∞

〓 ∞

1」 l

っ α 一０こ・０一０メＣＯＣＮぅ。つｏ一 ● ギ一一く

ＣＣ０〇一ｏメ一コっｏ ¨ ＯＣｏＣＯｍｏ一Ｏ，ｒｏ．０ｏ ● Ｃｏ ■ ● ¨ ′ ｏ ¨

‐

０ｏ３ ● ２ ■ コ τ リメヽ一〓ンｏ

ど ● ３ ● 〓２ｃ２聖＞

ｉＣｏ一ＯａＣＥＣ

ｃｏ〓０ ¨ Ｃ００ｏｏ一一 ● ゥ

００Ｃ ● 一

ｏメ ¨ 一３Ｅ ● 〓２一ｏ卜Ｎ〓ｏＥ

ＯＥｃ主．ｃｏｏ α メィｏ ● 工む一０〓０

Ⅲ 壼￨・ 2・島ぉ虔 ―」七・共

〓ｏ一０

ｏぉ

メョ一ｏ

● 。

ｏＥ

ｒ一ｏｏｒｃ〓・Ｏ。ａＮｘ ● ・Ｃｃ

．ＥＯ０メ ∽ 一０一〓メ０一Ｃ ′ 一 ■ ′ ｏ一

―

＞ｃ一 ¨ ０〇一り〓一 ′ ０＞

・゛００ＣＥＣＯメ０一

゛０〓 α

．一一０

・ｏＯＣ一ＣＯ〓一つつ ● ¨ ∽

∽ ｏ ¨ 一０

● ＯＣ＞ ∞ Ｃ〓〓０一。ＯＣ一メ一一０ｏＣ〓 ● ぃｏＮ

一 ● α ０

‐―

ＯヽＬＯ一

・Ｃ ●

Ｃ一

・ ∞ Ｃ ● ¨ ・０ｏつ ¨ ・ＣＣＣ一 ● ０Ｃ０ＯＣＯＣＯ一ｏＣ工 ● ＣＣ

Ｃ ● ＞。

ｏ ′ ｒ ● 一０工 ● ＥＥ ● ＣＯ

〇一Ｃ一

Ｃ一

り０ ≧ ¨

りヽ ′ 一ｏＣ ∽

つコリｏ

一 ― ―

‐ ・

・

ョｏつｏＯＣ ● ンＣＯ

Ｃ〇一

― ―

ＣＣＯ＞メ ● ● ＞００Ｎ。〓゛一ＣＯ０００￡っり０ ● 〓二一

・― 一

ンＣＯ０ ´ ′ 〓 α ０〓ＣＣ ● ′ ＣＯＥＯ ● ¨ ０・Ｅｏ ″ Ｃｔ０

ヽ

っＣ工０ｏ０ｏ〓Ｃ０Ｏ一ｏ０，一一〓口０一．Ｃ０っつ ´ 一一００￠ｏ一っｏＣ ′ りＣＯＣＣ ● ● 一００っ一ンｏＮつ ′ つっつ一年０一 α 一

″

一 ● つＣＮ．Ｌ。一〓ｏ一ｏ一ｏ〓０一ＮＣｏ工う。〓ｏｏＣＬ ′ Ｃ ● 〓

ｃ聖ンァこｃｏｒ３Ｅ００Ｅ

・

Ｅ一ｒＣＣ

ＯＣ ● Ｅ

￠

ｏ一

００つ０

■ つ

ｃｒ〓ＦＥ ■ ｏ ■ ・メ，ＯＮ

≧ ′ ．ＣＯＣ ● 二ｒ一つもＮ

‐ ― ―

ＣＯ ● ¨ ＯＥ

ｏＣ

． ● ● ● ０〇一＞〓一０工 ¨ ら、ｏ〓

ヽ

っｏつｏつｏ

● ００ ●

〓ｏ

゛一〓 ● ヨン００Ｎ一お

＞ｏｒＣＣ〓 ¨ ● ｃ〓ｏ

′ 一

つ

一０工

・書

Ｃ ● ＞Ｆ一工

一ＯＣ ● 一０一Ｃ〓０一ＮＣ ● ″ ¨

¨ ｏＣ＞ ● ００

Ｃ，Ｗｒ〓Ｃ

‐

３ ¨ ０ＸＯ一０〓ＣＯ〓 ● ＞ｏｏＮ。ＯＣ一〓００一ｏ、．・一 ● ０

０一っ０一ｏ一〓口．Ｃ ● メｏ０ｏＯＣＣＯ ● ≧ ¨ メ＞ ¨ 〓ｏＦ

い０一・ ¨ ・０一

---f ■ っ〓ｃ ● ● Ｅこと

，

・こ ′ ｏ

Ｃ．りｏ ′ ヽＣ ● ´ ＣＯ〓Ｌ

¨ Ｅ，ＮＣｏつｏ〓

ｏ ● ＞ ● う

り

一ｏＣ ′ ａＯｏ

，

。 ■ ．２二ｏヽＪ ● Ｏｃ ● ｏ２ｂ ■ Ｅ一

Ｃ ● 〓ｏＣＣＣｏ〓つ０

〓一コーフ一

′ ■ い・〓・・ ¨ ３セ ´ ● ¨ Ｃ

幾肇妻一メ狐〓一¨ 〓Ｆ・〓・ン鰤中０憂り￡糞誕誕

．００ ′ ヽ ′ 一９

Ｃ ● ′ Ｃ ● ″ 〓りつ〓一ｒ〓

，

〓一つ一０つ〓こり ¨ ＯＮ ● 一Ｅ

⋮

ｏｏ一っ ¨ ０

鍵Ｑ一弓一 ¨ 奨憂塾、ｃ聾聾勇 ■ 〓

０ｃＰ， ■ うつｃ ● 〓ン〓一っ ● ´ 一Ｃ一ａ二Ｃ〓 ′ ｏ一こ ● ● ● 一占３つヽ ′ 二 ■ ＣＯ，りｃ ¨ 一ｃ ● 〓ｃおｏＮ

こ〓 ● 一ろ ● つシっコ一贅〓Ｅこりピ０ ¨ 一つ〓お工Ｆ ● 〓聟つ０Ｅンヽ ´ ｏ ¨ つ ■ こ一っっＥシ３ ″ 〓一 ″ ３一

つ

つつつ

０

Ｃ ● 蒻ｃｏＥ

つ０００ョｏ一０Ｃ

⋮ 一

。 ● Ｃ一工 ″ 一っ ● ●

，一つンウ０っＦ一 ¨ ● ■ 一

Ｏ

一〓 ¨ ｏＣ ′ ウン ′ 〓０ ●

● ¨ ．ａ ′ ｏＬｏ

一一 ● 一一ｏつＯＦヒｏ

∽ ∽ ● Ｅ

０一ＯＥ

つ一っ〓

Ｃ．

″ ¨ ● ＞

ｏＯＦ

Ｃ一

ｏＯＬ

，〓〓め０

´ 〓ｏＥりヽｃ ● ＞

ｏ〓

停

，

０ α ｏＣ〓 ¨ ０

ｏ

， ● ■ Ｔ

・〓メ〓

一 ● メ ● ュっつ ″ 〓

・ ′ 一っ〓 ¨ 一 ″ ． ¨ Ｆ〓

一っ ′ 一〓一工事〓一Ｃ２。っ ● 事〓０一つｏ υ ０ｏ＞０＞Ｃ〓一工っｏ ● Ｆ ≧ 〓 ´ 一 ″ っ ● 〓 ″ つ鋤． ′ Ｃ一ＣＣ〓っＣ〓＞０ウ ● 〓 ′ 〓〓〓 ● ● ″ ● っ ¨ ０〓〓 τ

・Ｃつり〓ｏっ ′ つ ´ つウヨ０〓Ｃつし〓ヽヽっ一一つ一っつ〓レっ ″ ′ つ ¨ ● つ ′ つ鋤一 ″ 一〓っっ０〓一ン一． ● 〓〓工 ″ Ｃ ¨ ● ′ 〓ａ ′ つＣＣ〓ヱＯ ● ■ りン．Ｃ一〓一 ″ Ｃ工りつ〓つつ〓 ¨ ン一 ● つ０つつＬ〓 ● 〓一つ〓 ● 〓ヱ＞っコ ● ンＣ拗０一ｃ二り ● り ¨ ● ∞ つ〓

０ ´ １

´ ′ ＮＩ

ゥｙ

一

〓

α Ｏ・り一つ鋤〓〓一っＣ ´

∞ ニ

・〓 ● ● ウＯｔ ● ′ ′ 一Ｚ ″ ¨ っ卜．工 ´ 〓〓り〓〓卜一０〓 ″ 一 ■ ＞

ｏシ鋤〓〓〓ｏ一 ´ 〓一鋤 ● ´ ′ ０００

，３〇一一ョＣＯＣＣコ〓一 ′ ス一〓Ｃ ● ０一 ′ Ｃｏｒ〓・つ０ ´ 〓Ｏｍ一ｏゝ ″

メ ● Ｃ ■ Ｃ鋤。〓一 ″ 〓 ´ 〓 ¨ ＾ ′ ′ っ〓一つつ多０ ● つ鋤二っ３一Ｃっっつ３す ″ ¨ つ０っい ¨ 詮りｏＮ ■ つｏ ″ ¨ ｏ ´ 〓０りう〓ｏ一〇＞ ′ ．

υ 〓

＾雲・ ¨

● 鋤Ｃｏ

，ｏこ０ ∽ 事ヒＯＣ、つ

● ● ≧ 二Ｃ ● つ０ニアメＣ

．つ一ｏｏ一０ＥｏＣｃＣ ● りつ〓ｏ ● メ＞・

，〓つ ″ Ｃ ● つ〓 ● ． ′ 一〓〓 ■ 一〓ｏ ′ 〓

● ′ 一っ０ ¨ ０ ● ′ 一Ｃ ● 一つりｏ

α つ二一り３一ｏ ● 〓つ・ ′ ａ ● 〓ＣＣ〓〓 ¨ 〓〓Ｊ日

¨ ● ■

´ ＣＯ ●

′ ● ０ ●

０

鋤一 ● ＞っ一 ● 〓Ｃ ● つ勁 ¨ ● 〓・〓ウ ″ ＣＣＣ ● 鋤 ● ● ´ ´ ″ 〓ＣＣ ¨ ａＣ ● ∽

＝〓・ニフ

一

． ′ 一〓ゥン一っ夕 ● 鋤一りつこつ

〓〓つ。一 ● ● ¨ ′ ¨ ＣＯ一０お ●

● ０一０、

０〓０一ＩＣ ● Ｃ＾Ｃ＞，

〓ｏ ′ 〓Ｌｏｃ＾つ

あ二一 ¨ 〓ｃｃ ● ● ｏっ＞

ＥＥＯＥ一い ● 〓０・ｏり ● つＣ〓もゝｏレ

Ｃ ●

ＣＯう０つ

一ヒ

Ｃ ● ＱＯ０〓０一３０一００つ

０い００Ｃ０ ● ，，Ｃ ● ヒ ■ ｏＣ， ● 〓っｏ０３ ● ヽｏｏ ∞

３ ⊃ ０ ●

００一ＣＯＥＣ ● ｏｃＯＦヽ ● ０

一ｏ＞ｒ ● 鋤 ¨ っ〓Ｃ ● ンＣ ●

FOOa

０￡ｏン０こＣＣ ● つＰＣ ● ゅ

・０ｏｏ，りりｏＣ〓０つこ、っ α ¨ っつ〓Ｔコ０ ′ ＣＣい．ＥＣＣ ■ 〓っｏＣっ一Ｃ一り一つ〓ンン ● つＯ一、〓ｏ〓鋤な ■ ● っＬ ● ｏ一Ｃ一 ● ■ 一つｏつメメ〓ｏＣ協〓つＣ一 ′ 一〓ｏ〓 ″ 〓 ″ ● ¨ ヽ一〓つ ′ ｔ〓りっ ″ Ｏ一〓ア ″ 一３〓 ● ´ ．っも〓 ′ Ｃ ´ ¨ つ ● ′ つ〓二 ● っっ〓一００こい ● 。０・一一 ″ ０一 ¨ 一つっｏｏ ´ ０．つっつ璧一ＣＣ〓りつ，フ ● ０コｏ ● ● ｏお０ｏうンＣ〓〓Ｃ ● ン ● ● Ｎ ′ ０ＯＣ〓〓一っ一Ｃ ′ 一 ′ Ｊ ″ 〓〓〓 ● 〓〓つ〓 ′ 〓り０ ● ヽ＞助ｏ〓 ´ ０つ ´ 〓二パ￡Ｅつ． ′ 一一一¨ ゥ一 ″ 、 ● ｏ一〓〓〓 ′ 〓ｏ ■ 〓っ一 ″ Ｃっつ ● 〓〓 ● つつ ¨ Ｃ〓〓＞ ● ¨ ア一ｏ ″ ¨ 、〓っつｏこりレ ● 〓 ● ％一つ ′ し． ′ っヽ ● α Ｃヽメ〓 ● っワＯ〓 ¨ ゥ ● ｏ ● 一勁Ｃ一 ●〓Ｅ０ｏ勁＞〇Ｆ四〓つ一 ¨

u,..r lctugg $qr r0rna11aag6 ug^ Otluinb.trl

ＣＯう０つ００つ

８一Ｅｏこｏ３３聖Ｃｏ● 一０〓０〓ちＣ一 ● ５０ＣＯ＞０＞＞Ｃ，Ｏ一００工ｏＣ一，ｔ〇一ＯｏＯＣ３ ∽ Ｆの α ミ０ ● ０ぶ〓〓゛Ｊ ● ト０

０一Ｃ ● ＥＣ ● ｏｃＯＦ一ＣＯ ●

●

-r :t!rF' -:.-rrF.

.----

口誨竃Ｊ一一颯が一一一 ¨ Ｆ一議融一・．ｃ一〓ス一〓一〓中

。 ● ● ● 卜日０＞

珈

ｄ目

０

０ ● 一ヽ ´

● 一

，

０ ● 一 ● ｏ

ヨ ● つ０

０ ∽

華華騨轟義２

日０＞一ｏ〓０１１日明ｌ ¨ 一Ｑ・一丁４ｏＯＬ

・ ● ●

一ＯＮ

０一 ● ● ● ヽ０３０

０一

／

０ ∞ 〇り

００ ● ¨

o o

０一一 ● ＯＮ

)

( f 33,/N) uoortsuesuau

螢ヨ一６ ● ヽ ≧ ５０ ●

′ゴー′22

SFPE '「

T,"7

II)

Technology Report 79・ 5

A SIMPLE GRAPHICAL PROCEDURE ttO ESTIMATE THE M:NIMUM TIME TO EVACUATE A BU:LDING ＬＦ

D l',￨‐

猷￨

R:CHARD L.FRANCiS,Professor The University of Florida Gainesv‖ :o,Florlda

CANADA INSTIUI亀 ∫OR S.T.

AUじ 301979

0TTAWA lNSTIIUtt CANADlEtt D_E L'￨・ S.T

SOC:ETY OF FIRE PROTECT:ON ENG!NEERS 60 Batterymarch Stroot

Boston,Massachusetts 02110 price S3.50

A SI14PLE GRAPHICAL PROCEDURE 丁O ESTIMATE THE

::∵ ″ ・

1415t″

.1■

MINIMUM TIME 丁0 EVACUAttE A BUILDING R'icharci The

of this

paper

is to address

the question, "1,/hat i s the m'inimum time to evacuate a bujlding?" It is not too uncotnmon, i n bui I di ng evacuations, to f i nd that some evacuation routes are used more than others, in vrhich case it may be clear, i f I ess evacuees used such routes, wh'il e more evacuees used other routes, that the time to evacuate the bujlding could be reduced. In order to determ'ine just how much the time could be reduced, it would be necessary to know the mjnimum buiiding evacuatjon time. (Since the term "building evacuation time" appears often, we

abbreviate'it subsequently as B.E.T.)

of thjs

The

time would prov'ide a benchmark, or standard of comparison, enabling a better evaluation of actual B.E.T.'s. For example, if we knew that the m'in'imum B.E.T. \4as l0 mjnutes,

knowledge

s, Professor ', ty of Fl ori da , -:. GainesviI le, FloriOa',,a,"

Franc j Unl vers i

being designed, the design chosen will ., ' certainly affect the m'inimum B.E.T., and ,,i '

INTRODUCtt10N AND OVERVIEW

The purpose

￨:i

minimum

and observed that an actual trial evacuation took 20 minutes, then we m'ight want

to consider what steps could be taken to reduce the actual B.E.T. 0n the other hand, if the actual B.E.T. was ll minutes, then we might conclude the evacuation was sat'isfactory. Further, when a building

may be

of interest to compare minimum:'.::,

B.E.T.'s for alternative designs undel,,

consideration.

",,,r1

, . '.,1:,,'

';,,'

' '

-a.,

In what follows we develop and illus-. trate a simple graphical model whjch, given specifjed assumptions, permits the computat'ion of minjmum B.E.T.'s, as well i as allocations of evacuees to evacuation routes which

result in minimum

evacua-

tjon time. In order to develop'insight,

:j.

fjrst

consider the s'imp1e l'inear case where the tjme to clear a specified evacuatjon route is just the ratjo of tlte ',' number of evacuees using the route to the -1;r, fl ow capaci ty of the route. Thi s i nsi ght .. permits us to state the general graphical..g: solut'ion procedure. We then iilustrate... :;i, the use of the procedure by considering .1,_ the more realjstic case, based on actual evacuation data of Pauls, where the time to clear a route is a nonlinear function of the number of evacuees us'ing the route.. we

At the end of the paper we summarize general i nsi ghts and concl us jons obta'ined, poi nt out 'l jmi tations of the graphi cal model, and djscuss briefly means of elim-

'inating such I jm'itations by using detailed models.

(b)

time of INSiGHT VIA A LINEAR

MODEL

a bu'ilding has several different evacuation routes, the relat'ive numbers of people usjng each route can affect the B.E.T. As a simple example, suppose a bu i I d'i ng wi th 420 evacuees ha s two d i f ferent evacuatjon routes, I and 2, having flovr capacities of 42 and l8 evacuees per l.lhen

nute respecti ve'ly.

route, then 'it to "clear" that is,

each

'it

210 evacuees use takes 210/42=5 minutes empty, route I, whiIe

takes 210/18:l I .67 m'inutes to cl ear .l.|.67 route 2, resulting in a B.E.T. of m'inutes. However, 1f 294 evacuees use .|26 route I wh'ile use route 2, then it takes 294/42=7=126/18 ntinutes to clear each route, giving a B.E.f. of 7 minutes. fle now verify the intu'itively acceptable conclusjon that using both routes to capacity for the duration of the evacuation activjty yjelds a m'in'imum B.E.T.; 7 minutes in this case. (a) If x., evacuees use route l, then the tjme to clear route 1 is given by likewise, t^(x^) t,(x.) = x^/.|8 t I't' l' l' = x./42'. l' i s the t'ime fsr xZ evacuees to cl ear route 2. These two f unct'ions are 'il I ustrated 'in Figure 1. Provjded that ,l + xz = 420, the larger of t.,(x.t ) and js the B.E.T. t^(x^) t' l' -4-

) denotes the number who can clear route I in a

Now i

evacuees

・一ｒ．一一一一一中一

p.' (2.,

then 2., = pl (21)/42. and so pt(tt) = 422,. Likewise P2(zr) = 182, i s the number of evacuees who can cl ear .: route 2 in a time of zr. (c) if we denote the total number of - メ rぎ鼻 ■ 1 evacuees who can clear the building (using both routes) in a time of z bY P(z), then P(z) = n1(z) * pZ(z) = 422 + l8z = 602. Graphs of pl (tl ), O2Qr), and P(z) appear jn Fjgure 2 (d) If z* denotes the t'ime to clear the building by using both routes, then z* can be computed from 420= P(z*) = 602*: giving z* = 7 m'inutes (e) t^le can see from Figure 2 that 7 minutes is the minimum B.E.T., fon in a t'ime of z, with z less than 7, P(z), the total number of people evacuated in a t'ime of z by using both routes to capaci ty, j s I ess than 420, so that 'it woul d be imposs'ible to evacuate all 420 people in less than 7 minutes. ´ (f) Note from Figure 2 that we can read the allocation of evacuees to routes ` ■ giving a B.E.T. of 7 minutes, with xf = R., (z*) = pl (7) = ?94 evacuees usl ng route l, and x! = pz(z*) = o2Q) = 126 evacuees using route 2. 2.,

鎌 . :11

1■

Several parenthetjcal remarks about

the example seem appropriate. It is the case, of course, that there is a time 1agbetween when a building evacuation starts'

,t.t"li xrf r, for each route j. It can be demon-,;r.$ :i:.rf,i J J strated (l ) (using much the same approach'"iil;

and "hen evacuees actually exit the build―

ing;

it would be necessary to know this

for the example) that the minimum B.E.T.:, js K/R, which can be intepreted as the , :i',,

time lag, and add it to the BoE.丁 .:s com― puted above, to obtain a total BoE.T。

However, fOr our purposes this time lag

t jme

can be om￢ tted, both in this and subse―

people use a single - hypothet'ical - route' ';:,

quent discusslon, as it only changes re―

of flow capacity R. Further, the building;..

sultant times by an additive constanto

will

minimum

time

l K ': ,

if,

of the

for.4ii,t evac;.{,: :

In other , ,'i;.t: '" words, the number of evacuees to use eacl r:i:, route j yielding a minimum building eva-j'.i'" cuation time 'is given by xj = rrK/R. t.-.,.. cause r,Ki R may not be an integerr SOn'r€ :,.'

approach we are developing does not guar―

antee integer allocations of evacuees to For example, if there were 435

evacuees, the value of z★ would become

435/60 = 7.25 minutes, 91ving xl

be evacuated

each route i, the propontion uees using route j is rrlR.

second point to keep in mind is that the

routeso

to evacuate the bui l di ng i f

pl(7.25)=304.5,and xゥ =p2(7.25)= 130。 5, so that some (minOr)rOunding

rounding

to integers may be necessary...,To.

the proportionality djscussionrr-, l :. we state the Route Flow Proportionality lr'i Rule: When route evacuat'ion flows remain,; ' constant (at their capacitjes), the build-,.. ing will be evacuated in mjnimum time if,r the proportion of evacuees using each ・ Υ′ route j is approximateiy the ratio of ther i route j flow capacity to the total of all:｀ the route f I ow capac'iti es. summarjze

error ls lntr￢ nslc to the approach。

It is worth noting that the example i1lustrates what may be considered a mpro― pOrtiOnality result‖ 。 IVe observe that

,1l

the proportion of the 420 evacuees uslng route l is 244/470 = .7 = 42/60, which is

the ratio of the route l flow capacity

(42)to the total flow capacity for both routes (60)。 Li ke、 vise the proportion using route 2 is 126/420 = 。3 = 18/60, the ratio of the route 2 flow capacity to the total flow capacity for both routes.

that the propor輌 onality rule gives an二 11 intuitively agreeaЫ e result:evacueesFFI

This proportionality result is no coinci―

are allocated evenly among all n routes,1111

denceo

so that the proportion using each route llll:_:

■11、

It is interesting to note,when all ,1,I:￨ the route flow capacities are the same, 1111:

For the more general problem, sup―

pose a building has K evacuees and n evac―

is approximately 1/n.For examplqぅ

uation rOutes:

the route flow capacities, and the time

=11 there are two routes, each with a flow ´.:農す capaclty of r, then R = r+r, so the pro‐ メ portion a1located to each route is givep,〔

for xj evacuees to clear route j is

by r/R = r/2r = 1/2.

if rj is the flow capa―

city for route j, R is the total of all

lif.≒

‐5-

THE GRAPHICAL SOLUTION

nol general ize the graphicaf insights obtajned to this point, and state the generai graphical solution procedure. l^le

The evacuation problem

of interest is

follows: there are n evacuation routes, xj people (a nonnegative number to be de-

tlrrn'ined) wjll be evacuated v'ia each route j, and K people will be evacuated in total via all n routes. For each route j we assume that the tjme to evacuate x., peop'le vja route i is given by a knowi'.toute evacuati on time f uncti on, say t. (x. ) . Si nce the bui 1 di ng j s not JJ evacuated unti I al'l the routes are evacuatecl, the t jr're to evacuate tite bu j I ding, say z, is the longest of the route evacuatjon times t., (x., ), ... , tn(xn). The problem

Graphical Solution

say xf, min'imum B.E.T.

...

;' '

'l'::,;,.'i,l :t'' i

)

ts of

t'ime

)

by a dashed I i ne.

(2) In the first

quadrant defined by the axes just constructed, graph each route : evacuat'ion time f uncti on . Cal I thi s graph ;i Graph l. (See Fjgure l.)

, xf;, vi el di ng a

Several assumptions are needed about the route evacuation time functionso

Procedure

0n one side of a p'iece of transiu- . ..G cent paper, graph the axes, representing "U the horizontal axis (whjch has units of ' number of people) by a dotted'ljne, and representing the vertjcal axis (which has

of interest js to compute

:.::1,

the mjnimum B.E.T., say z*, as well as the corresponciing "route al locatjon," that is, the al locat'ion of people to routes

c'ity Assumption (A-3) that if V, js greater than x, , then ,j (rj ) i s greater'- ' than t. (x. ) : 'in other words , each route JJ evacuat j on time f unct'ion i s stri ctly 'i ncreasing. Assumption 3 may be paraphrased as follows: the larger the num- .:', i ber of people using a route, the greater ' :,-,u.. the time to evacuate the route. In order,:i'-ltii for Assumptions 2 and 3 to hold, 'it is t ,t.'1ii:;,i 'ir't suffjcjent that the slope of each route ,': .::;'i evacuation time function always be posi ,ij.i tive. (Figures I and 3 il'lustrate route t .-".i evacuatjon time funct'ions satisfying all . ,:.r-; three assumptions. ) I-Je can now state tne .,. :

PROCEDURE

For (3

each route j we make an lnitializati咀 Assunlption (A-1)that tj(0)= 0, that is,

)

Turn Graph 1 face down.

|,Jh

j'le

keep-

Assumption (A-2)that tj(Xj) iS a cOntinu―

ing Graph I face down, rotate the graph ". so that the vertical axis as seen throughi, the transl ucent paper i s now the dotted ' : line (people), wh'ile the horizontal axis I

ous function.

is now the dashed line(time)。

zero people are evacuated in zero time. For each routO j we make a Continulty Also, we make a Monotoni―

Leave・

￨￨:≒

t''', ,.':

-6‐ :∴

≧

:liF: ず:r.

:二

jll 手

丁ime inや Minules l

tl(xl)

7J ￨

￨

6￢￨

5J ￨

1は￨卜

≒

￨

り =籍

X」

21 1」￨

.t.....

600

■

2m Fi gure

400 Example

GraPh

.---) ftople

People↑ 囲

K=420

Plz)=60z 400■

■

´ ・一一一一．一一一ｒ一一４一一一一一．．・一一彎一ュ． ¨

200

一ヽ

xl=126 ~可

-T

234

￨

gure

--t---' Exampl e

-T-

Graph 2

―

→

7 Time

Minutes

、

一．

Graph

thi s

I in the position jt

poi

again the case where there are two evac∩ uatjon routes, route I has an evacuation fiow capacity of 42 people per m'inute, and route 2 has an evacuation flotr capac'ity of 18 people per minute. We have seen that the number of minutes needed to evacuate 一一 *l people via route 1 is g'iven by ，:i. 一 t.,(x., ) = xl /42, wh'tle the number of mjn- 一漱〓 utes needed to evacuate xZ peop'le via 一卜 route 2 is given by tr(xr) = xr/18. 一一，・ Figure I illustrates Graph I for this 一 example. Figure 2 illustrates Graph 2 for

now has from

on.

Notjce that the graph er each route j evacuat'ion time f uncti on - urhen vi ewed through the paper - gives the number of people, say p.(2.), who can be evacuated JJ via route j in a time of zr. For each route j 'it i s conven'ient to cal I p., (2. )

the people evacuat'ion function for route j. Call Graph I - as viewed through the paper

Graph

(See Figure 2.

)

example, showing the people evacuation functions p., (2., ) and p, (rZ). For route l, since 42 people can be evacuated per m'inute, the number of peop'le who can

this

(4) 0n Graph 2, construct the total (over all routes) of the people evacuat'ion functhjs function by P(z). P(z) is just the total number of people who can be evacuated via all routes in a tjme of z, so that P(z) = p1 (z) + ... + pn(z). tions,

and denote

Using Graph 2, jdentify the point K on the people axis, and use the function P(z ) to f i nd the poi nt z* on the t jme ax'is for which K = P(2"). The time z* is the m'inimum B.E.T.

(5)

Contjnuing to. use Graph 2, for each route j read on the peopl e axi s the number of people, saV xj, for which *j = pj(z*). The route allocation xf, ..., xf; gives the minjmum B.E.T., z*: any other allocation wjll give a bujlding evacuatjon tjme greater than z*. This ends the statement

of the graphical solution As a

cons'ider

o^(z^) '1' a' = lBz^. ( Thus the total

number

-8-

〓 ■ ●

people who can be evacuated via both routes in z minutes is given by P(z) = n1(z) + O2(z) = 422 + l8z = 602. If the bujlding has K = 420 people, then the time z* for which K = P(z*) is computed from 420 = 602*, giving z* = 7 minutes as the minimum B. E. T. Figure 2 i I I ustrates P(z), K, and z*. From F'igure 2 it also follows that xf = pl(z*) = 422* = 294, rvhjle xl = OrQ*) = l8z* = 126, and so the al'locat'ion of 294 people to route l, and 126 people to route 2, yields a minjmum bui 1d'ing evacuatjon time. NONLINEAR ROUTE EVACUAtt10N ttIME FUNC丁 10,lS

procedure.

sinple I inear example,

be evacuated i n 2., mi nutes i s just P1 (2.' ) = 422r. Likewise, for route 2,

ヽ一ヽ一ｔ ´ ¨ ¨ ｏ一

In the examples considered to this point..O

we have assumed each route flow capacity is independent of the number of evacuees

using the route. This assumption led to I inear route evacuatjon time functions, ljke those illuscrated jn Figure 1. However, there is considerable evidence to

jndicate that jn reality this

assumpt'ion may be

To find pl(zl)l:::::￨li

the number of evacuees who can clear l llキ

li,1

jnval'id. In part jcular,

f.,(x.,) = (.206)(wr)(f) JJ

.27

n町

width from the actual vrjdth. Pauls'

equa-

ratjos of x, to

one except'ion, l'ie in the interval between .l00 and 550: cautjon is recommended in using the equatjon when

'ljth

ratios are not in this interval. Let us 'il I ustrate the graphical

(角 )二

sol u-

procedure usi ng Paul s' equation. Suppose a building has two stairwell evacuation routes, I and 2, wjth wjdths of 66 'inches (l .68 meters ) and 44 i nches ( I . I 2 meters ) respecti ve1 y. For routes

tion

and 2 Pauls' equatjon simplifjes to ,-7 ,'l r.,(x., )t(.26.| )xt " and fr(xr) : (..|782)xr'' respectively. Thus if t.' (x., ) js the time to evacuate x5 people vii r6ute i, then

-9-

賓T。

(2日￢

total

the

日

)Tワ

route 2, We haVe p2(Z2)二

clear

The term w, in Pauls' equation is the ef f ecti ve sta'irwel I rvi dth , i n meters , obtajned by subtracting .3 meter of "unused" based on observed

面

Thus the

= (.206y(vrr)'73x.'27

w, which,

‐

:11‐

sumptions l, 2, and 3.

independence

Pauls (3) gives the foliowing emp'irica11y determined equation for the flow, in people per second, tj(xr), when x, evacuees use a stairwell route j:

tion js

we can see that the functions satisfy As‐

number

(

kttξ e,foril

f78222)洗

::1基

people who can.

building in a time of z

'",

given by P(z) : pl(Z)+p2(Z)

1 ● 〓

[(2田 f7丁

十(J782)丁ア乃z爾

If the building

has 800 occupants, we can' determine the minimum B.E.T. by reading..r.;

from Fjgure 4 the value z* for which ':.'.. 800 = P (z* ) . Equi va1 ent1y, we can so]ve ri 800 = P(z*) analytically for z*, obtaining z* : 359 seconds, on about 5 mjnutes and 59 seconds. The model allocates xf| = p,(z*) = 502.23 = 502 evacuees to . | Ir route I .'aand x*, : 297.77 , 298 evacuees ! to route 2. .,. Consider now the use of Pauls' flow equation for a problem with n routgSr , where w. is the effectjve width of each J route j jn meters, and l,l is the total of all n effective widths. For each route evacuees to clear the the time for x; J

● Ｊ

Time in I Seonds;

trur)

t,(x,

)

(XI)= ヽ

.73

.261 X2・

t21X」 =

￨∞ 1

1782

品轟Ю ∵ 輌 Functi ons 一ｉＴｅｍ

400 Fi gure 3. Exampl e of Nonl i near 200

一７３，

４

８００

ロレ)圭 253)z

・．

(。

７一３

ｐ

xち

Ｚ

〓

p2(Z2)=(

=298

′ ｉｌヽ．

２６

Ｚ

・

上７ 1782z2) ３

2001

￨(:)(:)

gure

Exampl

2∞ e of

-1-

3OO z* 40O

Nonl i near Peopl

- t0 -

- - - -1

e Functions

Time

Seconds

route, t.(x..), 'is given JJ

(x., ) = x ./f .(X, ) = (x./w, )'73 /(.206). t.JJJJJJJ

there are K evacuees, then by going through much the same procedure as before we find that the number of evacuees to use eich route i vrhich yields a minimum B.E.T. is given by xj = (w,/W)K. Thus the number of people the model allocates to a route i s directly proport'ional to the effective width of the route. Hence we can state the

二

uees alloCated to routes l and 2 are givenl 7:falleI:/1r: i3:(:Ole:;::ill: l羊

v:l`I

Til〔

ees, the model a1locates xi=(W1/W)K■ 5021: and X, = (W2/W)K

二298

l and 2 respectively.

evacuees to route,i‐

丁 he

minimYtt bui111lf

ingziV]CI:::;:.:IT'3;i.:::inttb首 ::isi:ic.:1::i:

Additional insight can be obtained by cOnl_ sidering the equation for z★ , the minimumll evacuation time, as a function of K。 _Tり ,,: equation implies that z★ increases aslK:￨:

increases, but at a nsublinear:: rate, be.1

ll'idth Proport jonal jty Rule hJhen route flows are determined by Pauls' equation, the bu'ilding vri11 be evacuated jn minimum time jf the proportion of evacuees using each route j is approximately the rat'io of the route i effective vridth to the total of all the route effective widths. Route

cause the exponent .73 is less than one。

.ll

Put another way, the equatiOn demonslrateS

there are"economies of scale"in K9,solチ :: that, for example, the time it takes toi=:: evacuate 800 occupants, 5 mlnutes and 59111■

seconds,is less than twice the time it'す takes to evacuate 400 occupants, 3 mlnute,

and 36 seconds,(keeping in mind the factl that we are computing the building evacu■

Further, the mi n'imum

B. E.

T. i s gi ven

ation time from the moment the first occur, lit「 pant ex'its the bu'i1di ng ) . :ノ

7't

z* = (K/l.t) . ," /(.206) ,

whjch is the time Pauls' flow equation would predict if all K PeoPle used a single - hypothetjcal - stairwell route of effective width W meters. As an 'i I I ustrat i on , cons i der aga'i n the examole with two stairwell evacuation routes, of w jdths I .68m' and I .'12m. , respect'ive'ly. The ef f ecti ve wi dths are

given by wl = 'l .38m. and w, = .82m. ' so that W = 2.2m. The proportions of evac-

CONCLUS IONS

At this point it

seems usefu'l

icitly

two additional rather model assumptions, the first of wel I be more stri ngent than the ''i A-l 2, and 3. expl

tions

Note that the model does not specify which .uuJu.., should use wh'ich routes:' rather it simply specifies a number of'

auuau.a, to use each route. Thus, to app1y the model to a specjfjc building evacuatjon, it must be the case that at least as many evacuees have reasonable access to each route as the model wjll specify. This leads us to state the Reasonable Access Aisumpt'ion (A-4): each evacuee has reasonable access to every evacuation route. This assumption seems acceptable 'if a bu'ilding has fairly central stai rwells, or stairwel'ls whjch are relatively cl ose to one another. However, 'if a bui I d'ing has, for example, 3 wings, with stairwell evacuatjon routes in each vring, it seems unljkely that evacuees in wing vrould have reasonable access to routes in wing 3. In such a case, in order to use the mode1, we might decompose the building evacuat'ion analysis into 3 problems, one for each wing, and consjder each rving to be a "bui1ding. " [.,1e rvould then expect the reasonable access assumption to be val jd, prov'ided evacuees use only routes in the wing which they occupy. 1

Another

implicit

assumptjon we

now

icit is the fol low'ing Route Independence Assumpt'ion (A-5 ) : the t'ime for

make expl

a route to clear

only upon the number of evacuees using the route, and is independent of the number of evacuees usjng other routes. As an exarnple of a situation where the assumption might be violated, cons'ider a building having two sta i rwel I s (evacua t'ion routes 1 and 2 ) whjch open onto a common lobby having depends

only one row of adjacent exit doors. If route 2 evacuees reach the lobby first, then the time for route I evacuees to clear the building may depend not only on the number of evacuees using route l, but a1 so upon whether or not route 2 evacuees are passing through all the exit doors .' when route I evacuees arrive in the lobby. In this case the lobby exit doors might ' effectively constitute a bottleneck. Hopefu11y such bottlenecks wll be precluded by good building design, but it is important to realize that the existence of such a bottleneck which 'is common to more than' , one evacuation route may make jt difficult to apply the graphical solutjon procedure. .,

jdentified two assumptions which restrict the applicability of l^Jhile we have

the model, in another way the model is really more general than is 'indicated by the exarnples we have seen. In part'icu1ar, the example route evacuatjon time functions we have considered have been ejther linear functjons or power functions. However, the model provi des an optima'l al I ocation for any I inear or nonl'inear t jne functions whatsoever which satisfy assumptions A-1, 2, and 3. More details concerning this matter are prov'ided jn reference I , as'

well as the necessary analysis to justify the fact that the graphical solution procedure provi des an a'l I ocat jon of evacuees to routes which evacuates the building in mi nimum

t'ime.

There -12â&#x20AC;?

is an additjonal

useful general

insight

we can obtain from the model,

At the expense of an increase in model

ty,itヽ

柿fact pos■ Ы∈,9,1難

over and above the proportjonal ity princ'ip1e and the econorn'ies of scale princiole identified earl ier. Recal I that once the mjnjmum B.E.T. z* 'is determjned,

COm口 e爛

the correspond'ing number of evacuees to use route i js iust pr(z*), the number of people who can evacuate route j in a time z* . Thus, not only i s the B. E. T . z* , but the evacuat jon time of every route is al so z* . Thr's I eads us to the statement of the Uniformjty Prjnc'iple: suppose the assumptions A-l through 5 are valjd; if a building 'is evacuated 'in mjnimum time, then the times at which the routes clear are all the same, that j s, there i s a un'iformi ty of route clearing times. llote that the uniforrnity principle is easy to mot'ivate: if some routo, sdy route h, is cleared later than other routes, then the reasonable access assumption guarantees that jf less people were to use route h, then

:l:::ielllltiり iall:lel:。

reference 2), to cOnstruct buildingl,y,'む

ず

ation models which identify specittclworヽ三

elevators, and exits,

;:, :::1芯 ili雌

Such models irftr=ギ

a ha11), as well as travel times betWeen various parts of the building of interO,t: ｆｆ

ｋ

ｔ

ｄ

ｅｎｅｗｏｒｃｔａｄｉｐｒｏｖｉｄｅｄａ，ｇｓｆｌＯｗａｌｇｏｒｎｔｗｒｋｌｔｅｔｈｅｒｅｓｕＳｖｏｌｌｅ恥ｅｓＯｕｔｌｏｎ・＞＜。ｉｍｅｈｏｗｅｖａｅｒｔ

ｎｅｉｂｕｂｅａｄｓｅｕｍ詢ｔｅｈｔ。鰤ｔｈｒ・叫ｖａｕａａｓｔ晰ｈｉｔｅｔｈｒｅｍ such people would have reasonable access 四ｉｏｓ to other routes, so that the tinte to ａｓｓｕ clear route h, and thus to evacuate the ｋｏｗｒ build'ing, could be reduced. A more ｔｔｅｎ extens'ive d'iscuss'ion of this uniformity ｅｖｅ pri nci p1e appears 'in reference I Final'ly, jt should be recognized that the model incorporates 'informat'ion about Ｔ the building being evacuated oniy to the い extent needed to determi ne the number of ｈｔｅ routes, n, as weli as the route evacuaｆｏｒ t'ion time f unctions (as i I ustrated by ｈｔｅ the stajrwell widths jn Pauls'equation). -13- 叩 ⋮

ｌ

ｏ

ｔａｎｔｎｅｔ

・ｄｏｂｔａｌｎｅ

ｃｕｅｅｓＣａ

ｎｅｔｗｏｒ

ｆ

ｅｗｏｒｋｐｌａｃｅｓ

ｃ

ｔ

ｏｎｋｒｅｐｒｅｓｅｎｔｌ

ｈ

ｅ

．．ｉ ⋮ ａ・一ｂｅｎ一耐．叫砕 ¨ ｉｄｕ．，ｂ・書噌〓・ ¨ 一 ¨ 一

● ｏｎｄｒｅａａｌｏｔｈｍｓｃａｌ

ｕｇ

ｏ

ｆｎｏいｈｅｎｔｋｏｒｗｍｒａｎｔｈｅｇｔｉｘｓｅｉｉｎｍｍｉｎｇ型ｃｓｓｕ

・ｏ・ハｔ１ｓｃｒｌ ν

ｔｏｔ，ｎｅ一・ｄ．．ｌｈｂｔｔｅｌｎｇｌｎｅｕｌｋｍｏｏｅｓｕｃｈａｎｅｔｗｏｒ

ｈ

ｅ

。ｓ

ｍｐ

ｎ

Ａ

ｎｏｎｅｅ

ｄ

ｔｏ

ｍａ，Ｋｅ

ｌ

ｔｈｒｏｕｇｈ

５，

Ｔｉｎｕ

―

ｒ

REFERENCES

(

)

一一二一 ■ ムず一一

s, R. L. ,

"A Uniformity Principie for Evacuation Route AlIocation," s ubmi tted to Journal of Research of the Nat'iona'l Bureau of Standards,

Franci

1979

(2)

Francis, R.L. and Saunders, P. 8., "EVACNET: Protype Network Optimization Model s for Building Evacuation," to appear as NBSIR 79-.|593, National - Bureau of Standards, Washington, DC 20234

(3)

Pauls, J. 1., "Building Evacuation: Research Findjngs and Recommendations," National Research Council of Canada, Divisjon of Building Research, ' Ottawa, Canada, May 1978, submitted for pub'lication in the book on Behavior and Fires, edited by D. Canter, J. l.liley & Sons, Publishers.

.:.:I

l;r ，

・一一 ■

・′ う・．

- t4 -

０ぃ

●

ABSTRACT i,lhen

bu j I d j

ferent evacuat'ion r0utes ' of people us'ing the routes can affect the

ng has several

the relat'ive numbers

d'i f

time to evacuate the buildjnq. A simpie graphical procedure

presented which determines the number of people

route

the building

is to be evacuated jn

minjmum

use each

time' The

procedure also determjnes the mjnimum build'ing evacuatjon tjme'

一〓一一一一一

瘍

∽ ０ ● ＯＣ ● 一 ∽ 一０一 ∞ コＯ０つ０ｍつＶ一 ● ０Ｃ０ｏ１一 ¨ ● ０２０Ｅ０二Ｏ一一０一０ ● ∽ 工０．一 ∝ く〇 Ш ２ ∽ ． Ш ０ ∞ 一００＞コくＺ ∝ つＯつ

四０中〓０一Ｌ ¨ υ 〓 ¨ Ｌ、一〓〓０一一〓コ一〓 ″ ヽ

肇

ｏ

〓０ ¨ ¨ ● Ｄ υ Ｏ

ご

〓ｏ ¨ 菫

卜一一 ● ＬＯ

ま〓０〓 ● ｏＯ〓イ０ ↑ ３０

０一 υ Ｅ ● Ｌ。Ｊつし ● 〓０一 ∝

，

塞

¨ ● つ

、ぃ ● ¨ ｏ ● ● 一 ¨ ぃ ● ● ＷＪ ● Ｅ、 Ь ● ● 〓ｔ一， ■ ぃぃ ● ● ¨ 一 ● ● ０ ● ¨ ０ ■ 一ｏ ● ¨ 一ｏ一 ● ● Ｅ ■ ■ ｏＬｏ ●

Ｅ ■

∝ ｏＥ〓

■ じ ⊇

Ｆｐ

一ｏ

，

ｏ

く

〓

〓弓

３ｏ

ＬｏＬ ● ｏＬ

ｏ ●

ｏｏ

０一

ｏ

´ ε

¨ Ｅ

〓

ε ■

華

ｏ

■ 〓い

華

ｏＬ ∽ Ｄ工 ¨ ¨ ■

警

驀華

，

ｏ

● 〓 ε

一ヽ工一

■

Ｅョ日

一ｃＦ

し

一一 ● Ｅ

Ｅ

ｃ

ョビ

ｏ一８

ＯＥ

■

＞ ●

一Ｅ一 ■

Ｂ

一Ｅ

２

ｏ一一黒

ｏ

Ｃ咤しＥ

５ ●

Ｆ

●

■

〓

● 葛

０

，

蓑

Ｅ

〓も

５０

´ ． ´

難

ど

〓〓

Ｌ一ｏ

一 ● 一ｒ一 ■ ● く

■ 目 ●

●

警

一 ε ヽｏＦ一Ｃ ´ ￡ｏ

Ｆｃ一 ¨ 讐

こ

し ●

． ● 躊ｏヒ一うＬ０ｘｏ０一 ■ ● ■ ● ● ■ コじ ■ Ｆ

〓一工ｏ ● ０〇一ｏＯ一ぅｏＬ ● ●

¨ 襲中誠劇響詳響

，

Ｏ

萎

套一筆一疵爆一砒華蓑華一詳一馨一畔章幽轟率葦毬

。ｏ一一 ¨

＞ ●

こＬヽ一，一一 ¨ モ８ ≧ バし

● ｏ一一 υ ＤＯＬ一Ｅ一．Ｆ

０＞ｏぃＣ一つ〓ョ ´ 一ぅｏ一＞ｏ工一０ ¨ ● 一〓

， ´ ｏ〓一 ● 一ｏ一。ｏぃｏ〓一ｏ ● ● ● ● Ｏｏ〓 ● ｏ

¨ 口ｏ工

請一メ議藝砕卍お疑曜慶詳盤肇 ¨ 一 ¨ ¨ 一 ¨ 轟〓試〓菫

一一０＞聖〓。

ｅＥ０００

● や ● ● 一０ ● ｏ０Ｅ喜＾

¨ ● 〓僣

目ｏ一一ぅｏ ● ，０ ¨ 目一一 ● Ｅ，

Ｅ一目 ¨ Ｅｏ４一〓

日〓ゅ目 ● ´ りｏ ¨ ＞ ● 一一工一ｏり ● ● ０〇一

● ε 一ｏ一卜Ｏ ● ｏ ● ｏ〓 ¨ Ｅ一 ¨ メ ● ０Ｅ ´ ｏ一〓どＦ

め、 ● ＥＥＥ〓事．暉 ● ● ２〓一ｏっ，ｏｏ ● ＞，ｏ ● 一 ¨ ｏ ● っ一一つ〇０一一 ● 目綸

喜．。メ轟鷲臨一り醜聾 ¨ ¨ 霊姜卜〓繋海 ¨ 為詳ギ一跳一 ¨ ¨ ¨ ¨ Ｆゃ ¨ 轟彙％一諄議議響

一華藝羹轟華難羹

１ ¨ ０〓 ξ ｆく，２ ∞ ＺＥＥｏ．１１暴誰菱裟睦睦 ´ 疑一．葬築薙景一・誕姜静一 ¨ ■ ≡ 距一一軽

し〓〓 ■ ■ ヽｃｔｒｏ ■ ¨ ι ｏ ● Ｌとヒし一〓〓 ● 〓 ● ● 〓Ｅ ● ｔ一 ● 〓ＦＦ〓〓・

ＯＯい

鎌理一 ¨ ¨ 一謹一＾キ妻砕一 ¨ ¨ 一型¨ 一。コ ¨ 一一．十み一 ¨ ¨ ¨ ヽ篭一熟一一語¨ 一 ¨ 鷲吟・凱 ¨ 一蒔Ｎ卍譲一。・ ¨ 賞一一調¨ 一ギ対一劇型諏゛鷺¨ ｒ鷲一 ¨ 一一 ¨ 一鐸一昨 ¨ 議．ｒ嗜一一凱］一一 ¨一一弱ぶ

ｏ ´ 。一。。。

゛工Ｅ一つ〓，

´ ● Ｏｏ ●

〓ｏ ¨ 一ｏＥ。い ■ 翠、Ｅく・Ｎ・一

Ｌ，

ｏ．ヽ〓ｏ一一卜一 ● Ｅり。０一目一０日ｏ一ぬ。

．

蜂

ヽＶ

。 ∽

０

ピ

・ヽ

拳響織蛾

一響押難

喜竜一ュ覆こ選虐

Ｖｌ

くモ

Ｃ￥

← Ｒヽ

要

＋．

とｔ＝

＋

繊

奮ミ

じ ¨ Ｌ＝

華

Ｆがヽこ

やＥ

ｏ

Ｅ選言

＝一

つＥ ● 重ぜ

じヽ導

鷲二響凱 ¨ 疑

一舗獅

聖

爆

一鮮粋墨

葉調一、快一奨萎バ換漱・い ¨ 理 ¨ 議一憂一一蛇 ¨ 一一鴫塾鶏一 ¨ ・一 ¨ 一打試辞．委一 ¨ 誕一

・ヽ卜

含

ｏぃｏ口 ■ 一 ● 。事 ● ● ｏ

＞。

口 ¨ 〓 ● ´ 。二一つぃ。一っ。

塾等量一華摯準一墓一 ¨ 嚢嶽謡難一一轟農棄

口。 ■ ● っｏ ● 卜ｏＯ〓一０ ● 一 ∽ 〓

お

＞。０ ∽ ０一っＯＬＯ工

ミ

．僣

一ぉ２Ｈ ● ● 一一っ。 ● ＞ ● ０一つ０

〓 ● 一 ● ０ ● 工一や。０一．ぅち。。 ● ∞ ＞ ε ● ● 一０ｏ ● 〓ヽ ¨ ¨ ｒ ● 目一一＾〓日 ● 一妥Ｅ。一工 κ ｇ一っ燿。．。工一 ∽ 一溜

．容 ← ご

一、日

ｒｉ

ゞ

← 三

一 ● 〓一。 ■ ● げ。Ｌ。〓

薔Ｅ＝、

ぃ。一ｏｏＬ

一

，

一

，，

。一０う０に０ ● 一一目 ● 〓Ｏｏ督一 ∽

＋

＋ｆ

３

ｏ工 ´ 〓２ ■ 一。 ● 目 ● ● 。〓。ＯＬ一 ∽ 苺ｏ一 ■ ´ げ０増

ｏり一。 ● 一けっｏ。

● 。。一一 ´ ● Ｏ一 ■ 〓 ¨ ● ．０〓，つ。一う０ ● ● ０聖Ｌ００

¨ 。日 ● ´ ● 一。一日ＯＬ

目 ● ｏ ¨ Ｏ ● Ｅ ¨ ∽ ● ぅＣｏ ■ 一一ｏ ● ● 一ヽ。〓一一一．し僣 ´ ●

一ェ一

日っロ

６ヽ２５２ＣＳ

０一 ● ＯＣ〓

、 ´ ● ｏ＞一 ∽ 一、ｏ一っＣＬ，

０一

＝

メ辞一一豪群詳・凛。鷲 ¨ 一識牽麟ＡピゞＦ三．翌ごＡＩ一摯一ギ理ざＡｌ議ｆ・鵜 ¨ 一詳霙・能麟一・ ¨ お．，展亀こ

¨や、ｏ

→ じヽ

うｏＣ＞ ●

＝ｒじ、ヽ

０一ＯＦ〓一０一一＾

＋

ε 聖８ａｒヽミ＝や

＋

ｏ工

ｒＯＬ＝

ョｏ一ｏｏ

Ｈ＋

〓一一のじ

＋

５０ ●

・¨ Ｒ

０し

ヽ晏ヽ〓 ● 嘔Ｓ〓ヽ０ヽ〓ｏ、ｏヽｏ ● 一〇ヽ ●

● ｏ

ヽ

目

，

● 一

，Ｃ ∞

´ Ｃ〓一一ｏぅ ∽ ｏ ● ● ´ つ〓 ●

目ｏ ■ ● ００〓 ● ｏ工一０，〓い

Ｆ饉瞥だ。鈍質辞ギ翼ｒ一義炉誡鎗課鳴 ¨ 義韓辞撃誌勇藤韓醐ザ．案２

騨覇層難姜鮮疑菫難轟癖 ≪ 駆難薄葺一事難中曜獣理警一難警轟。いｒ ¨ ｏ弊轟華〓じｃｏ誕一一自 ¨ ■ 桑嚢鐸¨ ００業〓工０ ¨ ．彙群薬〓，ぉじ

● ｏ一ＬＥｏｘ。で

壌〓一ｏｏ一● 煙葺一難華饉轟難華華嚢還築

二一ｏ嘔２ ■ ■

９ｏ涯一ｏ綱 → ３

日００ ■ ｏ日０ ■ 日一３

８ ● あヽヽ

〓っｏＥも ● 。日卜００ｆｏｆ。〓 ● 日、 ■ 電一も。つ ● 〓む一む。Ｉ ■ ｏ ● ゼロ一ヽ ■ ｏ。ｆ

鍵響華一難薮難嚢驀聾一難椰嘩

● Ｄ，ＯｏｏａｇＯ〓Ｄ一ｏ一． ● 一 ¨

０理 ´ ● Ｉ、ｏ一３

０も目０ ■ 〓。一３＆ヽ３日ヽ ● ０ ● 一遇っ０スＮ一＞ミ

。一一 ● ２。０、嬌ビ一つ ¨ ｏト，ｏ・＞ｏお

→ でＳ゛＝ヽ一一

Ｏい

０もｏゅｆ５，目２ｏｏ ● ｆ一ｏｏＮ．Ｅ ■ ｏ聖ｏ２Ｚ一。バぃ。さ．。．つ２ ● 。 ● ■ ２も一電ちｏ ● こ ε ● ｏ工貧ｏｏ ● 工ｏ０一ｏｏ ■ ∧ ２

〇一

Ｎ ← ０＝一

０卜０目目００ ■ ｏｏ喜一 ■ 一ぜ髯 ● ■ Ｅ

● 日一ｏ〓 ε ● ８一ｏヽｏ， ● Ⅵ ● ０９Ｅｏ０含理０じ ε ● ｏＴｏＬｏ ¨ ● 〓ｏＥ目ぃ工 ● ｏ〓〓一 ● ２一＋ ● 二 ● 一つ ε ● ｏ５ ● ＋ｏげｏ＾〓 ● ｏ ● ¨ じ ´ ２ミざ聖一 ● ´ 場Ｐ＝一ｏ３＞ ● もｏｏＬり ● ｏこじ日．０ｏ ● 〓一〓ｏ含ｏ，・００一 ¨ こｏ〓ｏｏ ● ＾ＥＥ ● Ｅｏ虐Ｅ嗜 ∽ ｏて〓ｏＦ ● Ｅ８２ｏｏ工ｏつ０も〓や２・ ● ０ｏこ．捏〓 ● 一 ″ もｂｏ一・ｏ〓つこ椰 ● 督螢ｏ ● ¨ → ｏ ● ，工一ｏ０

・萎躍ド謳肇．詳罐 ¨ 中一疑 ♯ 覆聾種評中建一叫嘩萎撃詳嵩・蟷．疑一駆郷一 ¨ ¨ ≒ 一 ¨ ギ妻ぁ〓訴２一２島配 ¨ｏずざ２霙詭● Ｅ。 ¨ 。露一増勇一嬰議ゞＩ一疑砕諄評一ｏ疑卜ざ・精・＝ ¨ 疑８角．ｏ▼

二 Ⅵ ゛ｖｌｏ．こ

＝ｓｔ・

ｏｏ〓ｆ

υ も

ｏこ

聖〓Ｌｏ一ｏＥ響

、ミ

り

Ｏ゛〓貧 ● ｏ ● ・ｏ，日一． ●

聖

ｏ ● ２じ ■

ＦＥ，む日 ´ 薔 ● りヽｃ漱聖ｏＣヽ一Ｅミゎｇｏｏｆや目ｏｏ ´ 工２い

虚ｏ ● やｏ９ｏｏ〓

。一、目〓ｏのｏ日ョく ● 、 ● りｏ２ｏ一ぅ ■ ョｏげＬ，ｃ目〓うｏ口＞ｏｏｏ〓 ¨ Ｉｏｅ〓 ´ ｒ ■ ● じ ´ 一゛晏ＥＮ＝ ● ｏＥ当ＣヾＩ一一目 ¨ Ｅｏ涯ｏ一一も〓ョ一 ● 一、ｏ０ ´ Ｅ，０ｏ２げ ● ● っｏ ● Ｌ

，

。Ｌ、ミ

５ｆエｏ．〓Ｌ計 ¨ ｏ一〓ｏ ´ 一旨 ● ¨ １ｏ ¨ 目８ ¨ ｏ ● ё 一０ョｏ８つｏｔ一ｃどｏ ¨ ■ ｅ一ｏ３コ電 ∽ 〓３０ ● ８ｔ ● こむｏぉ〓も事３＆針め

ち目ｏｏ

３ミ

］ ´ ８ｏｏＬ ● ● 。８

一２ ■ ち ● ヽ ● ｏ目ぅｏ遍

，

ミｒｏＮ一ミ０場０日ミふｏ暉選口ヽ ¨ ヽｏ日ミｏ ■ ｆ綱颯〓ｏ場ｏ ´

∽ 一一

〓 ε く上ｏぉ〓

ε 目日

一卜一一〆摯Ｔ

τ Ｏい ● 〓くＬ ∩

，

い０ ● ＥＯＬ ● 一０。一

→ じむ〓ｃＶ ● 〓

´ ● や ● ８一〓ｏ〓増〓〓〓〓〓 ¨ 〓こ・つ３一ぅ、ｏＨ ∞ Ｅヽ，８２。・ｏｏ工，Ｌ一ｏｏむ ● 一 ● ｃ〓一一Ｌｏｏ，ｏむｏ工いｏ一 ´・ｏＥ ■ ¨ 電ｏ ε 〓、っ一いミｏ ● ュ，ヽ響ｏミにｏｆミヽ〓 ´ 聖ミＳ・ｏ〓 ● 一 ● ｏトりヽ、 ● ´ 電め

懺醐曜緊審一灘駆鞭朧酢響慶灘灘懲

，

目 ● Ｅ

葬奎轟翌亜諦藻［灘講羹翼彙ュ一菫曇黎

あ八Ｓ螢一ｇ営ｏ ● し

，

一ざｃＯＦ ¨ ｏＺ一ｏＥヽ聖．ｒｅｏｃＬ τ 〓ぅ〓 ∽ ■

工 ● 〓じ￠色〓ｏ工 ↑

● ● 〓 ∽ ｏ ¨ ５ｏ

″ → ｏ ∽ τ 。二 ■ こつ ■ ｏも ● の、ロニｏ ■ もう ∽ 、ｏ＾〓Ｅあ口ＣＩ ¨ ¨ ｏｕくｃ〓〓一Ｆ ● つコピＺ ¨ ｇ２Ｏ〓Ｌ ¨ ● ｏ ● 一ｏ一 ● ぅ一 ● 一ｏ〓 ● ● ● 〓＞ｃ９〓 ● 〓Ｌ ■ Ｖ ■ ＳＬ一 ● 一一一一一〓゛ ● ヨａ０ ● ■ 一一共っｔとョあｏ ● い ● 〓〓，０ｏい一０ ● ２一、 ● し〓〓 ● 〓ＥＥ。 τ Ｅ ● 一口Ｅビ ■ 一ココ ● Ｅな３，ＥざＥ ε ■ τ ´ ど一〓〓 ε ¨ ● ュ日一 ● ● ● くｏ〓夕 ε ● ｃ

■ ピ〓ｏ

ｏ ■ ぁＥ〓一一Ｌｏｏ ●，〓Ｌｏ ● い ∽ ｏＥ ● 〓。資ュ ∽ ■ ２ ¨ ヽ〓工 ∞ 一〓一綱 ∝ 〓．Ｉ０ｃ〓Ｆ τ 〓〓Ｌ中ｔ ∵ ｏく０一Ｏ目０コ日ＬＪ〓ｏＯ ′ 日Ｌヽし日 ■ ● ● ５０二

ｒＮ

〓ちＥｏなｏ一０Ｑ ¨ ↑ 口２０く＞い ●

ヽ ∽ ざ卜２一あ〓ｏ〓日む３ ■ ● 安 ■ ご〓ｏ ● ●Ｔ〓〓じぅ ● 目めい ■ Ｅ● 増〓Ｊｏも＞８〓ヽ

お

〓ｏ ● ｃ ∞ ｏ

，

¨ ¨ 囃評辞罐．一０う。朔００一一工崚ｏ中計一一一議一〓

一〓 ■ Ｚ ¨ τ ｏ ●

ら卜一Ｌｏつｏ τ Ｏ滲ひ ¨ ．一ひｎ［い卜一 ∽ 餞Ｚ

ュ

∽ お社

のいＺ

工ｏご嗜 ■ ■ つｃ ● ＥＥｏｏｔ

∽ つＶ一ＥＥ ∽ ョ一Ｚ

〓

め

∞ ∽ ∞ じ卜 ● ＥＧ ´ ● Ｚ〓Ｃ〓〓一〓ｏ，一一０ ■ ヽ

〓〓 ¨ ¨ ３日０と

コｏ

一２っ〓しぃ

● ＞ ● ● 卜い・〓一 ■ い Φ ，日歌〓〓Ｏ二〓３ ∞ ｃ一０省 ● ２ ● Ｆ、ｇじ当ｏＺめお饉Ｅｏ５Ｐ〓ｃ ● Ｅ言もうし〓ｏＦ＞〓ｏｏン ∽ ２ちト〓Ｌ．工トｔ一ひ〓 ● ｃ［う一だハＥ ■ ● 〓日蒙 ● ｏＺ０ｙ ∽ Ｐ＞ｏ ● 一一Ｅ

τ 〓 ■ ● Ｅしぃ〓〓コ

お一 ● お

綱聟３ ● ｃｒＦ ■ 〓 ● ごＥ ● ０

〓 ∽ ｏ ∽ ０

∞ ″ ● 〓 ● ¨ ● ぉ３ ●

● ０ぃ。＾ｏ ε ｏいｏＥ ● 〓ｏ ● ● ´ ２〓ＰＥ〓ｏ３ｏ一 ¨ ｏ一

ｏ日０．８日 ′ 一〓 ● ２０ ● 〓 ● く ´ら毬、にご ●

ヽも一〓ヽ

∞ い

ｏく ¨ Ｌご〓Ｅ ∽

● おｇ ´ ｏ ´ ８〓含ゴ一ミ

一Ｌ

６じ

轟

屋一畔艶。嵯 ¨ 一ヽ謡評崎澤義郵配醜 ¨ 躍蟻 ■ 一一義

〓

ｏ一 ■ 〓３ｏＯ

難

一瑯

一攀

¨ Ｅ聖 ´ ●

● ０一一 ¨ ● ｏＬｏ ● ● 一っＯＬ ● Ｅ ■ つ ¨ ＯＣ一．０

轟

難

鐵

曜

攀

藝

０３ ● 一３ ● ｏｃョｏ一目。 ● ヨロ堀 ■ ｏ ● ｏ。む一ｏ ● ｏい目一〓２一 ε ｏｆ

嚇

¨ Ｅ矩ヮＥ ¨ 卜 ´

Ⅵ 一ぃＴ τ Ⅵ ｆｒｉＳざ Ⅵ ｆ・ギ

Ｅ聖漁 ●

。 ´ ２ ● Ｆちもり

一。 ■ 一一ｏこ ● 。。

ざ ■ Ｅ ● メｏＬｏＬ

ｔｆ ● ｏ．一 ● ● Ｌぅや８ｏ。Ｅ ● ｏＯ ´ δ じ ■ ｆ

．３〓・

．ｏりｏ一 ● 一。 δ 。じ颯一。一＝日０一じやｏ゛〓つｏｏ・一ｏ゛ ● 一メ０Ｅ。。つ〓ロ〓ョ０ ■ 一 ■ 一一２ｏ ● ● Ｚ一。目一

一 ∽ 目 ● ● むち ● ｏい三８〓ｏ ● ｏｏ

．聾

● ｏ一〇２ ● ２

む ● 〓Ｅ。一

一８

り０計〓

亀

一日８一一冒電

ε Ｏ暉 ● ● ２

¨磯疑ド塑一］司一 ¨ ¨ 識羅が一轟 ¨ 撃∝ 誠］繋誼 ¨ 一詢一 ¨ 一 ≒ 一詐黎 ¨ 中疑誠 ¨メ一・お評一 ¨ 略ゴ疇一要斡ぎ拠嘘訴いわ一一銅詳メ躇一憂３ｖ萎ぉ曖日建。義郷

〓〓，

Ａ

。 ´ 日 ● ●

、 ■ 〓

●

２８

つ一ｒ ■ 。Ｆぃ

０口一０ ● ８

ま

●

一撃ｉ

ｇヽ

，

ｏ〓ヽ３２一０一ｏ

縄ε

〓、 Ⅵ ｏぃでむ一 ● ・じＮミいちＥ，日 ■ ¨ Ｅ＝じさ。一．一ｒ ∵ ｒ

せいＯ゛ｎ，一卜・Ｆ一

。イヽ一ヽ ´

∽ ¨ Ｇづ

ε ¨ GH ●

G G

Ｉ饉ＩｌＣ可Ｌｌ■ 可 ● ■００１．１１・〓匡Ｏ３，０Ｃヽ０一１〓・Ｕ．Ｌｏｄ１ｍＬｍｄＯ ´ いＥ配Ｏ可．■ 〓ｏ３可 ´ 一一〓〇

飩

ｍＣぉｏｏ編卜

〓 ∽ ｏＮ一Ｃ ∞ 一一＞ｏＯこ〇Ｆ一一〇ｏｏ一こ二０〓 ⊃ 一〓 ∽ ∽ ョ一 ¨ トＣコｏ一〓卜Ｎ一一Ｃｏ一一Ｅ〓Ｎ ●，一＞０ ● 一〓ｏｏ

、 ∽ ０こ

〓 ″ ∽ Ｃｏ〓ｏＯつ筍Ｇ０餞で。だ。ヽ ′

ｏｏ

ハト t/b

¨ Ｐ，Ｉ

■

Ｉ、 ■

■

″

ど一之

二

ヽヽ嗅

υ つ

上・す

ヽもヽ一

▲ ζ ＳＩｏ

いや０ヽ ρ いヽ士Ｓ塁く疑Ｚ α ヽく。ヾと

，１１２

一峯旨嘔 Σ 〓．く

∽ ∞ 〓 ∽ 〓一コ一 ● Ｎ〓一 ∽

一一 ∽ 一．〓０一 ∽ ｏ〓

ｃ場 ∽ にＥＯ ● 聖 ∽ ｃト

∽ 一 ● っ一っ一 ∞ ０〓ｏ〓 ● 〓ｏ

ご〓 ● ＥＥｏ一い０ｃ一Ｅ〓ｏｏ口 ∽ ω

∽ 一〓 ∽ 〓 ∞ 一〓コＬヽつ

．０．０．ＣＯ一Ｃ一〓 ∽ ∞ ≧ ′ ．ＣＯ〓ｃＤＯ０一一一一ｏ ∽ 一Ｃ〇一 ● 〓一〓

〓ち一・３一２、もコ〓ｏｏ一涯〓一 ´ コ涯い ● Ｅく、一

∞ 卜一一

［ ¨ 一一 ¨ ´ 繁口０くロビヽ

一ぶ ¨ 一一辞一一摯一鵠

ｏ〓０コヽｏ・ ∽ 〓 ∞ 場ｏ一一コ一〇０〓〓０

〓属０００

〓

〇一Ｃ０００〇

Ｆ〓

一

〓０ｏ ∽ ｏ一Ｃ ● ＯＣ一に ∞ 一 ∽ 〇〓 ∽ ｒＯ ¨ 一ｃｏＦｏコ一

一〓た ′ 一〓一 ● 〓０ ■ 一つ〓００ ● 一一ンｏ

ヽミ、Ｑミ ■ くいヽヽ日ヾＦ ■ 漁％ヽミさ０ヽヽヽミ・ｔミヽＲヽヽ ◎

・〓一ミぃれぎ黙 ¨ 翁読磯

ま８ポいい総ざ軋

ヾミゞ０ミヽ ■ ヽミヽこ３とヽミヽ，ミヽミミ

―

・・・・・・・・・・．．．．．・・・・・・・・・・・・・・・・・Ｃ ∞ 一 ∽ 。

∽ 卜ＺロトＺ００

め

ト

―

・・・・・・・・・・・ ∽ ｏＯづ ∽ ＯＣ ∽ 口一一〓ぅＯ

Ｌ一０一。〇 ∽ ２一 ∽ 一っ一

＝ υ くＬ口

０ＯＣ ∽ Ｃ ∞ ∽ Φ Ｏ ¨ 〓 ¨ つ〓 ● ０〇一 ● ｏ口ｏ＞ＯＱ口〓・〇ヽ、 ″３０〓８お ∽ 〓房ＯＱｏＯｆ一一ｏ Φヽと一０Ｆョヽヽ Φ∽ おミ〓一ｆ、 ∽ こいｏこ

一議肇毎一轟難節詐銀舞擁難藝一壕

ＺＯ〓じ・ ⊃ ＯＯＬ卜Ｚ

。

ＣＣｏｏＦ ■ ¨ ０Ｏ＞Ｃ０コヽＬ。．一

Ｏ ● ＣＯ〓一 ● 一０〓嬌〓ａ写〓電飩一。一０し螢〓Ｆい。口〓

Ｌ卜ｏＬｏＥい安一

・

一

．〓 ∽ ∽ く

赫

″ 三〓

．０

義

８下０

日

翼

． ∽ 〇一０口

．Ｃ一Ｏ一一Ｎ〓Ｏ一＞づぬ〇一〇要０２．〓ぼ

∽ ∽ ｏ

ｏこト

ヽｏｒｏ ¨ 望ＦｏＯこ一〓ＯＣ

０つ０目。、一一だ〇ｏ一０

Ｃ一一０００

０＞ＯＣ

０ ∽ ０〓ち。 ∽ 〓一ｏｏ一Ｃ０一 ∽ ∽ 〓〓い．い。・

．〇

・・ ∽ ・・，ヽ・０・０・Ｌ・・Ｏ・・Ｌ・Ｌ・く・・・ト・ト・・Ｏ・Ｏ・Ｌ・・Ｌ・Ｏ・・Ｚ・Ｏ・・卜・・Ｏ・Ｓ・・Ｌ・Ｏ・・ ∽ ・・ロ・ ∽ ト ∞ ロ口 Σ 一ｏ一く〓ぼ ● ρ くＬ〓一．一〓Ｃっビ Я ロ〓く・〓ｏ

〓０日日ｏｏ、ｒ

欝

薔

ｏ＞一８〓

難

ｏ一つ舅 ● ｏ ■

＞〇 ∽ 串

ｏＯ〓ま

一一一嘔

．一０一〓０一０。一一ぃｏ一日

∽ ｏ０ｆ〓一ｏ一ｏ一〓〇いＯＱ ∽ ｏ

藁

辞

〇ぃ＞ｏ ∽

・

・０・＞・Ｏ・Ｅ・・・Ｏ・ｃ・〓・・口・一・。・・一 ● ∽ ・・〇・・ ∽ ・Ｏ・ａ・、・一・つ・Ｃ・・・ ∽ ・ ≧ ・ ′ ・ｏ・こ・０・に・・・・・一０・・０・・ ● ．ｏ ∽ 一口コ０ ∽ 一つ０Ｃ〓Ｏ ● ヽ

れ

縦

卜ＱＸ．くｏＯ。 ∽ 卜一 ∽ ∽ ｏ〇ｏ〓〓一 ¨ ● ●

ぼロエいＯＺ＜ ∽ ≧ ′ Ｏ口ＬＯＬＬくビトト００ＬＬＯＲＹ ∽ 口卜 ∽ ンピトロ〓ト一 υ ビくロだトくし〓く０〓０

。・・・・・・・・・一・・・ ● ＣＯ壺０＞〇 ″ 〓

Ｏｃ

０〓付一ｏ〓

ヽ

∽ ヽ多０口ＬＯＬＬ卜くＬくト〓 υ

． ´ く．ヽ ′ ．一ｏ ∽ ．〓Ｏｐ卜

い

●

● ＺくＬＯン０っ卜の〓

Ｃ〓

ト

〓・〓・ ∽ Ｌ ● 〇〓〓 ●" 〇・〓一 ∽ ｏ ∽ づ ●C ″ ・一 ∽ 〓Ｏ ⋮ 〓〇〓ｏいＪ ⋮ ∞ ・０ ⋮ ω ∽ Ｃ ω ０ ⋮ ぼ一 ● ００ ■ 一〇．。ｏ ● ０八一・一〓じ〓〇〓一〓口掟ｃ００〓＞一イ ∽ 一Ｚ。〇ゝｔ＞〓０ ∽ ・ｅｏＯ一ｏ口ｃｏ０〓ち一、ＤｏコＯＥ八＜一〓。 ■ ≧ ｒＥあ凸Ｏ ∽ じ一くｏヽ．〓３ンヨ。 ∽ ．Σ ｃＦぉ０ン〓２Ｅく。３３〓一ｏ、ぢ〓ｏこ一ち０＾一 ● 毬〓３一０〓、〓ｏ、０ｏ一 ¨ ０Ｅｏ ∽ Σ ∽ りも ● ｏつ一〓。〓巳つごＯｏぃｏ８。一ｎ〓こＬ ¨ぢトく ∽一一・．。マ。ｎト ´ ・〓

卜ＯＯＬＬＯ ∽ ＺＯ卜くコ ⊃ ０コくりＬＯ ∽ ＯＯ〓トロ Σ

ざ０ ∽ 一〓ｏｏｏ一ｏ選目ｏＬ巽要蟹・螂 ¨ 聾鋼囃一 ¨ 主 ¨ 一一凝

つ

ヾ

籍

―

き

‐

:_ _1

∞

Nめ

べ

・・・・・・・・ ∽ 〓一Ｑ

ヽ〓 ∽ ＣｏＯＯＣＯ〓

≧ Ｏ

ョ

ｏ

＞ｏ

∽ ｏ〓 ∽ 一。一０〓ＯＣ ∞ 一 ∽ ００〓 ● ０Ｑ００。

∽ 、ＦＯＯにＯｒＥＯ〓一＞一ｏｏＯ ∽ Ｏｏ〇一〓０一属００＞Ｏ０ｒ、ぃｏＯ００００ｏ０ａ〓 ∽

０〇ｏ一ヽ一 ∽ ４ ―

〇

〓

．．あおｏｃｏＥ一一ｏＯｏバ〓 ∽ Ｃｏ ∩ ．一

R R=

> ちヽ、、ヽ、もい．．．

● 〇〓トロ Σ ，一ロエト

０

〓 ● Ｏｏ υ く〇

０ｏｏ

〓 ¨

い。〓 ∽ 〓〓卜〓

。 ∽

場ｏ ●

０〓０ｏ〓一 ∽ ０専．

・

・・ ∽ ヽ

０日Ｌ ● ＯＺＬくＬく ∽ ンく卜３い国Ｏ０ＯくＬ ∽ Ｌ ∽ ＯくＬＺＬＯＯ卜 ∽ Ｏロ Σ トくロ卜ェ ∽ ト卜 ● 〓Ｚ２日Ｊ卜く．〓Ｄ＞ＯＪピくロり ↑ らＬく０〓０

〇いつい一ユ．〓Ｆ．〇．口Ｅ． ∽ 〓．・ｏ．〓一。．０一．一っ ∽ 〓・ＣＯＯ ∽ ∞ ＯｏｒｏＬＣｔ ∽ 一つつ０ ∽ 〓ｒ〓＞Ｏい ′ 〇ＯｃＯ０ＣｒｏＯ ∽ 饉ｔ ≧ 。 ″ 一〇ヽ一０つ ¨ ＯＣｏｏＥｏＣ。ｏＯ０口〓 ■ ヒ α Ｏ０一〓一０ｏ一Ｃｏ〓ＬＯ０Ｅ〓０〓〇 ∽ ｏ一ごＬ。一Ｌ ■，０一。ｏ ∞ ａ ∽ ３協０一ｃ〓０〓Ｏ一０ＦコａｏＯ０〓〓０つ一ｃｏ〓， ∽ 。ｏ００〓 ■ こｏ一０ｏ０Ｏ〇モい，げつＣ一ａコｏｏｏＯＥＥ９Ｅ口ｏ０〓ＯヽＦＦ四。ＬＲ ● ．。０いいい

・・・・・・・・・・・・・・・・・・・・・・・・・・・・・・・・・・・ ● ● ∽ ０一

一 ∽ ∵ に〓Ｃ一一りａ

いつ

い０一

め［

∞ つ

サいＮ

いＮＮ

Ｏ［Ｎ

いいＮ

=墨

口く

．．．． ∽ ０

∽ ＞く ≧ ゐ０く協く０２一卜くり一２つ〓〓０ ● ．Ｘ

い０ ∞ ● 一ヽ

ぃｏ一００

Ｐｒく〓０

RR=

０口Ｌ ∽ ＯＺく

∽ ∽ → ｗ。ａ

ｒ

、シｏ理一

こＣｔ０ｏ一〓 ● 一口ａ〓ＥＬ０００５一一〓 ● ３ｏ猥＾ＩｏＳ〓〓 ∽ υ

Ｚ口 Σ 一口＝Ｊく

〓 ´ ｏ

つ卜 υ コピ ↑ ∽

● ・

．Ｘ

Ｐ

鮮卜ヽく〓０

．。．卜Ｒ一一・

．

０Ｌ

∽ 国ビ ⊃ いく国Ｌ

∽ ヽＦＯ口ＬＯＬＬくビトト００ＬＬＯ ↑ Ｚ日〓ロンＯ〓

∽ ＯＺＯａ ⊃ ∞ ０

・・・・・・・・・・・。・・．．．． ∽ Ｃ ∽ 一ＯＸ〓Ｏっ一ａ０ＣＦ一０ ■ 〇一い一コｏ

∽ ０

口 ⊃

一０

ＬＯ

∽

い ●

い０ ∽ ０と

ついく口Ｌ

Ｅｃ０３一つ、ＯｃＯ ∞ ｏＥ

Ｚ

一一、ｏＬＬ゛〓ご０ｏ菫０ｏハＦロ ¨ ．０． ¨ ママ ‐ で

口〓卜Ｏ卜０ロトくコロＬ

ぃ ∽ ∞ Ｃ〓 ¨ 一つ〓〓ココ００

Ｏ覇口０

一一０

ｏ口〓０にｏ０ υ ロ

．一Ｘ＝ヒ

N■

中ｒＯ〓

∽ ０一 ● ０〇一Ｃ ● Ｆ口一ｏ一 ● 四

ａ ∽

ぎでゼざギ

く〓じ

くばトトＯＯＬＬＯ ∽ ぼロトロ Σ くぼくＬビＯＬロロ ● く卜ＺＯ日 ∽ 四０．Ｘ ● Ｚ口ヽく

⋮ ∽

⋮ ⋮ ．

ロト ∽ ロく Σ 日くＤぼ Σ くピヽＯ０ＬＬＺＬＯく卜にくト卜ト ⊃ ＯＯ，Ｌ０ビ０Ｏ日Ｌく ∽ 卜Ｚ，一口くだ Σ ＯくＯＯ Σ ＺＯつＺＬ，〓〓〓ＸＸ ∽ シＯ ● ＯＺＺＪ口日Ｌヽ一Ｌくく

∽ 田ＯＺロピ四Ｌ目

イ .11:

::11111::￨

０〓〇 ∽ 〓ｏ一０纏ｏ００ ■ ● ∽ ヽ ′ ＯＣＯＥ一一ｏ０・〇一〓０〓０＞Ｏ〓Ｏ

セミヽヽミＯ υ 一＞

墨壺壁墨昼頁汽罠

トト一

ト

つ００Ｑ ∽ ．ヽ

ａ［めヾヾ．Ｏ．．．ｒ．．に．．．．．．一．〇．Ｏ・・・・〇・・〓・０・属・・ｏ・＞一〇・〓・〓・〓・〇ｏ・、日・一・ ¨ ０ ∽ ・・目＞・ｏ・・・０・日・〓・・一・ｏ・ ■ ・・Ｃ・ ∽ ・一・〓・２ｏ ■ ・ ⊆ ・ｏＥ０Ｑ・・＞・〇ＯＣ、・Ｑ・〓・・〇００・ ∽ Ｑ・〓Ｃ・０Ｏ０・一ｔ０・一・一鱈で日，密〓ｏ ∽ ００・卜嘔国．い。

つ一

０ＬＬくにトトＯＯＬＬＯの口 ∽ く０

ＯＬ ∽ ＺＯ卜くごＤＯ日く００ ∽ く口．．＞． ∽ ロ，ゝト０ヽ日

く〓０

卜い寸ヾ．．．だ．．一一．Ｑ．．〇．・ ∽ ・〓・０・一３・・ｏ・ ∽ ・ ∞ ・ ∽ ｃ＞一 ′ ＣＯちＣＯＥ一一ｏＯｏＣ０３ａ一ｏ ∽ ｏ一０一〓ａめ

む

ｏ ∽ 聖Ｏｃコ００覇ａおＯｃｏＥ〓一ｏｏ

卜 ∞

いい。．．．．．．．・・・・・・・・・・・・・ ∽ Ｆ Φ 一 ∽

〓ｏ〓ＦＥ Ю ｏｒ一つ

ＯＺ ● ≧ Ｏ配０ント ∽ Ｚ国０〓０〓こ

・・・・

・・ ¨ ・お〓ｏｏ饉

Ｑ一０饉Ｃ一ｏＣＯ〓ｏｏ〓３ ∞ 一３ヒｏで＞一４

・・

０

。・・

・・

● ｍ。ｃ一 ¨ Ｏ〓Ｆｏ〓

∽ ∽ 〇

．．

Φ ＞ｏ目

Ｑロト０

０

，

ｏｚｏ卜くぼＯＬ．＞ロト餞く〓０

一〓ＯＥ

・・

Ｏ＞０日

・・・

０こ０目 ∽ ＣＯｔａｃ〓３ ∽ ∽ 一ｏ〓ｏ００〓

¨ 〓もおｏｏ〓 ∽ ｏ３ｏ ∽ 一〓ｏＯ ■ ザｏ ∽ 聖

一ｏ〓０門．Ｎ

０口ｏ０

ｏ・０＞３ｏＥｏつ０

ｏぃ

０ｔＥ一〓ｓ

つ０一Ｅ

ｏ七 ∽ ∽ 〓〇ｏ

ど

¨ 一ｏＯ ∞ Ｃ０２０ Σ

N NN

・・・・・・・

∽ Ｅ一ヨｏ一一０

国コ

・・・ｃＯ ∞ 日〓

∽ 疑ＣＥ一一〇 ¨ ¨〓０ ¨ 卜っ・・・ ¨ 〇・・ ∽ ・、・舞＞・許 ′ Φ ・ ∞ ・・計 ∽ ・ ∽ ・・〓Ｑ・ｏ・・一〓・ ¨ 一ヽ〓〓ギ一 ∽ 計．０〓一ａ０つ０．０卜 ″ ａ〓コ一一０ｒ〓キ ∽ 一ｃＸ嘔Ｅ〓ｏ畔 ∽ ０一０〓〇臨 ¨ 〓露一〓口蠣でｏＯ・れ４

一コ〇〓一一卜

ビト ∽ 田ぼＺ

卜い

黎

いｏ〓

．．

一ＣＯｒ

Ｏ

い ∞

ｕｒ

０ｏ

Sドミ

つ

Ｏｕ

Ｏ ∽

，

Ｏ理

Ｏ ∽

Ｏｏ２

く田ビト ∽ ＬＯ

一

ＯＺ目０ ∽ 田＞ Σ く

．＞ピロトしく〓０

≧ 口０く ∽ ∽ く飩

∽ ｏ一Ｑ一ＯＣ一Ｑ一ｏ ● ００寸 α

，

〓一 ∽ 日ｏと ∽ ｏ〓

２ Σ 崎 α

つ α

ト Cヽ

い〇

・Ｏ・ ∽ ・０ ∞

ＬＯ ∽ ゝ一く国

∽ ・ ∽ Ｑｏ工・一 ● 一・

卜 ∽

・

ヽ一一一〓ｏ日ヨＥ

Ｏ卜Ｚ口〓四＞〇

シｏｃ

ｏｏ日〓

′ ロエト

０一 ■・Ｅ鷺一ｏ一Ｅ一ｒ無ｏどｏ一Ｅ一ｘｏＱ色く

口 ∽ く０

． ∞ ヘ

色

∽ Ｃ〓ョ一〇〓０ ●

ビロトＬく〓０

ａ

０

ｒ〇一

口ＯＬｏｍ

つ

。．．．．．．・・・・・・・・・・・・・・・・・〓・つてメ

〓

コく０国 ∽ 二＞

Ｎ

０一ＯＬ一〓＞

Ｚ一

Ｅ

■ Ｏ τ ＞ｏつ〓〓き

Ｎ一

ヽ ′ ｏＦ

．〇．．．．、．．〓．． ∽ ．．暉．．ｏ．・つ・・・・〓・ ∞ ・・・一・〓・・・〓・・〓・・＞・・ ′ ・・ ∽ ・ｍ・・Ｃ・・一・・ ● ・・ｏ・Ｑ・・・Ｏ・・〓・〇・一・・〇一・・り〓 ∽ 一ｏ一 ● 〓Ｃ ● Ｏ ● Ｆ ● ¨ Ｃ０ ● Ｏ＞・・ ∽ 〇 ≧ 〓ごがＯｏＣＯＯ ∽ Ｅ０Ｏ ∽ Ｅ

一一ｏｏＬ

R8=

ｏ一ｏこ一ｏ〓 ε υ

CP 74/74

REPORT ON THE SPREAD OF FRE AT SUMMI二 LAND IN DOUGLAS ON THEISLE OF MAN, 2 AUGUST 1973

A S1lcock,ARIBA and P L Himey,FIFireE

On 6 - 7 August 1973 a team frqa the Fire Researcb Staticn carrled qrt an investigation d the dlsastrcus fire d $rmmedand in which 50 lives were lost. Following the investig'atio a report on the canrses a.nd means of rapid fire spnead, and o the escape problems was produced ln Nove'nber 1973. As an ofEcial equiry was to be held on the ffre, and the matter was therefore strb judice, thie report cculd nd be prblished openly at tqq+ tins

The official report of the Fire Commission has noqr been prblished on l the ieport of the FRfi investigatlng team is now reproduced as a BRE Orrrent Faper. Some phdographs and footnotes have been added as well as an appendix on the fire darm system in use at Srmmerland at the time of the fire, but the report is substatially as it was originally written.

Building Regearch Establisbment

Fire Research Statiou Boreharnwood

Hentfordshire WD6 2BL

■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ Ｆ壁撥騰聰圏 ■ ■ Ｆ駒熙園副留 ■ 闘圏曰ＩＩＩ＝ｌ，＝＝１国圏日 ■ ■

可可口ｄ可ＣＣ可可Ｊ０一．月醸Ｅ ● ＤＯくコ一〓コヽ口， ■ Ｌ ■ 司 υ 一 ″ 匡 ■ ０４口 ● 一〓３Ｄ ■ 哺 ´ 一ｄ一 ■ 〇 ↓

い一 ■ ンロ ′ ヽ

一〇０ ∞ 崎・トロ匡い

￨￨15 醤 i圧 ‐ X● Ψ ■.,

iO~2:￨12

kli10r■ .轟 IZ″ 望‐

い０一の一っ一一ちＬ

しり ¨ ´ 二 ψ ．ｒ一．， ′ ・いご

。０，。ＣＣＯ ● ００ ∽

バヽ

ヽミ二、む゛

ｒ６ｐり５い ● 一７Ｌト

■ 一い、一Ｔ ≦

く．

く＞２く＞ｏ α Ｅ ″ ｏ２つ口ｏ＞いｏ′ ｚロくロトェＯＮ α 。Ｌ餞∽ ０ ≧ ＞布〇

．〓卜．． ∽ ∝ 〇

ばヽ卜ヽ、 (り

］卜つ卜一卜 ∽ 〓一

う

(イン

ー、

ヽさ

・．、ｅヽいヽミ、ヽヽ０

く〓くトトｏ

ＸＯ〓０い０口 ↑ ∽ ．ど、こ４．０餞

．卜． ∽ ．一＾ヨ］ ∩ Ｚ］一く〓くのコ一〓 ∽ 〓一

、ゃＳいヽ ● ● ０ ● ● 属ｏ ■ ● 一 ● ■ ● ■ ト

一ｔミセヽい、ヽ

ヽもいヽ

錠へ、螢もヽ、、、ミ，Ｃヽ゛哺ヽミｓヽヽド、ヽミ０い

″ ０づ目 ■ ０＾ ● Ｏｏ夕ｏの，〓ｏに０ ● ０ ● ｏ一。っｍ一一目Ｏ

贅Ｆｏ３ ■ ● レＱＯＨ口ｏ〓 ∽ ｏ一ｏ・一 ∽ づｏ〓づ υ ∽ 〓〓０Ｃ ● つ餞

罐弔＾銭翼邦主 ¨ 甲督 ¨ 一炉醍憂指

＞Ｘセ、ミヽヽ９

６０ト卜卜トＮ一 ∝

．．

∞ ∞ Ｎ

ＺＯＨ卜くつ０く＞口一ンＯＺ口０ぼ口〓口

哺ヽ０いヽＬＯｔと、０卜

口。〓にＮ一

ミにゝｂミりミヽゞ

・・．・も〓 ∽ ・０ ∽ ５０・０・０ ∽ 鮎 ∽ Ｃ目だＯｏ〓０一ＯＯ τ ＣＣＥ。Ｆ一Ｆ口自Ｏｏ ∞ 守ｏ． ∞ ．配 ∞ ．寺

守 ∞ Ｎ

Ｏ ∞ Ｎ

崎 ∞ Ｎ

むミミ彗べ＞ｘ

∽ めＣ一づ

中０一

００Ｎ

卜のＮ

〇〇

∞ 〇

守

∞

い

・・・・・・・3・

....・

.・

・―・ ―・ ξ・

0 Z

・

・さ

∽ ∩ 〇ロトロヽ４

ロト餞く

餞卜餞く

∩ Ｚく

● 「

∽ 口０く ∽ ∽ くヽ

・．．・．．・．． ∽ ・ ● ● ● ０一一・一Ｏ〓うつ・０＞・ｏつｏｏつ・＞ｏ ′ Ｏ・によ一りのｏ〓 ● ● ０Ｆ．

ＺＯ嵐卜くっ０く＞口

∽ ロロく ≧ ″ 卜Ｚく卜 ∽ あ

∽ ０ ● 一〇５゛ ∽ ¨ ＥＯ

∽ Ｚ国口 α Ｏ ∽

幽

づづ ● ｏ υ ● ＱＱ ∽ ∽ ０ｏ

国卜餞く目０

ｏ

口ば

．

∩ Ｚく

∽ 。 ∽ づづＣ ●

Ｏ

．．．ＯＯ〓一 ∽ 一 ∽ Ｏｏ ¨

Ｅ ∽ ＯＱ一 ∽ 、 ω ト ∩ 鋼．

∽ ・０・ ● ¨ 一゛・・一 ■ 〇５一・・ ∽ 一Ｅ ∽ Ｏ・・ＣＯ・・卜ｔ一 ● 〓ヨ一・ＣｏＱ一 ∽ ｏめ〇 ∽ ∽ Ｃ一・寺もＮ・．０・ＮＮＮ

― ∼

ｄｏ国ヽヽ ○ 卜 ∽ 口 κ 日

＾

で

・．・．・ ∽ Ｃ００・０ ∽ ・ ■ 〓 ∽ ０口 ∽ ００Ｏ． ∽ 守

．． ∽ Ｃ υ ● ００一 ● ′ ″ Ｎ．寺

鬱 ● ヽ．ぴｏ〓Ｑ ∽ ．ｏｏ〓．ｒ一ｏ口〓 ∽ ｃ〓０饉一ｏ負Ｃ＞０ｏ ρ ０ヽ００．崎．ロ崎

め

"寺

留

=C aヽの 0

∽ く口

餞く〓０

． ∽ Ｃ０３ｏ３

０ ■

∽ Ｃ００

０ ■ Ｃ

つｏＣＯ一一

ロト

・・・・・・

・．． ∽

ＱづＣ ● ∽ ＣＯ

〓ち００ＣＣＯＯＣ ∽ に０

一

づＯＱ魚餞ｏ０ ∽ ∽ Ｏ０Ｏ

―

５０＞０

一

〓 ¨

。

∽

口

一

・一〓５

。 ′ ¨ Ｅ

辞一 ¨ 中 ¨ 一に

。

森一 ¨

¨ ¨ 一れ〓 ¨ 一一一 ¨ ¨

一

・・・

ＺＯ卜０ロト ○ ぼ餞． ∞ ［ α ロト餞くェ０

・

・・

・

・・・・・・・

・・・・・

・・・・・ ∽ ００・ ■ ∽ ０Ｏ一０００

・・・・ ∽ ・・００一・・つｒ０Ｏ・卜ｏ ∽ ■ ＞ ∽ ｃ０ ● ∽ ｏ００５日０ｏ日ｏ ε ０ｏｏ・ｏ〓 ∽ ∽ ヽ ′ ｏｏ目口ＪＺ ¨ ｃＣ一ｏＯＱ ∝ ＮＯ ○ ．．゛銀

―

・・・・・・・・・

００ａ出Ｑ ∽ い．Ｎ

・・・・・・・・・

・・・・・・・

・ ∽ Ｏ・ＯづいＥＯＯＯｍｏＣ〓ｏＣｏＯ ∽ 一協〓〓 ∞ ｏＱ ∽ り ∽ ヽ寺０Ｎ・．

〓一 ∽ 〓

咸卜Ｚ口 ∽ 口ぼヽ・卜

ず椰穫歳讐娘

ぼロトくａ

一

・・

・・・

いくぼくヽ口の

・・・・・・・・ ∽ ヽ ′ ０ ● どく戸

罠螢螢罠罠

∽ ＺＯ

・

―

罠ミ罠民畷罠罠騰・・

″ Ｏｚ昌国０

に

Ｎ一 ℃

∽ ＣＯ

∽ ∩ ○ 〓トロ

口。一

おギぶ

ON ∞ ゛ Nヽ

罠民民罠民ミミ曇民民罠爵罠

一

一０・

・・

Ｏ

一

・・

く

6● C

‐

・・・・・・・あ

０

守

・・

∩ α く

〓

め

トく

∽ 卜ＨＸ口

∞ め ∞

ロ ∩ ○

∞ ∞ ∞

・・

20 ..

口くｏ卜０くぼ餞０・

―

卜 ∞ ∞

蓋尉尉燿最誇謂

Ｎ．．四

∞ ０

餞．０

．．． ∽ ∽ 〓Ｃ〇を一一 ′ ¨ ０にＣＱ

一

ＮＯい

∽ ∽ 口００

ａ〓。５〓ｏ一〓づＯ口ｏｏ ≧ メｏｏつＯ

ＮＯＮ

ハ

００

ＯＮ

Ｏ

いヽいヽ ζ いもヽいヽく

゛ト〇〇

０

∞ ∞ 倒

・・・・・・

∼

． ∽ 〓。．゛〓．Ｃ υ ∽ 口め一づｏ〓

卜 ∞ ヽ

ハワ棗寺。召餞卜餞く目０

・・・・・・・・・・ ∽ ゛Ｑ００口００

・・・・・ ∽ ００口一 ∽ ●

・・．．・．・．・ ∽ ＣＯ ■ ●

にＱｏｏ ● Ｏｏ ∽ ■ ０に嘔 ■ Ц ２口０ ∽ Ｃ０Ｏ ● Ｃ ∽ 〓ＯｏＣｔ口ｏづ嗅一ｏＮ・ぃつ。ｏ

∽ ００Ｃ一 ∽ ●

・・・

ハに ∽ づ属

・・

口０一０颯

・・

０００

・

メ一場 ● 口０口ｏ一〓〓一０ＯＣ ¨ ■ Ｏ゛

目５口〓ｏｄ一ｏ〓Ｑ６ヌにＱ ¨ ＯつＯＣ口ＣＯｏ一出 ∽ ．餞． ∞ ―

N∞

Ｅ ■ ●

ロＺＯ卜ぼくＬ

ち智ヽヽ一ミゝ ξ ミミセ ℃ ヽミミ

ＺＯいく

・・・・・・ ∽ ＣＯづ〓Ｃ

・・・・・・・・．・・．・・ ∽ 〓Ｘ・・一Ｏつ〓Ｃ５ ∽ Ｏ０〓 ■ ∽ Ｘ０Ｏぅｏ一〓。 ∽ ρ 口０＞づＱ。麟・・・・ ∽

一 ■ 一 ∽ づｄ

・ヽ

セヽミ、０い＞Ｘ

ＣＯＱＯ国Ｎ。

・・・ ∽ 助 ε

∞

00ト

。

一

＞。

。 ρ Ｅ

一め．Ｎ

・

Ｏトバ

５ｏ

∽ 。

・・

。・・・・・・．．． ∽ づ

づ ∽ 〓００Ｏづｏ〓Ｏ０づｒ ∽ ５ＣＱ一〇 ∽ 一Ｃ ∽ ＣＣＯ。督 ∽ ０〓。ｏ ∽

Ｃ

ＣＯづ

・・

Ｏ〓

５Ｅ

づ

Ｅ

≧ ″

５ｃＦ口

守

っ０く＞国〇〇ＺＺＺく日餞 ∩ ＺくＺＯあ口 ω ［． ∞ 一ロト、く目０

∽ ロス国 ∩ ｚく ∽ 口ぃっ〇

・・・・・・・・・・・・・・・・・ ∽ ・ｏ・ ∽ 軍・口・。・Ｑ・口・．一Ｃｏ．ｒ・ｏ・・壺に・・ ■ に

■ ■ ｏｏＬＬｏｏ００

Ｃ一 ω ∽

N∞

・・・・・・・・・ ∽ ０００００Ｃ

● ● ● ● ● ● ● Ｏ ∽ 一匡

〓ｏ

・・・・・・・・．・．・．・ ∽ ｏ

住 ∽ 一 ≧ ′ ∽ Ｏ０Ｏ ″ ０ｍ ∽ ∽ Ｎ．．・め餞 ∞

● ・ ● ● ・ ● ・・，・・・・・・・・・ ∽ し， ¨ 一 ● ∽ υ

．一ｃｏＥＯ＞ｏ日ｏＥ〓執工 ■ ■，。ｃＯｃ２ ∽ 一＞Ｏ餞尋

・・・・・・・・ ¨

・・・・・ ∽ 一一 ∽ ００

・・・・１，囁お〓 ∽ Ｏ

８一 ■ づｃ ′ ｒ。日以０日

ｏ＞と０Ｑ０ｒ綱Ｏ κ 〓０ ● 〓 ´ ∽ Ｃ０ ∽ Ｏ ∽ 。ｏＯＯ目ｃａ０。。Ｃ ■ 脳罵ｏ０ｃ ∽ 。０Ｒ口 ∽ ∩ 口［銀 ∞ 守め．．寺．守．守守

・・

。

・．・．

・・・・

・・．・．・．・ ∽ ■ ｏ．・づづ一・．．ＯＱ．・０ ∽ ∽ Ｘ０一。０ ● 口嘔 ● 。００口 ● 。一一ＸＸ餞ロ

崎

０＞ロ

tD卜

・・・・・・・・・・ヽ・昼３Ｃ日Ｏ ∽ ∽ ■ つ嘔ＱＱづ〓 ∽ ０８ＣＥＳ・０ ∽ ∽ ∽ ０〓 ● ２一Ｅ０ ∞

露

最尉轟

露最駁轟饉駁駁至 SSS露憂翼壺

ヨ翠露まま燎漱

ＬＯ ≧ ９り，ＦＳ運スコ υ 哺ミざｂいいｍ０ヽＯミミＬ ∽ ０なｔ≧ ０ｓヽ０ミい０いいｎいス０ヽヽヽ哺ヽヽ０ヽぃＬ ζ ミゃＦぃ

撃

ＺＯ卜くっ口く＞口一ｏＨ目ＯＺ００口口くｏＨＺ〓０国ヽ一．０ロト飩く目

∽

卜い口くの国ぼ

０

哺喜一５〇一ｄ ● ０づＯＣ口

．一一口０口 ● ∽ υ ＞饉

。・ ∽ ０一・

。

Ｈｏ一０〇一〇 α 一０

一一２に。一 ● 、景喘ヽ ∽ 。Ｃ

Ｃ

０一０。Ｑめ一

‐

口 ¨ ５ ∽ 〓国

・・・・・・ ∽ づ〇〓一０口

・・・・・・ ∽ ＣＯ ¨ ∽ ０〇一属 ● 切 ¨ ∽ ∽ ００Ｏ ∽ ０ ■ 日。 ● ００ロ

N∞

〇卜 ● 、く． ∞

髯〓ｏ日

・・聖ｏ．口ｏ ■ ．ｏＯ ∽ ∽ 一〇ＯＬ２Ｏ＞日〓ＯＣ ρ に。０む一 ∽ 書００Ｏ ■ ｐＣ０ｏ ● Ｏ口。 ● 一Ｃ〓巳づ ■ 髯 ● ＞＞餞Ｈ図め守．． ∞ め

・・・・・ＣＯ ■ ０〇一０七Ｑ

・ ● ・一．・・ ● ・．・．・．．・．・．・・・・・・。０いさヽヽいヽれヽミや、卦゛ヽ、

」 I凱懇nsぶごこ :Ln量 lLl鰹

Ottawa,tlDntariO,Hl■ oR言 ,Canュごこ

NBS― CCR 77‐ 92

1:i甲 Ч lotingi:リ

High

リ甲an Beha宙 Or

RIse Buildlng Fires:

MOdeling Occupant Mlovement 丁hrough A Fire― Floor from

initial Alert to Safe Egress

ｔＢＩＦＩＥ ξ

Prepared for

Center for Building Technology Institute for Applied Technology National Bureau of Standards Washington, D.C. 20234 and

Department of Health, Education and Welfare Washington, D.C.

~~~T 轟

羹

~ ~

‐~~~― ―

―

二五 1・轟轟亀轟鑢轟轟 ≧ 重墨二二血、墓

―

■

‖Bs_114A(REV

鰯

il:[1ど

■

NBS-GCR

￨.Redplent｀

1l:e2

『

Ъ;7 00 COde

葬● &Pσ bttng Orga話

N∝ 一一一 ¨

RESS

illEiill"Zcrant

ORCINIZA

TASK B,P・

―

t{oodside,

A■ cs■ On No

a"Rep

SIo stahl 翻

評llitial 評』 ::∬ 議漱ま予

硼醜脚罰

CCeSSbn

0。

No'

NO.512225

諄 :r::RcPσ `&Pい

に,″ め

@dComPlete Nat

ional

五 ng

Bure-ag

DeptttIIlent,鳴

-o L

Agencソ

^Sjandaras

切跳1:e

of Health' Education

and welfare ;;1"'

a sicnificant

ot most

ffiu*"rY

覆菫:iV[嶽

:鸞

■ 警菫弯

ζf r wilding

像

′ げ口

鮮 ,■ :[:鰍 Pずと:ち

餌CUN■ 最もSS

ff uttitit"d

い rM b Servl∝ " :罵炉

二為轟鏃、

二・―

F PAGES

UNこ ASSFE2

'` Do Not RelCase tO NTIS P・ n由 3 0riCe 虜 :ヽ lmen｀ :∫ :お二

EE For 6fficial DiSttibiti。

繰

註tehavior in fires: ::冨品 ‐‐ ヽV AILハ pェレ● Iv,rtt-,tstl-ttY

・ ■ ■ ３ ■ ■ ■ ■ １ ● ■ ・ ■ マ

Iご糀 ;Iffi 爾瑶聖

番

・４１ ■ ７

′ NTIS｀

‐ ‐ ――

鱚

lLsLASS

UNCLASSIFIED

~翻簿T丁

ffi'..'--oc

zsotz-P7'

■驀

TABLE OF CON丁「

,「

● ●

PREFACE

lil

TABLE OF CONTENTS

LIST OF F16URES

I I.IG THE il:::NlIPl。人,I T: DESCRIB

I:1 :llitillilCI::[1:l::AllEitill‖ BUlLDlN6 FIRE SYSTEM RISEI BUILDIN6 FlRES Huヽ AN 8EHAV10R IN (H16H― 111. RISE BUILDIN6 FIRES lV. sIMυ LATING HUMAN 8EHAV10R IN H16H‐ AND CAVEATS IMPLICAT10NS′ FUTURE RESEARCH′

20 30 35

Vl. su椰へRY

REFERENCES

FIGURES

CLOSSARY

一一・オ ´ ⋮

ヽ

≡ 電

_H:―

冨

丁丁丁蔦III壕

~薔 :

: 二 [■

『

これe pau15: 良ボ,17日 I Bldョ M― 告￨・ Naticlnこ l Research C口 undl carlada

Ottawa,Ontario,Xl■ oR8,Canada

NBSIR 82-2480

Time‐ Based Capabilities of

Occupants to Escape Fires in Public Buildings: A Review of Code Provisions and Technical Literatu re

U.S. DEPARTMENT OF COMMERCE National Bureau of Standards

National Engineering LaboratorY Center for Building TechnologY Washington, DC 20234

Apri1 1 982

Prepared for:

Home and Public Building Safety Division U.S. Fire Administration Federal Emergency Management Agency Washington, DC 20472

u.s.DEPT OF COMM・

BIBL10GRAPHIC DATA SHEET (See instruction

Pelform:ng Organ.Report N。 13・

PuBLiCAT10N OR

1。

REPORT NO。

April,

NBSIR 82-2480

TITLE AND SU BTITLE TIME― BASED

CAPABILITIES OF OCCUPANTS TO ESCAPE FIRES A REVIEW OF CODE PROVIS10NS AND TECHNICAL LITERATURE

5.AUttHOR(S)Fred Io Stahl,

i. FFnronNtNG oRGANlzATloN

」ameS

IN PUBLIC BUILDINCS:

Jo CrosSOn, Stephen To MarguliS

f ioint o, otnu

l. CoavacdGrant No'

thon NBS, see instructions)

瓦 ■ ,Fと

‖

::I11∵ llttliF:」 VLlltlDARDS W ASHl‖

L982

o(Report&PeriOd COVered

GTO■ ,DoC。 20234

ty,S tCte,

NA|4E AND COMFLEIE Au

@oN Home and

Public Building Safety Divislon

¥と :Lrilrim:lEき llit轟 :il:lment

Washington, DoC._

Agency

0477

io. suFFGMENTARY NorEs

高高語葛丁 T五面石FTttTT面扁焉十漱鶯三載議張讐ぎ沢尋キ轟留寺 liさ橿菱 F7■ 膏嘉「 DoCument desCribes a computer pro8ram:SF‐

:35,F:PS電。ftWare Summary,iS attached.

t includes a Signi″ COnt

i葬

bibliog「 o,"yo「 litercture sur

― ―‐

‐

― ――一二^`

■―″ 十ハ ●マ■

' faC■

■i

litieS Of bu■ lding OCCupants 鸞

響

li肇

fires. in te■ ation tO evacuatiOn time available during :::::illl:[:ii::il::i:::lii:::[:Ii:li:[i::::[::::I:ilh S■ Ons )r■

a are exam■ ned in relatiOn tO COde prov■

::ni:[::[夢 :lili::i:[ti:

:[:::liSil::

〕 Code

r:f ::::rit。

[::::illtill[サ

:1:::ring

are evaluated。 lie

lili:'c:1111:::][el::il

3isions are presented,

dⅢ ∝Ce■ y

Ⅲ

S""市 d¨ り

EmergenCy egress; fire prOtectiOn3 'ire Safety3 human be,aViOr in fireS; human

factOrs;:Lile Salety Code, means of egresS` 14.NO.OF

P RINTED PAGES

13Jvatuegtutrr ￨￨:嵐ltT:I」l:1lillギ目」￨::InmentPrmT:fflT'WashlngtOnl ttG 重露露 il￡

国

order From National Technica!lnfo「 mati,n SerViCC(NTiS),Springfield,VA。

168 15。

Price

15.00

22161

USCOMM'DC

奮

・・●■ 11::

ギ i「

TABLE OF CONTENTS Page vii vili ix

ABSTRACT ......。 .● 0● ●●●●●●0● ●●●●00000● ●●●●●●●●●●0● ●000● ●●●●●● ACKNOWLEDGEMENTS ...。。。。.● ●●00● ◆0● 0● ●●●●●●●●0● ●●●●●●●●●00● ●00 EXECUTIVE SUMMARY ..。。.0● 0● 0● ●●●●●●●●●●●0● ●0000● ●●●●000● 0● ●●●

o....... o........... o o...... INTRODUCTION ........................... l. I Pf Oblem . . . . o . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . o . . . . . . . . . . . . . .

1 1

o...... .... ..... . . ........ o................ 1.2 .. . . ... .. ... . . . .. r. .. . ..... .. the Report ....... 1.3 QrganlzatLon of o . . . . . . . . o. . . . . . . . . . . . . . . . . . . I .4 Tgchnlcal Approach . . o . o . . . . . o . . . . o.. ..... ............. 1.4.1 Study Deslgn and Tagk Organl.zatLon .. . . . . .. . . . . . . . . I.4.2 Llteraturg Rgvigrr . .. .. .... . .. . . . . . . .... .........' 1.4.3 Behavloral Assurnptions Peer Revlew Procedures

ObJectlve and ScOpg

l a5

Strmafy

...........

a....a...

o......

a...

aaa

4 4 5 5

a.. a................

........ o....... o...... o.. o... 2.L Appltcable Code ProvlslonS ............. 2.2 Underlylng BehavLoral AssumPtlons .... . o.... . o. . . .. .. .. .. . . '... 2.2.2 Assunptions Relatlng to the Transfer of Responses Learned During Drll-l-s to Actual Fire Sltuatlons ........

７

PROVISIONS AFFECTING PRE-EMERGENCY TRAINING AND PREPARATION

2.3

７８

Comrngntaf] o o o.. . t. o.... ..... .. o... .. . ..... ... . . . ... .. . ... . . .. . o 2.3. I Problgm ... o.. o.... o o o.. o. o............................ o.... o o....... 2.3.2 tlnderlying Behavloral ModelS ............. 2.3.3 Assessment of Behavioral Assumptlons Based on the

Technlcal Lltgratur€ o....o.......o..................... 2.4 Summary of Gaps ln the Technlcal Literature roo...........o.... 2.4.I Research on the Predictlons of Occupantsf Responses DurLng Real FlrgS ... . . .... o. .. . . .. .. .... .. ... . . .. . o .. .. 2.4.2 Reeearch on the Transfer of Training .... .... .........., 2.4.3 Research on Occupantsr Attltudes Toward Exit Dril-ls ...o 2.4.4 Research on the AcconrnodatLon of Tralning Programs to SPecLflc Oecupancy Requlrgments ...............o.....

2.5

Sunmary

o........ o.....................

o......................

PROVISIONS AFFECTING TI{E PERCEPTION OF THE EMERGENCY O ENVIRONMENT AND TI{E RECOGNITION OF EGRESS FACILITIES ......... " " .... o.... ..... .. o.... .... ....... Provtglon8 r o. Appllcable Code 3. I

' o

o. . o........... 9.2 Underlylng BehavLoral AssumPtlons o............. and of Door the Effect Relatlng to 3.2,L Assurnptlons Wlndow DesJ.gn Upon Egress Route Perceptlon .. .... o...... 3.2.2 Assumptione Relating to the Affect of lllumlnaElon o. Level Upon Egresg Route Identlficatlon .............. and 3.2.3 Asstmpti.on Relatlng to the Role of Visual Slgnage Dlrectlonal- Infornatl-on ln Egresa Route RecognLtlon and the Fomatlon of Energency Egrese Strategies . . . o . .. o

ili

７

８８９３８８９９９２。

２１２１２２

う乙

００４つ

TABLE OF CONTENTS Page

Relatlng Eo the Abillty of Audlble ::d Vlsual Alarn Signals to Effectlvely Alert Bulldlng : 0CeUpaAtS tO a Flfe Thfgat ......o....................... 3.3 COmmentaf] .. .. o... ...... . .. . .. ... . . .. .... . ...... .. . .. t. .. .. ... 3.3.1 PfOblgm ....... .. .. ... .. . .... ..... . .... ..... .... ...... ...

3.2.4

Assurnprlons

Research on DirectLonal Slgnage ..'..................... Rgsearch on Alarm Slgnals ..r...........................

３９

SUnnafy

IdgntlflCatloo .

a...

...........

a.. a...

a...a

a. a...

........................

!.....

a.....

a....a

' ' ' "'

t. " " " '

5.1 Applleablg Code Provlsloag .or................................. 5,2 Underlylng Behavloral Assumptiong .................o........... 5.2,L Assumptiona concernlng the Influence of Designated Leaders Upon Egreee Tlme Durlng Flre EnergencLes " " " ' 5.2.2 Assunptions Concernlng Pedestrlan Movement' Under t{igh Denalty OCCUpanCy CgnditLgns .....,.,.............. 5.2.3 Assuuptions Concernlng the Effects of Building Configuratlon and Architectural ObsEructions on Efftcient Crowd MovemenE .....o............"..'"'

5.3 Commentary ... r.. o.. r.o....................t......... "r " "t" t... '. ' "' 5.3.1 Probleo .... r.... o............................ .o .......... . o .. t. . ... 5.3.2 Underlying Behavloral Models '.... Ehe on Based 5.3.3 Assessment of Behavioral Assumptions TgChniCal LltgfaLUfe .....r............................. ■V

３８３８３８

３９

PROVISIONS ASFECTING EGRESS STRATEGY FORMATION .... t... .. o....... . .. 4.1 AppltCable COde PfOvlSlOnB . . . .. . . . . ... . o. . ... . . .. . . . .. . . .. . ...

PROVISIONS AFFECTING DISCIPLINED EGRESS BEIIAVIOR AND CROIID . .. .... o.... .. ...................... ICIVEMENT ..............

３

３９

4.2 Undeflying Behaviofal ASSunPCIOnS ..... o....................... 4.2-,2 AssumpElons Relatlng to occupaaEsf Ablllt'les to Det,ernlne the Safest and }lost Accesslble Escape Route Under Potentially Streesful CondlElons .. . .... . ... .. " " " o.........................." 4.3 Cormentary ............... . . . ........ . . . . . . . . . . . . | . . . . . o . . . . PfOb lem . . . . . . . . . . o o . . 4 .3.I .o......................... Modelg 4.3.2 Uoderlylng Behavloral 4.3.3 Assessment of Behavioral Assunptlons Based on the TeChnLCal LlEgrature ............r...................... 4,4 Suonary of Gaps ln the TechnLcal LLterature .... ......... ...... 4.5 Sumary .. . . . . . . .. . . . .. . . . . . . . . . . . . . . . . . o . . . . . . . . . . . . . . . . . . . ' ' ' 5.

２４２４

3.4.3 3.4.4 3.5

２４

UndeflyLng BehaviOfal !1ode18 . r......................... Aseesement of Behavtoral Assumptions Based on the TeChOlCal tLtgfatut'g .............o..................... Surnnafy of Gaps ln the Techaical Llteratufe .... . . .. .... . . .. . .. 3.4.1 Reeearch oo the Effects of Door and Wlndow Design ...... 3.4,2 Reeearch on lllunLnatLoo Requlred for Egress Route

3.3.2 3.3.3

3.4

２４

４０４０４０

4l I

46 2? >4

55 55 55 56 56

55 57

TABLE OF CONTENTS Page

5,4

5.5

ln the Technical Literatur€ .... o........ ' 5.4.i Research on the Effectiveness of Trained Leaders in Facilitating Rapid Emergency Escape and in .. . .. . . ....... . .... Avoiding "Panic" .. . .. .. .. . 5.4.2 Research on Lhe Effects of Occupant Loading and Discharge Behavior on successful crowd Movement ........ 5.4.3 Research on the Effects of Archltectural Inpediments

Summary

Gaps

and Physical Obstacles Upon Crowd Behavior and DisclpJ.ingd MOve'mgnt ................................... c..................... .......... o......... SUfmafy .............

PROVISIONS ACCOMI'{ODATING OCCT,PANTIS CAPABILITIES TO SAFELY AND RAPIDLY NEGOTIATE EGRESS WAYS .. ... . . ... ........ . .. ... .. .. . ... . 6.I Appllcable COde Provlslons .......................o............

COrrtmentaf| o. o .. . . . .. . o. o. ...

.....

. . . o .... .. .. . .. . . . .. . . . . .. . o . .. . . ....... ..... ....... o. . .. .........

PROVIS10NS GOVERNING THE CAPACITY OF MLANS OF EGRESS

.. o.. ... . .... .... . . .. .. . .. . ...... 7 .L Appllcable Code Provisions o................ 7 .2 Underlying Behavioral Assumption6 o........... 7.2.L. Assumptie:-s Pertainl-ng to the Influence of Architectural BarrLers and Other Obstruetlons Lo Egress Flow ........ . . . .. ......... o..... r ...... r..... r.. r..

６８

6.3.1 Problern ... r... ... ... .. 6.3.2 Underlying Behavioral- l"lodels ........................... 6.3.3 Assessnent of the Behavioral Assurnptions Based Upon .. .. r. the Technical Literature . . .......... .. . .. ... ...... . ..... nary .. ... .. of Gaps in the Technical LLterature 6,4 Su on Deslgn 6.4.f Research on the Affect of Stalr and Rarop Oceupantsr Capabilitles to Safely and Rapidly NegOtiate EgfeSSWayS ......r............................ t,he Affect of Stress and Fatigue on 6.4.2 Research on Occupants I Capabll"itles to Safely and Rapldly NegotLate EgfeSSwayS ...........'...................o..o 6..4.3 lleseareh on the Ef f ects of Doorway and Door Hardware oa Occupantsr Capabilltles to Safely and Rapidly . ' o.... .. ... Negotiate EgresswayS .... ......... ....t . . .... . .. ....... .............. 6.5 Sunnary .. r... o...

7。

６７

6.2 Underlylng Behavioral AssumPtions ............................. 6.2.L Assumptlone Relat.ing Eo the Effects of Stair and Ranp Design on Occupants' Capabil-ities to Safely ....... and Rapldly Negotlate Egress Ways .... ..... o Effects of 6.2.2 Aseunptlons Relatlng to the Phystologlcal and Psyehological Stress UPon Occupants I Capabillties to Safel-y and Rapldly NegotLate Egress WayS o.........r....................... 6.2.3 AseunptLons Relating to the Biomechanics of Exit ... ...... o. o... ...... . Door Operation .... .... ........... 6.3

69 69

12 I

72 13

73 73 ７８８６

，００′

８８８８

90 92

■

TABLE OF CONTENTS Page

....... ........................................... 7.3.1 Problen .. .. .. . .. . . .. . o .... o .. ... . . . . . ...... . . .. .... . . . . . 7 13.2 Underlylng Behavloral Modgls ........................... 7 .3.3 Assesgment of Behavloral Assumptlons Based on the TgchnlCal Lltgratufg ................................... ..... 7 .4 Sunmary of Gaps ln the Technlcal LLterature .............. Barrlers Archltectural of Influence the 7 .4.L Research on and Other PoEentlal Obetructlons to Egress Flow ........ 7 .4.2 Research on the Flow CapaciEy of Egrese Channels . .. .. .. 7 .5 Srrtnmafy . . . . . | . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . t 7.3

Cornmentary

............................................ 8.1 Sglgction of Code ProvLslonS ............................. 8.2 Underlylng Behavioral AssumPClons ............................o 8 .3 TgChni.Cal Connentaflgs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3.1 PfOblgn StatgmeotS .o................................... 8.3.2 BghavlOral MOdels ...................................... 8.3.3 Assessment of Behavloral AssumpEions Based on the Technlcal Lltgrat'ure ..............................t.... 8.4 Sunmary of Gaps ln the Technical Llterature ................... 8.4.1 OvgrvLgw ... . ... o. ... . ... . ..... . ... .... .. . .. . .... ... .. .. 8.4.2 Pre-emergency TrainJ.ng and Preparatlon ................. 8.4.3 Perceptlon of the Emergency EnvLronment, and r........... Recogoltlon of Egress FacllltleS .......r... o..'.........o..o....o........ Formation 8.4.4 Egress Strategy 8.4.5 Dlsclpltned Egress Behavlor and Crowd Movement ..... . ... 8.4.6 Occupantsr Capabillties to Safely and Rapldly NegOtlate EgresS Ways ............................o..... 8.4.7 Ttre CapaCity Of lleanS Of EgfeSS ........................ o.. .. ..... 8.5 Conclusions ...... r.... . o...................... 8.5.1 Ovgrvigw r....... ..... ....... ..... ...... o. .... r.. ... ... . 8.5.2 Provislons Affecting Pre-energency Training and SUI,IMARY AblD CONCLUSIONS

PfepafatlOn . . . . . . . . o . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8.5.3 Provlslons Affectlng Perceptlon of Ehe Emergency Envirocment, and Recognltlon of Egress FaclllEles ...... 8.5.4 ProvLslons AffectLng Egress Strategy Formatlon ......... 8.5.5 Provislone Affectlng DisclPllned Egress Behavior and Crowd lttovemgnt ................r....o................... 8.5.6 Provlsions Accommodatlng Occupantsr Capabillttes to Safely and Rapldly Negotiate Egress Ways ............... 8.5.7 Provlslons Governlng the Capaclty of I'Ieans of EgfeSS .....

........................

.......

o...

93 94

1ll 111

112 112 113 113

114 115 115 117 119

128 128 128

129 130 130

l3l 131

132 132 132 133 133

133

134 134

●●●●000● ●●00● ●●000● ●●●00● 00● ●●●●●●●●0000

135

APPENDIX A: MEMBERS OF PEER REVIEW PANEL .e c c.● oo 0 00000● 00● ●●●●0● ●0000● APPENDIX 3: INSTRUCT10NS TO REVIEWERS 。。。。。。..。 000000000000● ●●0● ●00● 0● O

A-1

REFERENCES

。。..。

.。

。.00● ●●●000●

0●

.........

B-1

′丁―″∂

P― S

Amounc蛇

lntemational Conference on Building Use and Safety Technology

FOR PAPmS

襲

】Ⅱユ FOR

PAPERS

Hosted by:

WEユ FOR PAPERS

TIIE NATIONAL INSTITUTS OF BT'III)ING

FOR P螂

Dotes:

FOR PA―

Locotion: Sheraton Interaationd Gonference Genter

FORPAmS

R.eston,

―

"t internadonal conferenceg are 'ligcqrragd. at previous papart may addrees a broad range of topical issuss induding. but not linited to. the economics

m FORPAPR 1-一

of ioae+hince recommendationi. liability igeueg, risk aesegsment modale, techaology tralsfal' .giog f aisaf,itity fectors, public education, hazardg of building producte, etc. Participa$oo-dJ6oot"""o"e 6v individualg in the fietde of building design, code anforcement, standardsnatiqg, product ianufacturiag, governnent, b-uilding operation & maintenance, Earageqsrt' niilier education is actively 6ncouraged. Coneuneigroupe and the public at-large wiU be ""a welcome gession attendeeg. Papert are now invit6d from interegted individuals on subfect a1eas euch ag: 6.Human Factors&Beha宙 or During 1. Orientation & Path Findiry E■ ergency Situatons Management & 2. Crowd Dynamics 7.Egress perfomance Criteria 3. Aseembly Seating 8.Ar&Noise QuaLty 4. Movement Safety 9.Per30nal Se(■ nty a cOmfort S. Alarms & Communication Syrtens Participante from many countries will b€ attending tha thry+day event. A eeesion is planned for briif preaentationaof current rrsearch initiatives and priorities within foreigp reaearch

CAH叫

ゴ重ず庸妻鯛懇 CA 必

・日 CALl ｌ CAI 側ｌ

Virdnia

The International Confer€nce on Building Uge and Safety Technology to be held ia November of iSSe -iU bring togetber a broad crnss€ection of ressarchere and other experte m aubiecte relating to building uge end building safety. The obiect of thig confercnce ie to exenine uew information and technical findings in eciences wbich inlluence the use of buildinge. The confer€nce orgaaizers recogaize that eome rt'stBt+ pt"ri"* information."y-b" nscsssary ar a poiat of departure, but papere preeeated .*i

FOR P螂

叫

SCXET{CES

Novembet 78,1984

FOR PAPR

―

― ‐‐

一〇

cslterg.

Abrtrrctl

Authors are requeeted to subnit 5 copies of a 250 word abetract identifying and explaining_the sigaificance of iheir topic. Abetractg ahould mention tbe sponsorehip of thsir author's worl ae -6U ae deta rworrcei and regsarch methodologfes. A ecreening comnittea will male final decisions on pep€rs to be prceented. Subrbdo dcrdtbo: March 1, 1984. Acceptance letters will b6 igeued bry March 30, 1984.

翻 ∬ 聯 C

For addiflonal infornaton csll Stwen L. Biege! AIA at the Nadoad Ingtitute of Building sciences. 1015 Fifteenth strest, N.w., washington, D.c. 20005{2021347-5710.

Gmlcrucr Co8porcr:

Natmal Fire Pmtection Associailm. lnterastional r TheAnarica! In86tute of Architestg Foundatim o

o National Research Couacfl of Canada o U.S. D,epartnent of Hougiag aod Urban Dravelorpmeat o Building Ownem and Managare Asgocieton hteraaflooel

r Envirmmentd D,Bsig! Ressercb Aggodaton n_F

Neme

Add―

SL●_______Z口

_____=nЮ ttL_____― ―――――

Atteldatvt6iling I&tplr' dotrch .ad 1g5a thb Docdj6 of tlc olmloon [t o 106 .r Dcdbb. fir wll.m! rl.qsrrlt enotorcocc dhlbodo [rL scod cmphtrd lqn to: NIls, rolt lltb s[lt"], N.w., s[ft. ?lto, wrr&tto, Ir.G loot SPealer-

υ片 Cl′

ARENCE Lo NICODEMUS & ASSOC IATES 255 No. Lincoln Street Centor Complex, Suito l Dixon, CA 95620 (916) 678-4949 March 4, 1982

Richard Seaton School of Architecture University of British Columbia Vancouver, Canada v6T 1W5 Dear Mr. Seaton:

This is a follow up to our brief meeting during the International Seminar on Lrfe Safety and Egre at the UniverAt thac Eime I spoke of the need to perceive the population as a continu-m of people with varying needs. In the specific case of people _in that continum who have disabilities, the actual numbers of people varies over time, depending on seasons of the year (winter skiing accidents decrlase t[e numbers of people able to ambulate well for example) BBe, wars, disease, etc. On a shorter time scale, many people who are involved in a building fire situation, esPecial!Y a high rise-ltilding fire, bec5me temporarily unable to function as well as before the iire. The reasons are manifested, but are a direct result of the anxieEy, the smoke and its associated contaminants or the physical changes in the environment. Anxiety induced psychological and physiological stress are tolerated differbntly by each member of the population the resulEant effective ais-aUitity can be mental (disorieitation, confusion, uncerEainty, indecisiveness) or physical (cardiovascular dysfunction; stroke or heart attack). Combustion products like smoke, carbon monoxide, hydrogen cyanide, etc. aiso have different effects on people through ailergic reacEion or sensitivity to-_minute quantiEies in the breat[ed air. These Physiologie effects can be so dibilating as to render a Person- unable to walk or speak or even dial a

telephone.

Finally, as the Physical environment changes, 19"s planned they move information becomes available to direct people as(and therethrough a building. Smoke obscures the ceilings (and therefore fore iights;, the tops of door waYg and halls exit si[ns) and even door knobs and "crash" bars (which can

~′ /θ

March 4, L982 page 2

serve Lo indicate exits) as the smoke volume increases. The effect of this physical change in the environment, changes the "rules" governing egress; visual crues or even vision are gone (as with a person who is blind or near brind), the breathable air space does not permit walking at a brisk rate, but rather crawling (not unlike the rate of a person with severe arEhritis). My purpose in reiterating these rather obvious facts is to help crear away some of the "smoke" that tends to confuse the question of who is the "handicapped group"? It is a large group (as much as 307. without fire involvement), it varyi wich time, and it especially increases in numbers during an actual building fire involvement. My hope is thar your thoughts about building design and retrofit or egress requirements will include the total conEinr$n of population. ,:

Along that same idea, the use of the "safe area concept" (or compartmentation) as an integral plan for protection of elevator lobbies is an extremely attractive one for dealing with the disabled. If total building evacuation can be avoided, and the occupants protected for a period of time then'rescue'can be effected from a protected area (i.e. the elevator lobby). In addition, fire and smoke are denied the remainder of the building via the elevator shaft, and fire surpression personnel can addreis fire fighting activities from a prorecred staging area (i.e. the lobby). The 'safe area'concept is not new, of course, but it has been associated with high costs, inflexible design, eEc. That is not the case any longer. Thankfully, free enterprise has moved into the area and produced a product that makes low, flexible comPartmentation a reality in new construction, but especially so in reLrofit and building rehabilitation. This product is a flexible fire barrier, fully tested and listed by Underwriter Labs (carries a I and L-L/2 hour rating). It is a mulfipass barrier with wide open (up to 400 feer)'applieation and is automatically deployed when smoke is detected. It is self contained, does noE depend on building power and does not interfere with normal HVAC operation when not deployed since it stores in the wa1l. The enclosed literature provides additional information. Please contact the manufacturer directly for specific questions on cost installation, etc. I bring this to your attention because I think it ean help to resolve the question of need for persons with disabilities involved in a high rise fire. I have used the fire barrier in a retrofit building and I know it works. It has recently been smoke tested under the auspices of ICB0; the test supervised by Gus Degenkolb. I am certain that he will be willing to express an opinion on the potential usefulness of Ehe fire barrier.

March 4, 1982 page 3

in Maryland - it was a for attending !h".seminar in your area of expertise and pf""!rri"--f'or *" to meeg pe[ple weil. rf you have.quesrions ro be able ro"'Jnii";t-i'deai as noE hesitate to write or call' do of me in this *"ti.t,'please 1 am looking .forward- tb a continuing involvement with the Thank you

egress grouP.

Sincere . NICODEMUS tation Engâ&#x2013; neering Consultant Rehabil

cLN/yr

ノFEE

n high safe彎五

ttise St a matter bu重 lcttings is no ttongeF j■ ■

卿

of conscience or codes。 ¨

一脳一ｒギ

らｔふ It'sこュロ ■ natteF

Of Su籠田van.

チや中

Photo counery ol D.lly lieus

1翻

o do醜電Ⅷ「馨證t atty岨 1壼血蠍.abovo

People around the worid tte

thatthe ne対 ble fire barriers necessary

becoming increasinglソ aware that

to create these sale areas were proba‐

mostlives lostin high rise fires are a

blyalon9wlyinthefuture.

result ofSMOKE,TOXC GASES, and PANIC.Three prob!emsthat are

VHAT ヽ

ing smoke and toxic gases One

Fire ottcials have agreed for some bundingS depends on: 1.rapid isOla・ tion or olovator si、 afts to

that passes tha low back‐ face tempera‐ ture rise test in other u′ ords,one that、 vin remain cool on one sido

、 vhile an inferno rages out of cOntrol

on the other An automatica‖ y acti‐ vated.easy to operate(even for the handicapped)multiple pass‐ through, continuous closing barrior that operates on its own fail‐ saFe power supply.(Dne that can alも vays be ovor ridden manua‖ y,retracts into theし va‖

and nover intorferes with the nol‐ rnal

trattc or usage of the btllding (Dne

that can span unhmited widths.is remote!y contro‖ able from central 1:●

‐ 卜 li l゛手=il・ 1:l`:.1■ ■ ・ ==IF111だ ■千

。

smoke,2.in‐ place protection or・ SaFe

Refuge''areas for those who And themselves trapped on the upper 100rs.

Unfortunately,most alsO agreed

，

申ｎ０ ①ｒ，Ｗ耐０一

refuge and elevator isolation, but for a wide variety of applications for both life and property safety. Says State Fire Marshall Grant Walker, "ln my judgement, the flexibility and fire protection advantages of this new flexible barrier represent one of the most significant breakthroughs in recent memory." Fire code analyst John G. Degenkolb, says, "The model building codes are in basic agreement in requiring clevator lobbies of high rise buildings to be separated from the remainder ol the building by fire"rateci barriers. This UL R6799 Fire-Guard barrier is the best means for accomplishing that requirement."

engineering spaces and can assuro contamination‐ free safe areas irom

[:<'r f urtlter itrforll: r,licrrt contact

central status monitoring prevont vertical spread of ire and

2nd

their use not only for high rise

That is,a UL tested,l and l%hour

diricult to resolve.

tirne that life/6re saFety in high rise

ｅ

Fire officials and analysts who witnessed the testing of the barrier at Underwriter's Laboratory, are seeking

iabeled retractablo Flre barrier that is alsO exceptiona‖ v eFfective in block・

CAN tVE DO ABOUTIT? THREE SIMPLE THINGS: 1.PROVIDE SAFE REFUGE" AREAS ON EVERY FL00R, 2.PROTECT ELEヽろ0ご ORS, 3.REMOVE PttIC.

♂ａ］ヽｄ一＝一３

\'{)rti ilf('it \Vflft-l),,,,I

THESE SAFE REFUGF:Ait::■ 「ヽ Af?FN()W POSSII〕 li l(}:1 BOTH NEW AND 11)く ISl li・

it i)t"l(:lltiltivg

or i'.rll toll-frce i-iiitlr-,lir3-8494.

l´

BUil.1)lNCS tVITH l・

llli til´

LABEL_1:D R6799 FHく li‐ Gtlハ itr) Fl_EXH31´ 1l BARIRIEIモ

1865 South 3480 West Salt Lake City,Ctah 84104

Circle No. 51 on Reader Sorvice Card ド

llヽ

1]j()t'1ヽ

NAl.― 一 Si二

:ン

1'E、 11)lll1

1981

・ 71

９

Conference onl,ife Safety andttre Handicapped

XpSO

メ I確￨￨

1980C《，一いＶ

andd■ (

一 ■ ■一一離

SEP 24'80,

ぼ一ヽ ■

′

ｒ 111予

「

AIA Reseorch

Corporotion 1735 New York Avenue′ WOshington′ DC.

N.VV

20006

C.A. Tiers, Acting Dir. School of Architecture University of British Columbia Vancouver, BC v6丁 115 CAitADA

￨■

1■ ￨■

￨.｀

:￨:i:￨￨

y,せ ● ■ヤ′ : ヽ、.

■ １１１１

鉤

‐

￨￨ ‐ `' 1,I

II.37三

イ ,ザ

品島品 J静ふ

The National Task Force onlife Safery and the Handicapped, in conjunction with the AIA Research Corporation, is presenting the 1980 Conference on Life Safeqv and the Handicapped. This national confbrencewill rake place ar Howard Universit-r'in V'ashingtorr, DC., from Sunday, Z6 October, to Vedne sday,Z9 October. Ia maior goal,lies:in the dltimate creation of a life-safe environment forthe disablecwhich■ 11l a13o benentthe able_bodiedcommuniwinⅢ e .

紺

一 ■ １ ■

雫9,IPtlle耳 9▼ ■rd

Й::￨:鍛熱 1二毎亀質薄鸞轟轟‖IⅢⅢ中￨:学

unWeゃ ityinHttamDeel聟

ビ,11:

職鰤嚇椰 i籍激欄螂靱聯 i藩瀧謝灘 : :繭鷲畿蠣蓉壻:l轟構鋏鰈1霞暴鼈善塞塞妻 ili[鐵]脚聯難 1爵

鰍

=1中

￨￨

れ早総二 l::Ⅲ■ ゞ :望 ,1聟 t肇￨

￨ 11帯疑二

LⅢ ⅢⅢ11。 ⅢⅢllllⅢ Ⅲ 戴

藝"111壼舅

ゃオなSf■ ■マ

響盤鶴樹＝出 ¨

)ヽ 1麹

￨

θノ￠クアー

努鷺 ♂ 壼

鐸羹

髪

デ

．緊

≦ 幕

罵一 ¨ 鮮 ¨ 曜撃璃評識Ｆ誕理．轟〓０コつｏ

００つ一コ００〓０一二〓

ｏ工一。Ｃ ¨ ０００００００ ● 一Ｅ。一。ｏっ〓。。上Ｃいレｃン一。Ｃｃｏ ´ 〓ｑ

・０ ● 中０一一 ● ０〇一．０。 ● ０ ∽

一一 ●

０一〇０一の０ ● 〓二

一ＣＯ

一。

Φ 。〓

Ｃ一＞一ｔ二Ｏ一

群鯉一農墨嬌 ¨ 肇だ砂¨ 〓紫鵜酬譴警轟調謡

詳墜。Ｌ襲躍。謀髪・詢調

ＯＣ ● ０

難轟癖曜畿疑羮

静誕

訪譲

詳

¨ ．疑

巽響・誕．配唱 ¨ 凄弊 ≒ 誠 ¨

巽

０〓

ｏ一〇０

００

い０

∽ 一 ∽ ＞一。ＣＣ一一ＯＣ０〇

一 ∽ ００

ＯＣ一〓Ｃ〓＞０

Ｃ〇一だっっ０。０Ｃ０コ一Ｏ．。ＯＣ０一＞０。 ∽ 一ｏＣ ● ０ ● Ｃと一〓０ ● ０

∽ ∽ 一〓一。＞ ● 一〓００

．０つ０一〓０ ● 上一一 ● ０〇一 Φ 〓。一ｏ一〇００一 Φ 〓 ¨

罷雲れ議委慶霙腱囃露妻礎冒馨

鰐一〇誕〇証醍ＯＣ駿嘲一ｏＯ難莉一ｏＦｏ試擬工 ● 鵠 ●群襲¨ ¨ ０静 ● ヽギ ∽ 一＞ ¨ Φ読聴つの０鈍配〓 ∽ ∝鵠製

離確〓義慮畔粘脚搬

●

ユニ三電 0

● ニ菫ユ書一 ● Ｏ ● ０ ● ● ● ● ● Ｅ ● ＥＥ ● ｃ ● けく ●

ｈ一０口づ口 ● 一関０一 ■ ０ ● 一 ● ． ● ０一〇０一０ ● ０一０

● ＣＣヽ ● ヨョコい０、一。０

ｏ卜 ● ● ョｃ一ＥＯＥ〓 ● ぃ ● ● ンく

〓〓Ｅ ● Ｃ ● と０００Ｃ。 ● ● Ｅ二 ● ｔｑ〓． ● Ｏ出Ｅ。。 ● ● 〓 ● ０

● ● コロ＞ｄ ● ● ０一口綱一ａｎ

．．． ●

∞ せ寸

卜ヾ ●

ト

∞

∞ ０ ●

０００

００ ∞

寺

０

ｒ一

い０．Ｏｒ

００．寺ｒ

卜 ● ．Ｎ０

０ ● ．ＮＮ

せ０． ● ゛

ｒＮ． ∞ め

００．Ｏｒ

００．Ｎｒ

∞ ０．一Ｎ

００．０

００．一

寺 ∞ ．せ

ゆ．Ｎ

ＯＮ．０

００．Ｎ

０．一

００．０

つ０ＯｏＣＣ。。一ＯＣ

０一 ● ０ ■ 一

ｏ、と０い０。 ● こけＯａ。

Ｏ Φ ００ＣＣ。。

０ ● ● ● ｔ一

● し〓０・一０ｏ０〓一 ● ０。

一ｏ Φ ｏ。 ∽ 一つＣＯｏ一ＣＣ ● Ｏ一と ●

ding Research Establishment Information Paper

CI/S,81976

レ FFl薔翔督◎蒙曹躙轟量:③ 鶏

υF― D

K〕〔

August1979

The incidence of residential fires in London-the effect of housing and other social factors S E Chandler, BSc The Fire Research Station has used data for residential ftres in 1972 in 32 Greater London boroughs in order to see to what extent fire incidence is related to housing and social

conditions. ξ Ｌ

towards protecting these groups if this was thought to be desirable. It may also be desirable at a later stage to deter. mine to what extent conditions in London mirror those in the rest of the country, so that nationally, consideration may be given to the most effective use of available resources.

INTRODUCTION lVhile studies have shown there to be relationships between multiple fue fatalities and age of buildingr, there has been

little published work on the relationship between fire incidence and social factors in Britain. American studies have, however, established relationships between residential fues and types of urban neighbourhoods2'3.

In a recent study, data for residential fires in London in l972have been analysed in relation to a number of 'social indicaton'to see to what extent fire incidence and fire casualties are related to housing and social conditions. Iondon data were used as they were readily availablea in a suitable format for analysis; 1972 was chosen as it is the middle year used in other sociological analyses in progress. In the full study, the following factors were considered.

FIRE DATA There were 8284 fires in dwellings in Greater London in 1972. These fires have been analysed for each borough and related to type of dwelling (house or flat), sources of ignition, circumstances leading to ignition and other factors. The City of london is excluded as relatively few people (four to five thousand) actually live there. Figure I shows the fire incidence n 1972 in dwellings per thousand population for each borough. The average

Housing tenancy (owner oocupied, private rented etc), density of population in residential areas, lacking basic amenities (hot water, etc), shared households and rateable value per person.

Sociat sociocconomic group, unemployment, country of birth, age groups, family stability. Family stability was considered to be reflected by the proportions of children in care and illegitimacy.

The factors chosen were those that could be reasonably expected to influence fire incidence and for which data were available. This report considers only the more important initial results; it is intended to publish a detailed paper later.

11::::::::::1111::

Three quarters of all fire deaths occur in dwellingss. One objective of this type of study is to seek those sections of the population that are at higher than normal risk from fire. This would enable any national resources to be concentrated Euilding Fesearch Station Garston

Watford VVO2 7JR Telephone: Ga.ston (Herts) 74040 Telex:923220

Fire Research Station Borehamwood Hertfordshire WDO 2BL Telephone: 01 -953 61 77

r.oo

o.75

Fires per thousand population

For reasons of claraty, borough names are omitted

Figure

Fires per thousand population in Greater London boroughs (1972) (Inner London boroughs shown by dotted lines)

Princes Risborough Laboratory Princes Risborough

Euilding Research Esrablishment Scottish Laboratory

Aylesbury Buckinghamshire H P17 9PX Telephone : Princes Risborough 31O1

Kelvin Road, East Kilbride Glasgow c75 ORZ Telephone: East Kilbride 33001

fre incidence in the GLC area was l.l3 lires per thousand population, the figures for Inner and Outer London being I .42 and 0.88 respectively. The domestic fire frequency in the whole of the United Kingdom in 1972was 0.95 fires per thousand population. domestic

Hackney. Camden

Kensington

〔

The ten boroughs with the highest incidence of residential

Iires per thousand population were all Inner London boroughs and are as follows:

Hackney l.9l Camden 1.88 1.86 Tower Hamlets Kensington & Chelsea | 39 Islington 1.62

Westminster

hmbeth

Wandsworth Hammersmith

Southwark

1.60 1.59 1.47 1.47 1.37

In the following analysis, the above data are related to some

ＣＯ一 ● 一 ● ０００Ｃ ● ０っ０工

rBarking

００ ● ０一

of the social indicators listed in the Introduction. Details of these indicators are contained in the GLC Annual Abstract of Statisticsa.

SOCIAL PARAMETERS

(a) Orvner occupation The boroughs with thehighest level of non-owner occupation were:

Hamlets Westminster Hackney Southwark Islington Tower

〔 -

20 r .

98Vo 89Vo

Percentage of non-owner occupation

Figure

100

Inner London boroughs Outer London borouqhs

88Vo

Relationship between fires in dwellings per thousand population and non-owner occupation in London boroughs

87Vo

87%

All these boroughs are on the list of those with high fire incidence. The average level of non-owner occupation in Greater l,ondon was 60 per cent, the proportions in Inner and Outer London being 82 per cent and 46 per cent respectively. Figure 2 relates fire incidence to the percentage of nonowner occupation in the 32 boroughs and shows a strong relationship significant at the I per cent level between the two variables (conelation coefficient 0.91). The notable exception is Barking where there was large-scale council development in the inter-war period and it is possible that tenants in this development had similar behavioural characteristics to owner occupiers. However, the existence of the correlations in Figure 2 and elsewhere in this paper

The average population density in the GLC area was 135 persons per residential hectare. The correlation coefficient between fire incidence and residential population density over all32 boroughs was 0.86, also significant at the I per cent level of significance.

(c) Lack of basic household amenities The measure used was the proportion of households in 1971 that lacked the exclusive use of hot water, fixed bath or inside WC. The boroughs with the highest proportions of such households were: Islington

497o

Newham Hammersmith Hackney Haringey

457o

4s% 39Vo

39Vo

do not imply direct causal relationships between fire incidence and social or behavioural facton.

(b)

Density in residential areas The residential population density is classified by 'persons per residential hectare'(260 hectares I square mile). The most densely populated boroughs all appear in-the list of those with high fire incidence and were:

319

notable omission is Tower Hamlets, where as a result of major redevelopment, provision of basic amenities by the council had been considerably improved, but the fue incidence was high. This suggests that the provision of amenities will not in isolation reduce fire incidence. However, the consequences of fire, in terms of fatalities may well be less serious than before such amenities were provided. Earlier unpublished work showed a correlation between fire fatalities and the lack of basic household

3O7

amenities.

Westminster

441 Kensington & Chelsea 368

Hamlets Hammersmith lslington Iower

Apart from Newham (formely West Ham and East Ham) where domestic fire incidence was l.l0 per thousand population, these boroughs had high fire incidences. A

332

The correlation coefficient, significant at I per cent, between the 'lacking'factor and fire incidence was 0.82.

In the whole of Greater London, 24 per cent ofhouseholds lacked exclusive use of one or more basic amenities in 197 I .

(d) Children in care It is posible that there

is a relationship between fire and family instability. To examine this theory, the proportion of children in care is used as a measure of family instability. The borouglrs with the highest proportions were:

Tower

Hamlets

2.65%

DISCUSSION AND CONCLUSIONS A number of housing and social indicaton are shown to bc statistically correlated (at the I per cent level) with lrc incidence in dwellings. However it is not possible to establish direct causal relationships. Although such relations mry exist, further analyses are necessary, using more sophisticated statistical techniques, to attempt to unravel thc complexities of these relationships- It is also possible that the real underlying factors which give rise to the housing and social conditions considered are economic6'?.

Kensington & Chelsea 2.477o

Hammenmith Camden Westminster

The social indicator most strongly correlated with lire incidence is the proportion of children in care, which is thought to reflect family instability.

2.07% l.98Vo 1.967o

All

these boroughs have high fire incidence. The average proportion ofchildren in care varies sharply between Inner London (1.77 per cent) and Outer London (0.57 per cent). The corelation coeflicient with fire incidence is 0.94 and statistically significant at the I per cent level.

(e) Country of birth It is believed that the life styles and cultures ofparts ofthe community might possibly increase or decrease the likelihood of fue. Preliminary examination of the data showed no correlation between fire incidence and the proportions of people bom outside the British Isles. A similar result was also shown for people bom in the Indian sub+ontinent. However, there was a statistically significant correlation at the I per cent level (correlation coefficient 0.64) between fire incidence and persons born in the Caribbean or Africa. The boroughs with the highest proportions of people born in the Caribbean or Africa were:

Brent Hackney Iambeth Haringey Wandsworth The average

9.4Vo

8.lVo 8.lVo 7.5%

Fire frequency appears to be independent ofrateable value per person and the age distribution of the population. However, the incidence of casualties is generally highesl amongst the young and elderlys. Casualties will be discussed in the

full report, which will also examine the most prevalent types of fire and their relationship with socioeconomic goup.

ACKNOWLEDGEMENT

The assistance of a number of members of staff of the Building Research Establishment, the Greater l.ondon Council and the Home Office Statistical Department in the detailed statistical analysis and interpretation of the results is acknowledged by the author.

6.2Vo

proportion bom in

these areas living

London was .5 per cent in 1971 3

Of the housing variables examined, owner occupation and the lacking of amenities are the most strongly correlated with lre incidence. If the real underlying cause of fire is carelessness, it is feasible that people who own their own homes may be more careful than those who live in rented property and therefore may be less likely to have fires.

in Greater

TYPES OF FIRE Some relationstrips between fue frequency and types of fire are investigated in the full report. There are indications that in areas where high fue frequencies occun, the following types of fre are more frequent than the United Kingdom average:

(D fires attributed

to children, matches, smokers materials, malicious ignition or unknown causes

REFERENCES Chandter S E- Multiple death fires. Fire Research Technical Paper No 22.lnndon, HMSO, 1969.

Munson III J. Residential fires and urban neighbourhoods: an empirical analysis of Charlotte, North Carolina. Princeton Univenity, I 977. Munson Ill J and Oates W E. Community characteristics and the incidence of life: an empirical analysis. Princeton University, I 977.

Intelligence Unit, Greater London Council. 1972 Annual Abstract of Greater London Statistics. London,

GLC,t973.

(ii) fres where

things were overturned, flared up, spilt,

carelessly placed or carelessly disposed-

Home Office. United Kingdom Fire Statistics 1977 -

Iondon, HO,1979.

If the proportions of fues recorded in these categories in Inndon boroughs could be reduced to national level, there would be a reduction in the annual number of dwelling fires in london of about 600 (8 per cent) most of this reduction would occur in Inner London.

Ash M. A guide to the structure of London. Adams

Dart, Bath, Somerset, 1972. Department of the Environment- Policy for the inner cities. Cmnd 6845. London, HMSO, 1977'

structure hit, load peaks of very short duration were not of

very much significance, the average force largely determin-

ing the extent of damage. 'Corrsiderable damage might however be expected to result from vehicle impacts of this

kind'he warned. A colourful reminder of the hazard to tall buildings from aircraft collision was given by Dr Mainstone in his reference to the infamous case of the 825 Mitchell bomber which flew into the 79th floor of the Empire State Building in 1945. This had caused considerable damage at lhe point of impact and in a lift shaft'down which one engine fell', but no progressive collapse. Again the author sounded a warning note.: 'Bearing in mind that the structure of this building is very substantial with all the continuity desirable to permit the bridging of local damage, and that the impact it suftered might easily be exceeded in severity more than tenfold today, there is now a real possibility of a far more

He considered that the basic hazards could be reduced

in obvious ways such as safer gas installations, and a greater concern for safety in road-traffic management,

aircraft operations and construction processes. He noted that one of the simplest provisions in building design that

could reduce the risk of major damage lrom runaway

vehicles was to protect vital load-bearing elements of i structure by bollards or banks (as recommended in BS Code of Practice CP 110).

The chiel way of limiting the gas explosion risk was, he said, the provision ol adequate ventilation and vents -the first to limit accumulation of flammable gases, the second to vent any explosions that occurred at a safe pressure: ,As yet no statutory requirements have been laid down but some guidance is available and research on the factors determining the build up of pressure is continuing at BRE,

catastrophic accident.'

Looking at the likelihood of buitdings being hit by aircraft, the author based the foltowing conclusion on some estimates in a recent paper published by the US Department of Commerce. 'For a tall building near a city centre the risk during a GGyear life might be between 1 in 10c and 1 in 104 depending on size and location.' 'lf presenttrends continue' Dr Mainstone added, referring to the pressure for heavier road transport in the UK, .the chief change to be expected in this country is an increase

,.1"\

The fire risk

in the proportion of seyere impacts from heavy vehicles'.

Safer building design

To conclude the paper, the author

commented . on the implications of accidental loading on building design. He thought that, on the evidence he had found, even gaseous explosions might sometimes result in loadings greater than those that had been envisaged by Regulation D19 of the 1972 Building Regulations and that still higher loadlngs might arise from aircraft impact: 'These hazards must be given special consideration if protection against them is thought to be necessary'.

3 Measured (upper curves) and estimated (hottom curves) loadings from horizontal impacts with a building by a heavily laden lorry and an aircraft, and a vertical impact by a heticopter. Loads are plotted non-dimensionally as reactionlstatic weight

uT -ots-

The risks of fire and the measures that can be taken to reduce them were the subject of a conference paper by Margaret Law, until recently a research worker at the Firi Research Station of BRE and now with Ove Arup and

Partners. She concentrated on the 'human' aspects of fire hazard, for example, the social acceptability of various types of fire risk, the amount of money that should be spent on fire protection, the attitudes of people to fire hazards, their behaviour in fire situations, and the role of education and publicity in promoting fire safeiy. ,Until recenily'the author noted 'these aspects have received litfle consideration compared to the physical aspects of fire protection'. Fire risk and society

considering the risk of death by fire, the statistics prepared and published annually by the Fire Research Station were quoted. These revealed that between g00 and 1000 people died each year in the UK, mostly in small domestic incidents, and about three-quarters of them in fires where there had been only one fatality. Comparing these figures with those for other causes of accidental death, it was clearly shown (see Table) that there was a much higher risk of dying during road, air or even rail travel than in a fire.

The importance of society's attitude to comparative risks was emphasised: 'There are certainly indications, from recent public disquiet in this country, that the risk of death by fire in hotels, which is an order of magnitude higher than in dwellings, is not acceptable even though it is still lower than the risk of death from travel by car. The special problem posed by tall buildings is the possibility of the loss of a large number of lives in one incident, that is, there is a potential disaster should a fire occur. Judging

by public reaction to previous disasters, society demands that considerable effort be devoted to preventing their

reoccurrence, and indeed most building regulations can be

traced back to disasters.' The author added that this presented the fire engineer with the dilemma of anticipating

the special degree of economic fire protection that would be demanded for tall buildings, taking into account people's feelings that death by fire was particularly horrible.

Reducing the fire risk 'Does the specific quality of tallness affect the risk of fire occurrence?' In posing this question, the author thought that it could be argued that special measures were needed to reduce the chance of ignition and fire growth because of the number of people likely to be at risk and certain other leatures that tend to accelerate fire and smoke spread.

sprinklers are required for life safety anyway, then why not exploit their value in saving the structure?' She went on to suggest that what was needed was some means of tradingoff active against passive protection on a rational basis so that the optimum level ol fire protection, which minimises the sum of costs and expected losses, could be provided, Criteria for assessing the trade-off ol passive against active protection have been developed at the Fire Research Station. Commenting on these, the author said that before they could be put to practical use it would be necessary to undertake an analysis of the costs ol sprinklers and detectors and of providing fire resistance of the structure. lt had been shown that the important factor was the cost of reduc-

ing the probability of failure, which had received some attention so lar as structural protection was concerned, but little had been done in defining improvements to active

Regulations could be strengthened by placing controls on

protection and the costs. lf loss of life is to be included in calculations of the above kind, a monetary value must be placed on saving a human life. A tentative estimate of â&#x201A;Ź50 000 was given in the paper tor the overall value of life as calculated by the Fire Research Station allowing for the risks that an individual was pre-

extra hazard.

pared to take.

the amounts of flammable contents such as furnishings and fittings used in tall buildings but the author speculated whether occupants would be prepared to allow ihe regular inspection of their premises that this implied and whether they would agree to discard contents deemed to create an Another approach suggested in the paper was to reduce fire incidence by giving people a better understanding of fire hazards and safety requirements. Little measurement had, however, been made of the effect of education and publicity on fire safety (unlike road safety) and such information was needed if the best advantage was to be gained from limited resources.

'There is a strong argument, from a fire safety point of lor making sure that a tall building is provided with trained staff, possibly with an internal fire brigade, who would play a positive role in fire prevention and control', the author thought. 'Conversely, where there is a policy of reducing staff, serious consideration needs to be given to increased automatic detection and extinction equipment'. Despite the long-standing aim of the fire engineer to compartmentalise a building to contain a fire, problems of smoke and fire spread still remained for tall buildings, cautioned the author. For example, there was the risk of external spread ol fire from floor to floor because ol flames emerging from the windows, and also accelerated smoke spread in tall buildings caused by the stack effect that occurred when there was a large difference between inside view,

and outside temperatures.

ln view of this, many people considered that, for life safety in tall buildings, the basic aim should be to attack fire automatically at its source by the use of sprinklers. 'The great advantage ol a properly designed sprinkler system is that it restricts the size of a fi re so that the amou nts of heat and smoke are limited and the measures to deal with them can be designed accordingly'. Although doubts about the reliability of such systems were often expressed, the author added, there were weaknesses in all systems, whether active or passive. The work of Roger Baldwin of the Fire Research Station in this field was referred to, in particular his suggested levels of design reliability for fire

protection systems

for different types of

occupancy industry, houses and storage lor example. He suggests that a return period of '10000 years (the average time elapsing between occurrences of structural failure caused by fire) would be appropriate to set the level of fire-protection design reliability. This has been adopted for other potentially disastrous lailures such as the flooding of dams in Holland. The values thus obtained make quite reasonable demands on the systems.

Active or passive Since sprinklers reduce the size of a fire many people have argued that if they are installed, there can be a corresponding reduction in the passive methods of structural fire protection - in the compartment walls and floors for example. In line with this thinking, Miss Law posed the question .lf 14

Behaviour of people caught in a fire 'When an attempt is made to assess the effectiveness of methods ol assisting people to escape the fact must be faced that little is known or understood about how people will actually behave when confronted with a fire situation'. To illustrate this point, the author added that it had only recently been realised that currently-adopted escape provi-

sions might not be suitable for holiday centres specifically designed to provide separate entertainment areas, since people would not go to the nearest exit if a fire occurred but would try first to locate their family. 'Some thought has

been given to possible methods of communicating with the occupants of tall buildings, but even if a satisfactory way could be devised, would these people accept instructions or reassurance from the fire brigade or would they panic?' Miss Law considered that the limited amount of data available on people's behaviour in fire was not adequate lor assessing the various possible measures {or control such as education, fire drills or regular visits lrom fire prevention omcers, Types of building which gave rise to increasing public concern over fire safety were pinpointed in the paper's concluding remarks. They were the covered shopping mall in a town centre development, the large holiday centre and not least the tall building. 'The problems were not only the physical ones of providing fire containment, structural

protection, detection and extinction, but also those of

assessing the effectiveness of various protective measures

when human reactions to fire conditions in these new situations were not understood,' Risk of

latal accident (UK)

Cause

Situation

Accident rate fatal accidents/ person/l@

exposed hours Fire

All accidents

in home in hotel

0ă&#x192;ť

in home able-bodied man in home

British industry coal-mining travel by car travel by train travel by air travel by motor-cycle rock-climbing

40 57 5

240 660 4000

0F-D/3

HOWE :li彗横

RESKY● UR LIFE‐ Ⅷ ■

織滞欄

Iサ

棚麹協認

Joe never travels by air. Fred forbids his son to enter an amateur boxing contest. And Bill gives up hunting after reading about a fatal accident. In each case, they're scared of the risk. Yet Joe smokes a pack a day. Fred lets his son swim at the cottage. And Bill goes on a canoeing trip. None of these activities scare any of them.

The hard facts tell a different story. Smoking is about six times as dangerous as air travel. Swimming is about

five times as dangerous as

amateur

boxing. And canoeing is abotit l0 times

as dangerous as hunting. Surprised? We all go through life

worrying about the risks we take. Yet seldom do we do it intelligently. We apply a mixture of fact, fancy and superstition, add a large dose of misinformation, and then make our choices. The result: most of us end up trading off the risks in our lives with all the precision of a drunken goat. There's a more sensible approach, and that is to develop a pragmatic individual attitude to risk. You simply dismiss all risks below a certain worry threshold as not worth worrying about.

levels, you weigh or other benefits against the risk taken. You know, for instance, that the automobile is a highrisk method of transportation. But because o[ convenience and comfort you continue to favor -its and habit use. And if- you wish to risk your neck

For higher risk

pleasure, convenience

一コキ一Ｃり０ハ ″ 一＞ ● ２〇一く卜りっＪＪ一

Living beside o nucleor power plont

Wolking in the

woods

Deoth in hGd:odults (house fires only)

iAI:ド

ー農轟■協港翼

. ヽ `.

′す―θ20

NBS|R 81-2438

B41Jι h ra. tr.J stah('.

Human Behavior and Fire Emergencies: An Annotated Bibliography

December 1981