This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
■|'
: ‐
11 11′ .: `.
:`‐ il1 1. :4■∵
:■1,、
)t
`↓
,
■
火災安全性評価に蘭|す ぶ基礎由耕究・ |
別冊資料編"火災安全性評価に関する文献1 ‐..1111‐
_
■ l卜 |‐ン『
r絆一 一一・ ・一〓一r 一・一」一
11ヽい・1, ■ .
キヽ、,
ヽ―
く
一¨西野悦子
●
た
な
一一 , 一
一■“一
: ヽ
二l:=:Jl,′ :■ 'I=:
・ 1 1
| ド :
iF.ギど
1 1
l●
L・ 、 .I,:1を 1■ _■ ・
‐
1111
113キ l t fl liflヤ 1 ) '111 1 1l F tt ι「 . ・I .111ど : F l「 、ゥ ヽ
El il. 16 *E[fiffiHntil.aBo*ifr,tro*ttc :0.5 (c€l,/gf .dg\p:0.5 (g/a;t). t=l
(crn)l.: 1.8 x ltrr (cal/qn.
min . deg)o{€
“■■■
―
一| | |
委 員 菅原 進一
委 員 神 忠久
◆建設省建築研究所
東京大学助教授
自治省消防研究所室長
第 5研究部室長
第 5研究部室長
第 5研究部研究員
第 5研究部研究員
建築試験室研究員
幹 事
幹 事
幹 事
幹 事
幹 事
若松 孝旺
三村 由夫
田中 嗜義
森下弥三郎
最上 法二
忍
港
林 垣
小
中
員
員
委
委
東京消防庁予防部予防課
損害保険料率算定会研究
部長代理 ノ
佛竹中工務店技術研究所
研究員
委 員 小日・勝男
―-323トー
(2η )
1は じ め に
建設省には総合技術開発プロジェク
トという制度があり,建設技術に関す
る主要な研究gI発課題のうち行政上重
要であり,且つ多領域にEIる総合的凛
層について,日及び民間等の緊密な薔
力のもとに計目的,組織的に研究開発
を進めることになつているが, この一
環として「住宅性能総合評価システム
の関発」というプロジェク ト研究が工
業化住宅認定制度の整備,拡充を直接
の目的として昭和484か ら認年の 5年
間に亘つて実施された。工業化住宅認
定制度というのは簡単にいえば工業化
住宅の購入者が住宅を選定する場合の
便宜のために,公の機関が嘔入者に代
つて住宅の性能の評価 (評定)を行い,
これを建設大臣の名で認定 し,表示す
る制度である。
ところで,こ の「住宅性能総合評薔
システムの開発」の中には,住宅の居
住性.安全性′耐久性′経済佳等の間
題に対応 して多くのサプテーマが設け
られたが,防火の問題も「防火安全性
能評価システムの開発」と称して安全
性に関するテーマの中に加えられたわ
けである。
一般に大規模威it高層の建築物が災
害に対 して種々の危険をはらんでいる
ことは常 注々目され,ま た折あるごと
に指債されるところであり。最近まで
の建築防人のOF究 も主としてこのよう
な大規模.高層の建物を対象としてき
たように思われるe然 しながら住宅の
ように小規模な建築物も。それなりに
防災上不利な点を多く含んでいること
にはもつと注意が向けられて然るべき
である。住宅は例えば多くが燃焼 し易
い木質の材料でつくられ。多くの可燃
佳家具を収納 し,炊事,崚房.入浴等
のために火気が多く使用され, また老
人。子供.痛人も含めてきまざまな人
々が寝起 している。加えて小規模なた
めに感知,消火設備等の設置や種々の
法規制が馴染みにくい等の事情 もある。
事実このような事情を反映 してか。わ
が国の建築火災による人的,物的被害
の中に住宅火災のそれらが占める比率
は非常に高いのである。
‐方。このような住宅火災による長
害を如何にすれば有効に減少させるこ
とができるかといつたような問題には,
最近徐々に関心が向きつつあるものの
未だ十分でなく.一体損害の減少を目
ることが現実問題として可能か否かと
いつたこと自体。余り良い見洒 しが得
られていない.
今回の研究「防火安全性能評信方法
の開発」は主として昭和504~認年の
3年間に各界の有裁者による審議の形
でとりまとめて頂いたものである.こ
れは,或は失礼に当る面があるかも知
れないが,ま だ初歩的な段臓にあり,
上記のような問題に対 しても十分な解
答を与えられる段倍に至つていない。
然 し本研究は住宅の防火安全性の問題
に組織的に取組んだ初めての試みとし
て評領できるように思 う。門題は広範
で多岐茫洋としている上に初めての試
みのため試行錯誤に■やす時間も多く
て,必ず しも当初意気込んだl■ どの成
果は得られなかつたかも知れないが,
今回揮起 した問題や開発された成果は,
たたき台としてより良い防火対策に奮
するところが大きいと思われる。筆者
は本研究の期間中,建設省
"幹事の一
人としての任にあつたので,その開発
内容の概略を示 して■者の批判を請い
たtヽ と思 う.
2 住 宅 防 火安 全 性 能 評 価
の た め の 予 備 的 検 討
一口に・ 住宅の防火安全性能の評価″
といつても即座に具体的なイメージを
構くのは彙 しい。問題はプF常に拍象的
で漢然としているが,一方われわれに
lt具体的な評
“
方法の開発が義薔付け
られている。従つて,先ずわれわれは
住宅防火安全性能とは一体何か。現在
の技術,デ ータを基に してどの星度の
ことが可籠であるか。といつたことを
考慮 しながら研究,30発の方針を定め
なければならない.こ のために今回わ
れわれは慨略次のような立0を採つた。
1 住宅の火災倉破慶と火災の道晨性
住宅の防火安全性の.Y・価というのは
住宅の火災危険度の評価の裏遍 く表 1>火災フェーズと秘
しの表現である。火災危険度 と
い う語は火災により損害の発生
する恐れの程度′より具体的に
は損害発生の■率咸は期待億を
意味 していると考えられる.但し特に住宅火災では損害として
人的なもの,つまり死傷者の問
層がかなり大きな比■を持つ.
結局`あ る条件を有する住宅が
与えられたとき,その住宅で火
災により人的′物的損害の出る
確率或は期待億を求めることが
できるとしたら,これがわれわれの最
終日標ということができよう.
然 し,これは最初の試みとしては明
らかに腱しい問題であるから,こ こで
は先ず第 1ス テップとして住宅の火災
の遺晨のし易さ (火災進展佳)を評薔
し,火災jt険度(裏返せば防火安全佳)
の評価に代えることにする。これは次
のような理由による.即ち,火災損害
の発生には■然的なものも含めて多く
の要因が00与 しているから,火災事例
を個別に猥察すれば火災自体はそれl■
どでなくても思いがけない損害が出る
例もあるには違いないが,全体として
使計的にみれば火災が進むほど,物的
損害は勿論人的損害も出易くなること
が推測される.大規模な火災は小規模
な段階を経て成長するのであるから`
如何なる損害があつても小規模な段僣
でとどまつた火災がより危険であるこ
との積極的な理由は考えられない。従
つて火災の進展の程度が損害の程度を
定め, また火災の進展のし易さが損害
の発生の し易さ′即ち火災危険度を支
配すると考えてほね支障はないであろ
う.
2 火災のフェーズ
ところで,このような火災の進展性
を評価 しようとする場合,先ず火災の
進展程度を表わす尺度を通切に定める
必要がある。若 しこの尺度として火災
損害を採ればここでの議論は堂々巡り
に陥 る恐れがあろう.す く・に思いつく
尺度としては燃焼面積などがあげられ
るが,実は建築火災の場合にはより遍
切な尺度が導入できる。周知のように
建築火災は成長が継続すると途中で何
度か様相を一変し途中最つかの特徴的
燃焼性状を覆出するのが一般である。
従つてこれらの特徴的盤熾佳状に対応
する段階を各々火災遺晨程度の尺度に
とれば良いのである。今回はこれらを
火災のフェーズと称 し,夫々表 1の よ
うに定義と名称を与えた。 このように
することにより,火災の進展はある火
災フェーズから次の火災フェーズヘの
層移として捉えらltる ことになつた。
IL
デ
ー
ーまた火災のこ晨性は火災フェーズの遍
移薔事と考えることも可能である。
3 住宅鶴火安全性菫欝●の体薬
「住宅の防火安全性能」lt当 初漢然 と
した概念であつたが,上のように して
火災危険度を火災の過展性で代薔評僣
することにし,更に火災フェーズとい
う概念を導入 したことによりかなり合
理的に整理することができた.即ち,
ある一つの火災フェーズから次のフェ
ーズヘ遍移が生することは,対応する
段階の火災の進展を意味 しているが,
それは火災損害の出易さを支屁するか
ら,若 しある住宅があるフェーズの選
移を助長する性質を有するなら,それ
は住宅の火災危険の一つということが
でき,逆にフェーズの選移を抑制する
性質を有するなら防火安全性能の一つ
とい うことができる。例えば火源曇燎
フェーズから室内局部燃焼フェーズヘ
の選移は出火という一つの火災の進展
であり,その選移のし易きは出火危険
度.遷移を抑制する住宅の性質は防火
安全性能である。同様にして各フェー
ズの遍移に仕した名称はやはり表 1に
示 している。
以 卜のようなユニに基づけば住宅の
防火安全性能の評価システム:=長 1に
列挙された四つの火災フェーズの選移
を抑制する性能を軸にすべきであると
考えられるが,われわれはこれに避盤
安全佳能を加えた。これは明億な形で
は如何なる火災フェーズの選移にも関
係 してはいないが,人命安全上重要で
あるとの判断で特に加えられた性籠で
ある。従つて結局今回の住宅防火安全
佳籠評薔システムは五つのサプ評
“
シ
ステムより構成さオLる .即ち出火。初
期拡大,層彙拡大.類熾の緒防止佳籠
及び最羞安全性能である。
次にIt性能の評領の基になる政饉と
しては如何ようなものが望ましいかと
いう問麺であるが,各サブ評
"シ
ステ
ムが概ね火災フェーズの選移に対する
性能を評価するものであることを寺え
れば,火災フェーズのE移確率或は遷
移のクリティカルな条件を与えるもの
住宅の防火安全性の評● 口中暉凛 P●●● 39
L初日火薄な フェイズヨ人人の原因とをう可曖性のある火凛″犠歳 していう陵層
`全住戸1邸睫フェイズ●
た111孵と島重棒泉騒
|
2室内局二颯彙フェイズこ火災室内で=墨 してい
大卒13未 薔火
ゝ亀び 毬彙 フェイズ●出大馬:ti:晨 ι
人で,籠tt戸
∈0)
:.ll.1
7:可燃物係数
然る後,これらのパラメータ
の積.印ち
κ≡α.p.r
をつくり,こ の値を使つて出
火防止性能を評●する.目 1
はこの考え方を目解 したもの
である.こ こで ●.′′rは
対応する要因の平均的状態か
らの隔りを表わす性格のパラ
メータであるからκも同様の
性格を有することになり,在
来住宅に対する平均値は 1∞
となる。問題は α,′,7の定め方であるが, これは各々
E(■)‐ 1∞
rlllヽ
ヽ
)ヽ
1うヽ1‥‥IL
安全
(2崚)|平均的‖:
(3a)皇鬱考籠κ,1(ゆ
`4a)
(5“ )
く日 1>出火危険度評価の流れ日
などが理想的であろう.然 し現在のと
ころ任意の住宅における火災のフェー
ズの選移の薔率やクリティカルな条件
を求めることなど不可能といつてよい.
従つて上のような数値に基づいて評価
が行われることは将来の課題であり.
そのための努力は必要ではあるが,現
段階のそれがただ火災フェーズの選移
を抑制する上で平均的なものより有利
であるか不利であるかといつた傾向を
示すだけのものにとどまつていたとし
ても,ま た止むを得ないことであろう.
3各 性 能 別 評 価 方 法
今回開発された防火安全性能評価シ
ステムは既に述べたように五つのサブ
評価システムよりなる。以下ではこれ
らの概要を述べるが,実 lt夫々主担当
者が異るため手法に一貫性を欠 く嫌い
がある。燃 し今回の評価システムlt第
1ス テップと考えれば′異る人の多様
の考え方が盛 り込まれている方が,逆に興味深いともいたよう.
1 出火防止性能
(1)諄饉の考え方
一般に火災は,火気の熱が何らかの
興機で可燃物に伝活 され,その着火を
もたらすために発生する.従つて評任
方法の中にも。この「火知 「可燃物」
「契崚」に00す る要素が盛り込まれる
ことが望ましい.ま たこれらは室の用
造により大豊のあるものであるから,
拝
"は用途の異る空間毎に行 うのが良
い.
②評価方法
先ず「火知 「可燃物」及び「興繊」
の三者は実際には互いに■雑な関連を
有すると思われるが, ここでlt単に次
の三つのパラメータ ●,′`ア
を導入
し,これらが各々独立に三者の条件を
代表 し得るものと考える.
α:火気スコア
′:管理係数
次のようにする.
0火気スコアα
住宅内で使用 される種々の火気とし
ては,住宅の型式 設計仕様等で定ま
つているものもあれば,居住者自身が
好みに応 じて持込むものもあろう.その際全体として火災を起 しにくい火気
が使用される結果となつている住宅ほ
ど出火防止上有利であることはいうま
でもない.火気スコアは住宅内で使用
されるこのような火気の トータルでの
出火させ易さを表わすバラメータであ
り,評価対象室プを指定 して次のよう
に定義される.
",=ofi;-,*rm (1.1)
ここに O′ は椰領対象室 プの年間出火
事 (件 /年・菫).F′ は室 ブと同用途
の室における年間全出火件数 (件 /年 )
また ″J It室 プの全敗 (統計サンプル
数)である。ところで具体的な評価対
象空間プの 0′ は次式で計算される.
0フ =Σ`′
″P`+0´ (1.の
ここに′"は
菫ブにおける火気 jの保
有数
“
D, 0,は 評●される火気以
外の原因 (例えば放火)による出火事
(件 ノ年・∋ ,P`:t火 気 ずによる年
間出火事 (件 ノ年 。1)である.式 (1.
2)中 の P● ●″ を火気 づからの年間
出火件敷 r`(件 ノ年). 火気 jの総
数
“
(日), 評
“
される火気以外の原
因による年間出火件数 r,(件 ノ年)
等の統II性 を使つて
P`=r`′″“o,=rヵノ″J (1.3)
のように表わし式 (1.1)を 変形すると
●J=OJX(″′/FJD x l∞
=:Σ`′
,P`(rJ/FJD+0,(″ノF】 Xl∞
=lΣ`′
“(r`/FJ)′ (Nツ″J)
+《 r"′FDI xloO (1.4)ここで各々の火気
'に対する(r`ノ F′ ),
(“ /どD及び (r"′ FD it現在のと
ころ未だ■度に幾分離点はあるものの`
適当な統計賢料を使つて予め求めてお
くこと力ヽできることに注意すれば,実
40 ,o● ● 住宅の防火安全性の静0 日中暉=
(3/)
際の評価がなされる時点で α′を求め
るためにわれわれが知るべきことは諄
伍対象空間の火気の保有個数のなとい
つてよい.
●lt理係数 ′
火気と可燃物の管理状菫lt本来出火
に影響が深い筈であるが,われわれは
諄領に当つて住宅の条件 しか知らされ
ないため,居住者の気費等にも依存す
るような管理状態は合理的に予測する
ことができない。従つて今回は′は常
に 1と する.すなわち管理状態の諄●
は行わない。
0可燃物係数 r
この係数は住宅内の可建物の状態に
関するパラメータである。可燃物とし
ては内姜材等の住宅部材と収精可燃物
とが考えられる。然 し後者は出火の着
火物として占める割合は高いが,住宅
の性能として捉えることに難点がある.
そこで今回は前者のス評債することに
し,7を不燃化事に応 じて表 2の よう
に定める.
く猥 2>可燿物縣 数
不儀化率 1 可建物係数
1級 1 095
1:1105級 1 105
0グ レーディング
既に述べたように,出火防止性能は
κの評点を基にグレード分けされる.
この結果を表 3に示 した.因みに既存
く表 3>出火防止性能のグレード
住宅はκの平均が 1∞ であるから,E存の平均住宅なみの性能が 3にランタ
付けされることになる.
2 翻露拡大防止性饉
(1磨価の考え方
初期拡大防止ttlt室 内局部燃焼フ
ェーズから全室内燃焼フェーズヘの選
移の防止に関する佳能である.住宅内
の主要な室である台所・DK及び居室
について火災事例から口べてみると次
のような火災拡大パターンがこの段階
で■著であることが解る.先ず台所・
DKでは油類が主要な第一次着火物と
なることが多く,
0"一次着火場―→船直面―→天丼
面
嚇 一次着火物― 天丼面
の二つのパターンを経ることが多い.
また居室ではより多彩となるが,総貨
類が第一次着火物となり
(a′)第一次着火物一→始直面一―天
井面
(b′ )第一次着火物一→家具一→鉛直
面′天丼面
の二つのパターンとなることが比較的
多いぃ
以 卜のようなことから, この段階の
火災拡大に内姜材の防火性が演する役
覇が大きいことが示崚される.そ こで
この段僣の性七は内贅材の防火性を中
心に諄薔することにする.居室におい
ては家具もまた初期拡大に影響が大き
いと思われるが。住宅の性能と考える
ことにはやはり困難がある.ま た同一
住宅であつても室毎に内姜は異る可能
性が高いから,本性能も室毎の評●を
行 う.
感知威は消火設備が一般の住宅に備
えられるなら,初期火災の拡大防止に
も相当有効であろう.然 し現在までの
ところわが日の住宅では設置例が少 く,
その効果の程度を知ることができるだ
けの資料が十分でないため,今回は極
めて大雑把な評価で満足する外ない。
2磨
“
方法
次式で定義されるパラメータ●を不
燃化率と称 し,住宅の室内の防火性の
程度を表わしていると入なす。
α=Σ19`S`/Σ`S`
(21)
ここに ,`:室 内表面を構成する材の
称類・性能に応 じて表 4の ように定め
られる数値
く猥 4>各邸位の不盤化僻数c`
:tnl il !.t t グレード
:不 燿 材 料
月璧内贅 |●不腱材料
仕上材料
1撃
燿
冒
け
罵
1防火戸(甲■,乙■)|
田口3口 |口 晏 戸 |等 |な「 li阜
ツシJIS4:室内の表面仕上材 Fの面積
係数 c`は内姜材,開口部材の防火
性が■ くなるに従つて大きな値を与え
られている.こ のため不燃化率が高い
ほど,菫の平均的防火性は高まり初期
拡大に有利に●くことが期待されよう.
昇る材料の配置の違いによる差などは
缶親 しているから,少 扱々いが組い恐
れもあるが.グ レー ド分けが適切なら
相当うまく現実の火災を反映させられ
ることが認められている。
一方.感力・消火散●等については
政量の有無のみで拝僣を行う.
0グ レーディング
火災の初期拡大防止性能については
内姜材の防火佳tと感力・消火設備に
分け`夫々表5,表 6の ようにグレー
ド分けする。
言時火気
使 用 菫
8以上
7以上 8未ロ
6以上 7未潤
く表 5>内装材の初期拡大防止性籠グレード
菫 別 グレード
国 2に例示するように,圧焼経路に
入る可能性のある各室には番号をつけ
るetた各室間の区画壁等は両日の空
間の番号を並べて位置を示す.
e■.■ :室 1の火盛 りな綺時田
,1.:室 1と=2の
日の区日のこ
燒風止時間
(D出火室の選定
最初の出火空間としては,玄田′暉
■ 階段。便所等出火頻度の低いとこ
ろIt避け,居室′全重′台所。浴室.
鵜戸等に限定する.
0火菫 り継餞時間 rの算定
火災室の火菫り継続時間 rは火災菫
の標準可燃物量″を燃焼速度■で餘 し
て得られる。印ち
,=″/R
更にP,″は
R=■ 5Σ`η
′`ν
π″="F+ι +″
(3.1)
(■ 2)
(■ 3)
ここに 4`:開口部面積(m3)
「・
:開口部高さ(m)●:単位面積当り標準薇載可燃物量
(kgノ m3)
F:各室床面積(m8)
ι:各室の可燃内姜重量(kg)″:各室の可燃下地重量(kg)
7:開口部の防火性能に応 じた開ロ
低減係数
0区画の延焼阻止時間 ′
区面の構成材料・構法が単一の場合
にはその種類に応 じて表 7の ように定
める.区日が2■以上の■成材よりなる場合は最も弱い要素で評
"する。
●)延焼動層
□ 2に例を,日 3に手層の流れ目を
示すように仮想出火室から始めて rと
,を比較 しながら住戸内の工燎動態を
目べてゆく。但 し圧彙を受けた室は一
く表 8>通費拡大防止佳籠のグレード
| 'rv-r25未瀾
25以上■0未瀾■0″ lo o″
1● o″ ao″a■ o2tt
1
2
3
5`以上 6未■
4未瀾
●時火気
使 用=以 外
61■上5以上 6未満4以上 5未満3以上
`未瀾
3未ロ
1
23
4
5
く衰
`>墨知・消火重日の初期拡大麟
止性籠 グレー ド
餞●刷 感知消火投備の種類
凛知器の
設置の有
麟
消火舷備
の故量の
3 騒捜拡大防止性腱
{l17・価 の考 え方
二燎拡大防止佳能は全室内燃掟フェ
ーズから全住声燃彙フェーズヘの選移
の防止に関する性能である.
住戸内の一室で出火 した火災がその
室内だけで鎮火するか,或は室外へ居
焼するかIt,室 内の火災性状と室区画
の性能との関係で決まる.両者の比較
は一般には凛雑であるが,い ま若 し簡
単に火災性状を火盛り継続時間 rで ,
室区日の性能を延燎阻止時間′で代表
させ得るとすれば,両者の時間の比較
の問題に簡略化される。即ち火災室と
その口室のうちのあるものとの間の区
日の二崚阻止時間夕が,火菫り義侵時
間 rょ り大きければ工焼せず,逆の場
合には延焼する。
次に圧綸を受けた室をその時点で新
たに火災室とし,更にその周ウヘの圧
境を考えるという手順を繰返せば住戸
全体での延燒拡大の動態を評
“
できる。
12)評価方法
0■.部位の符号
熱・題=知
轟併用
燿 愚 知=熱 凛 知 彗 ・
た し
1
2
3
`1
2
3
4
5
一夕
橙
水
〓
し
か
火
畑
火
ス
消
防
消
な
有缶
3
司 F
ι :
: :
く図 2>延腱動態の算定例
ωυ
く表 7>区面構成材・構法のこ幾働止性壼
試■の有無 覆 用 用 二 彙 阻 止 時 口'(分
)
試験による場合
区■壼
醸 火 構 遭
コウリ ト ■火時間+"
"分合格 50
10分合格 "
遺
等
構
目蜃
火
菫
譴
土●
コ
口 ■
甲 ロ
乙
"Hガ ラス戸。木襲戸
薔
”
”
Ю
3
換気ロ
目 ■ 式劉口o
■
饉
億なの
ロ
ア
そ
■ 呻 田 放裏百月日 lm以内に,火物有 10
″ 無 ‐
試験によら
ない場合
石 青 ポ ー ド
厚15mm以上
12mm以上15mm未 ヨ9mm以上12mm未 ヨ9mm未 満
コ
a
・5
5
口継強 fヒ タイプ石膏ボー ド
石 組 ス レ ー ト
石
“
セメン トケイカル板
合 板
石青ボー ドの場合の'+5
石青ポー ドの場合の'-5
グラスクール・●ッタウール 厚150mm以上 lo
5mm以 上mmmtttn s
プラステック発池材 0
瞬に して火盛りに連するものとする。
圧燎勁態の追跡が終つたら,次式で定
義される延崚拡大係数εを計算する。
`=И/β
ここに
4≡辮
X爛 :圧崚
面積率
β :歴焼拡大終了時間 (分)
13)グ レーディング
圧焼拡大防止性能は rの儀によつて
表 8の ようにグレーディングする。
4■ 燿防止性
籠
0)諄個 の考え
方
ここで類焼と
いうのは,一つ
の住戸から他の
住戸への圧焼拡
大のことである。
凛崚の原因とし
て支配的なもの
は熱輌射である
から,層焼危険
は炎上家屋から
の僣射放熱L僣射を受ける側
の住宅の外月部
防火性能及び百
者の最何学的関
係に依存する.
燃 し,こ の中で
住宅の性能とし
て評価すべまも
のは外月部防火性のみでぁる.
12)評価方法
住宅外月部の構成材を現行法規の規
定を基準に級別 し,級別された各々が
雇燎防止条件を清足するために必要な
形態係数を並記 した.因みに並紀され
た形態係数に応 じた隣枚間隔を求める
に:t次のようにすれirよ い。
(i)火 炎の形状・寸法
隣棟が炎上 した場合の火炎形状を計
第の面易のため,長方形と仮定する。
火炎僣●lt炎上家屋の見付僣とし,こ
住宅の防火安全性の諄優 田中暉= ,G90 41
:● 火災室 とetど ,1^=
護鯉皐肝を・ 咸十う●1〔
PIr4(1‐ 総てのには||)の=t
な'曹
imFふ 禽
爆鷺11ふ謹書を'llろ =彗崚雄火百嬌 A
く表 9>類焼防止性能のグレー ド
部位 一 部 位 性 能グレ 1延 焼部分への― ド 1使 用 の 可 否
必要隣薇FD隔 (下饉)
裸未造 1裸ホ造以外
外壺
・軒裏
可
可
可
部
否一
可
可
可
部
否一
■
根
関
口
部
防火構造 よ り優れ ている
防火構造
土塗壁同等輌遺
土塗壁同等輌造 より劣つてぃる
≦030 ≦●55
≦0∞ ≦035S● 10 ≦α15
≦Q“ ≦010
甲腱防火戸ょ り優れてぃる
甲■防火戸
乙橿防火戸
乙籠防火戸よ り劣つている
可 ≦050 ≦055可 ≦030 ≦035可 SO.10 ≦α 15
(一部可) ≦005 `● 10
防火構造 よ り優れている
防火構造
警材が法定不燿材料葺材が法定不懺材料以外
≦Q50 ≦055≦030 ≦035≦010 ≦●15
≦0∝ ≦010
さ″は次のように算定する。
裸木造の場合
″=:X(全 見付面積)
裸木造以外
″=与k星根.開口部等燃え抜けが
予測される部分の面積和)
(二 )必 要隣棟間隔の算定
上のようにして火炎の形状が定まれ
ば,形態係数の適当な計算式或は計算
図表を使用 して必要な形態係数に応 じ
た隣棟間隔を求められる。
13)グ レーディング
住宅のタト周部構成材の表 9に示すよ
うなグレーディングを類焼防止性能の
評価とする。
3 遍“
安全性籠 |(1)評価の考え方 |われわれは租々の調査費科によつて,
住宅火災時の緒問題を部分的に知るこ
とができる。例えば.住宅火災による
死者は建物火災全体に対 して,F常 に高
い比率を占めること,就寝中の事故や。
老人.幼児の事故が多いことなどであ
る.従つてわれわれが住宅の選難安全
の問題を考える場合, これらの特質を
等閑視できないことはいうまでもない=
然 しながら一方でまた。われわれはこ
の住宅における選難の問題に本格的に
取組もうとする場合,基礎的なデータ
が絶対的に不足 しているという事実に
も直面せぎるを得ないのである。住宅
火災は全国で毎年 2万件近 くも発生 し
ているから,住宅火災に遭遇する人も
相当数にの`=る
はずであるが,それら
の人々が火災に対 して如何ように働き
かけ,ま た避難 しようとしたかなどに
ついて組織的に調べられたデータは乏
しい.こ のため, この問題に対するわ
れわれの知識のうちには,類推や想像
の城を著 しくは出ないものも多く 翻つて考えてなれば,われわれの現在の
知識で最饉安全性能の評価を行おうと
すること自体本質的に無理であるとも
いえるのである。従つて今回:ま この住
宅火災時の選燿の詳細に分析的に立入
ることは最け,平面計画などから判断
される選難経路に基づいた口便な評価
をすることにする.
121「価方法
避難安全性能の評価は,後に示すと
ころの所定の手順に従つて算出された
・ 選難経路の信頼度″の値に基づいて
行 う。但 しここでの い信頼度
″は確率
で定義されるような厳密な量ではなく,
ただ安全らしきの程度についての序列
を表わすだけの便宜的な尺度に過ぎな
いから,信頼度の値が 2倍でも, 2倍
安全というわけでなく,よ り安全そ う
だという意味にとるのである。
また避難の信頼度は同一住宅内でも
例えば 1階の室と2階の室では大きく
昇ることが予想されるので評価は各室
毎に行うものとする。
■難経路の信頼度の算定手順は以下
の選りである。 1(1)住戸内の各
=`中間城及び安全城
をノッドと考え,43籟度を算定しよう |
とする=か
ら安全城までの経路を目 4
のように全て網墨する。但 し,安全塊 ‐
中間域等は次のように定義する. |危険城 :火災発生の自′火炎や煙に
■われるIL険のある住戸内空間。 |安全城 :原則として地上等の最終的
■難先。但 し集合住宅の隣戸領域等, |
実情に応 じて解釈できる。 |中間域 :危険域と安全域との中間に
あるパルコニー, 1階部分屋上等の薇
衝城.但 し, これは整理のための呼称
であリノフドとしての役割は室内各室
と同様。 :(二 )各 ノッド間の個別の経路の信頼魔
避=経
路の ネッ トワー クく園
`>避腱経路 の ネッ トワー タの例
Po●● 住宅の防火安全性の諄饉 田中暉晨
くよ'0>閣
別II路 信恒度
,“
日
獣 百鰤
平 面僣 m曖 ● ■
菫留腱く
,―口翻鳩ζ
:ロロ T ~L「
―
エ
カー'レ
諄′つても● In・・
22依 引遣い●
II2餞
'1■ド7
″ある●3t●Hl可
3.1枚 ドア
王 制,■
,共
″い
′‘可
ヒ引用
4.臓餞 ヽ 1
麻r〓薇一‥―〓●攣こ
: 2枚 II● い戸
副L] α8メ (― ツ)|― :h,r'
2手綱付2枚引違い●
麟凛] トリ)←―:
3.1枚 ;i
■ ]・型 |
|
工
年
■
〓
〓
1 バ jL,■ ― HIIII
』
‐
“
b
2フ ラ ット電級
k口 lr雫大なら力
|
3傾斜薔僣
ヽ 国トエのうちス
=うカ
|(1-城)1-unθ
ヽ電
1 ・・●コニーから■|●地「 |
へ
||
ドこのうち人なる方
|い 1-:)
|
摯
磐
E
〓
日ア、聾
― ―
下記′)う ら人なる方
108メ (1-})
3 '` ル=ニ
ーカらF● ●
E キ|
11■ rt ●6
■,' 1.●
フウ ト暉根か|■ ,t」1′ヽ 町1ト 1誰Triソ |
嵌,
′、′地
フ韓
らn段 を ,‖ rl
子‖颯
6こそ艶i興
ロカらに't 味。ED
7傾斜口観から直織,,上ヘ
‰方
滋■
”嚇
儡別理路 僣籟度
′′′′if菫=;,I喜
:ミミミ1`ヽ
(3主)
を表10か ら引用 して与える。
0対象とする空間から安全壌へ至る
こ彙経路のネットワークを解き,全体
での信頼度を算定する.これは僣別経
路要素 j′ ′の信頼度が ■,7Jであ
るとき,合成の信頼皮 島フは
は ず,ノ が直列のとき
R“ =71・′′
0経路 ど,′ が並列のとき
′`′
=1-(1-74)(1-均 )
であることを考慮 しながら□ 5の よう
に して順次計算を遺める。但 し計算の
日便さのため,各 ノ,ド間の経路は一
方向 しか経由 しないものとする.
0グ レーディング
算定された「量難経路のネットワー
クの信頼度」を,表 11の ように級別す
る。一般に 1階の各室では信頼度の饉
が高 くなり0.9以上くらいにr・●るが 2
階の各室では低くなリパルヨニーや庇
に出られる場合には0.5■ 庁′それら
が全 くないものでti a 2以下になるこ
ともある。
く衣11>避難安全性能のグレー ド
遷難経路の信頼度 | グレー ド
1 ~0808-06●●‐α
`α4-0202~ 0
1
2
3
4
5
4ま と め
今回の住宅防火安全性能評価システ
ムの開発研究では,火災の一連の成長
過程を幾つかに分割 して各々火災のフ
ェーズとし,火災の遺晨はこれらの火
災フェーズの選移であると考えた。こ
の結果,住宅の防火安全性能は,これ
らの各々の火災のフェーズの選移を防
止する性能の総合として考えられるこ
とになつれ
この経彗を考えるなら,これらの性
能の評鋼の基礎となるデータとしては,
本来火災フェーズの選移確率或は層移
の限界条件などの意味を持つ数値が望
ましいであろう。現在のところ。これ
らの数値を評価に取入れられるl■ ど研
究やデータの書積はないので,今回の
諄● システムlt全体的にみて,各性籠
に対する任意の住宅の優劣を示す程度
の段階にあると考えられるが,将^的なより良い評領システムを考慮 した研
究。データの蓄積が望まれる。
この目的のために本研究の範囲内で
行つた研究も幾つかあるが,任面の椰
合で省略 し,最後に研究開発に携わつ
て頂いた諸氏の名を記 して御苦労に■
したいと思 う。
く目 5>選菫経路 ネ ッ トワータの信籟度のII“例
菫」●ロ ー火安― ■■
' 1...‖」見:鷺
晏二K二轟1, |・メ栞量: nH警
庁 |, ■鳳■― 菫東大手 |
: 末:。1 1:霊摯讐率纂定倉
|″ , 3久 自治●清
"●究所 .
0 ● ■●拿[.三 村由夫.●上餞牢。日
■●■.■下弥~8(鳳
上 菫餃`ヨ饉研究所)
● ,■ 艘
43
“
建・
椰 :Ct―“
システム
●●兒―安全性に目する辞●・・A晨●ヨ定法の口,
(,火安全性経の
“
●方法の
“
■). ●力前年,
疇 .253年 .′ く
=餃省菫藁研究所防災研究室長>
住宅の騎火安全性の鮮饉 口中暉晨
喫 )
:I-5
Developnent of Eval uationProtection Performance of
l4ethod for FireIlwel I ing Houses
TAKEYOSHI TANAKA, Dr. Eng.
Research MeuberBuilding Research InstituteMinistry of Construction
fbaraki-Pref. Japan
1. INTRODUCTION
This study is to develop an evaluation system which can eval-uate objectively fire protection performance of prefabricated dwel-ling houses and inform- the authorized results to prospective con-sumers so that they can wisely choose their houses-
There have been few systematic studies centered on developmentof evaluation method for fire protection performance of dwellinghouses up to now, but the accumulation of study and data on relatedfield wai e*pected not so little. Then we decided to develop theevaluation slstem in principte by applying the existing studies anddata, while complemental investigations and experiments were car-ried out in case they are necessary.
This study is conducted for four years, fiscal L974 L977,as a part of a-project of B.R.f., Development of Evaluation Systemfor Total Performance of Drellings, in fiscal year 1973 L977.
The outline of the course of the four year study is shown inFig. 1. In the first fiscal year L974, the study was conducted bythe B.R.f. members as a direclly iontrolled study, but in the fol-lowing fiscal year 1975 L977, it was placed !n charge to theBuilding Centei of Japan (B.C.J.) and Fire Protection PerformanceComrnittee organized in B.C.J. took over the study:
Fire Protection Performance Committee (B.C-J- )
chairman Kunio KAWACOE′ Science UniversitY of TokYo
Members Takashi ASAI.{I, Tokyo Fire Departmento shinobu KoBAYASHI, I' !{inato NAI(AGAI(I, Fire and trlarine Insurance
Rating Association of JaPanKatsuo oGttNI, Takenaka construction companyShin-ichi SUGA9|ARA' Tokyo University
' Tadahisa JIN, Fire Research InstituteBЧilding Research lnstitute
Π-59 (3三:
ltanager Takao $INU{Mi\TSU,
l{anager Yoshio MIMURA , Building Research Instituten TakeYoshi TANAIG, '" Yasaburo MORISHITA, 'i Koji !{oGAllrr rr
2. PRELIII{INARY DISCUSSIONS FOR FIRE PROTECTION PERFORI'IANCE OF
DWELLING HOUSES
2.L Phase of FireTo evaluate the fire protection performance of dwellings is_r_in
other words, to evaluate the fire risk of them. The "fire risk"can be interpreted more concretely as the probability or tle ex-.pected value-of loss we suffer by fire, while the loss of lives isiarticularly important in case oi home fires. Therefore, it may beiaid that to eslinrate probability or expected value of damages andcasualties that we may suffer by fires while llving ln a given Pre-fabricated house'is a final aim of such a study as this. But atpresent, to attain such a aim is, however, undoubtedly- very diffi-lultr so here we consider about fire spreadability in houses.Since fire loss is dependent on various factors in fire' among manycases, there must be some srnall fires that happen to suffer an un-expectedly large loss. But amount of loss in usual cases is ex-pelted to be almost proportional to the extent of fire spread,including loss of lives as well as ProPerties. In other words,extent oi spread of a fire generally determines the amount of theloss by the fire, and fire spreadability in a dwelling house-deter-mines Lhe probability of loss by fire, i.e. the fire risk, of thehouse.
When ne try to evaluat.e the fire spreadability in a dwellinghouse, what we need first of all is to introduce a certain aPPro:priate measure to represent the extent of fire spread. An ordinaryexample of such a melsure will be burnt area. However, since it iswell- known that building fires abruptly change the asPect of theburning several times in the course of whole growth, to use someother measure corresponding to the aspects of the burning is moresuitable and convenient.
In order to obtain sucha reasurer w€ divide the whole Processof building fire into some stages each of which is named as shownin Tab. I iccording to the burning characteristics. Thus, thegrowth of building fire can be reduced to the transitions of firephases, and spreadability of fire can be interpreted as the proba-bility of the transition of the phases of fire2.2 Structure of Evaluation
Owing to introduction of the idea of fire phase as the measureof extent of fire spread, "Fire Prcitection Perfortnance', which hadbeen an obscure concept at the beginning of the study, was inter-preted in fairly reasonable way. Since a transition from one phaseto the next phase lnplies a kind of fire spread, the properties tofoster the transitlon of the fire phase comprises the fire hazardfor the corresponding stage, and conversely the properties toprevent the transition comprises one of the fire protection perfor-mances of the dwellings. For exa.mpler the transition from the phaseof fire origin burning to the phase of partial room burning is akind of fLre spread, which is usually called outbreak of fire.Therefore, the properties to foster or prevent this transition canbe regarded as the hazard or protection performance of outbreak offire respectively. The rest fire transitions likewise mentioned
I-60 (361
and named in this study are also shown in Tab. l.It may be said that, in principle, an evaluation system for
fire protection performance of dwelling houses should be based onthe four performances, which are derived fron the transitions offire phases mentioned in Tab. 1. However, the present evaluationsystem as shown in Fig. 2 includes another performance on humanevacuation, which has no direct relation with any transitlon offire. This evaluation subsystem was added to total evaluation sys.tem for its importance was recognized by Fire Protection PerfornanclCommittee.
The next thing we have to do is to examine what kinds of valuetare desirable as the bases of ttre evaluation. Since the most per-formances constituting the total evaluation system have their ori-gins in the respective transitions of fire phases, some values togive informations on the probabilities or the critical conditionsof the fire phase transitions will be perfect. But the fact is,none of us can possibly tell the probability or the critical con-dition of a transition of fire phase in an arbitrary dwellinghouse. Therefore, admitting the importance of such bases for afuture evaluation system, it is unavoidable that the evaluation syrtem here is still in so early stage that we can only know a certainhouse is comparatively superior to another but we cannot know howmuch it is superior.3. EVATUATION !,[ETHOD FOR RESPECTIVE PERFORI.{ANCE
As already referred to, the present evaluation system for fireprotection performance of dwelling houses is composed of five sub-systems for respective fire protection performances. And the fourof them are related to the corresponding transition of fire phaseand the rest to human evacuation. The following is the evaluationmethods for the five respective performances.3.1 Prevention Performance against Outbreak of Fire
(1) ViewpointsGenerally, outbreak of fire in dwelling houses mostly result
from that fire-using appliances get chances to transfer hazardousheat to combustible-malerials and ignite them. Therdfore, in prin-ciple, the evaluation system of prevention performance against out'break of fire should include the three factors, i.e. fire-usingappliances, combustible materials and chances connecting the bothttro. And since rooms are expected to differ in the condition onthe three factors according to the usage of themr w€ should eval-uate the performance for every room with different usage.
(21 ProcedureIn fact, the above three factors are considered so intracted
that we have not known well about the mutual relation. Then tosimplify the problem, we introduce the following three parametersc, B and y, and regard them independently represent the conditionsof the three factors.
c : fire-using appliances scoret : arrangement factory : combustible material factor
Thus, the prevention performance against outlreak of fire is eval-uated by grading the values of K defined as follows as the product
I-61t37)
Of the three parameters.
K = 。・β・γ
The outline of the procedure of the evaluation is shoWn in .Fige 3。 The values of ●′ β and γ are determined as f0110Ws.
(1) fire― using appliances scOre
llx::eto th: 1::lli:: t::S:ime ofn the house afterward by the user。he leSS hazardous appliances on
:i[:L姜:薫
露 :E::::鮮 ::言 i:壼 奎 重 :華 奪 電 :::iti:ii:m
data.
十
■
¨ は … … ……
川
Where 38 name Of the roOm under evaluatiOn
Qj= the
Fj8 the
fire occurrence rate in room jnumber of fire in rooms of the same use with j
Mj t the total nr-rmber of
while Qi is calculated as
Qj=12jiPi+。 jo・ ………・……・…・…・…………・… :00・ … (■-2)
wiere 2..8 the number of fire― using appliance i in rOom j〕ユ
P■ 8 th:e ile occurrence rate CWing to fire―using appli―
and ag,in, Pi′ Qjo are
Pi ヨ TiノNi and Qj6 〒 Tj。/MjO・ 。・・・●●0● ●000● 00。 (■~3)
where Ti : the number of fire owing to fire-using appliance iNi 8 the tOtal number of app■iance i
‐・=_Tj。8:::st:t:th:賞『二
°f fire in room j∝ ing tO then the appliances selected as the
=_ ■ 1・ :_eva■ uation factor
Thこi′ rewriting Eq. (■―■)by uSe of ttqs。 (■-2)and (■ -3)′
。 CJ = Qj X (Mj/Fj) x ■00
8 [三 二jiPi〔lj/Fj) + P。 (Mj/Fj)] X ■00
‐鴨i x■岩 +(十月X Ю∝……口
I-62
rooms of the same use with i
(38ゝ
Therefore′ substituting
Si = Mj/Ni′
into Eq. (■-4)gives the
Si, Wi which defined as
r{i r ti/tinext expression
Cj =〔ItjiSiWi +(Ъ♂j)]X100。 …………………………0(■ ~5)
ft must be noted that the values of S; , lf{ and T*^,/F.r can beobtained in advance from appropriate statiSticir datai" rTherefore what we,have only-to-do in the evaluation for the perfonmance of a room is to count up the nr:mber of each fire-usinq appli-ance in the room. The 51 and-w1 shownin Tab. 3 ire-ttieGiif,plEt-for houses of exclusive use of living.
(ii) Arrangement factorArrangement factor B |s a parameter to represent the arrange-
ment conditions of fire-using appliances and combtrstible materiitsin dwellings, which is suppoied Lo have much influence on outbreakof fire. However, the condition largely depends dn various factorse.9. userrs temper which can hardly be iegaiaea as performances ofdwellings.Therefore, here we consider B as always unity, in other words,
hte don't evaluate the chances that connect fire-using applianceswith combustible materials.(iii) Combustible material factorThe factor y is a parameter to represent the conditions oflive and dead combustible materials in-rooms, which are very impor-tant on outbreak of fire. But we determined the value of y onryfrom interior finish as shown in Tab. 3, because we have few meinsto evaluate conditions of live combustibles.
(3) Grading
- Tt-re prevention performance of outbreak of fire is graded asshown in Tab. 4 according to tlie value of K. Because eich of theparameters o, B and y is a parameter that indicate a distance fromthe average state of the corresponding factor, K as well wears sucha characteristic and average of K for usual existing dwellingsbecomes 100.3.2 Prevention Performance against Initial Fire Spread
(1) ViewpointPrevention performance against initial fire spread is relatedto the transition from the phase of partial room burning to the
phase of whole room burning.: Statistical Investigation into fire spread at this stage makes
us known that in kitchen room (K) and dining kitchen room (DK), thefollowing two fire spread patterns are dominant,
(a) tnitial i,gnited material - interior wall - ceiling(b) initial ignited material - ceiling
and that in living or bed room the next two are dominant,(a) initiat ignited material interior wall ceiling(b) initlal ignited material - furniture - interior wall or
ceiling
I-63t[37 )
Tbelefore, it is supposed that interior finish material will Playan l.portant role tL-tirU spread at this stage.- While furniture issuppoled to play an signiflcant role too in living roomsr we neg-feii their elfelts because to regard furniture as an inherent per-formance of a house is not always adequate-
If equipments to detect and,/or extinguish fire_are installedin usual tiomi houses, they witl work very effectively. But we havehad so little data on their effectiveness in home houses that wehave to be satisfied with a rough evaluation.
12, Procedure(i) Incombustibility number of interior finishWe call the parameter c defined by Eq. (2-1) r "fncombustibil-
ity ngmber" of inlerior finish and regard it as rePresenting theexlent of fire protectiveness of interior finish of a room.
' c = [igi51/[isi .o....o........o...............q....(2-l)where gi : number specified as in Tab. 6 according to fire
' ptotection performance of material :Si : area of used interior finish material i
,'' ..i.r,Since the better fire protection performance a material has,
the larger value gi is given as in fab..5, generally.speaking, -thehigher a of a room-is, the more protective the room is expected tobecome against the initial fire spread. Though such a treatmentby incombustibility number, which is simple average of materialsw-ith different fire protectiveness, is somewhat rough, the numbercan be expected to reflect the fire protection performance of aroom in actual fire by properly grading the number.
(ii) Equipments for detection and extinguishment of fireThe evaluation on equipments to detect or extinguish fire is
done simply by whether they are installed or not in a house.(3) Grading
Grading on the prevention performance against initial firespread is shown in Tab. 6 and 7 respectively on interior finishand devices.3.3 Prevention Performance against Interroom Fire Spread
(1) Viewpoint' Prevention Performance of interroom fire spread is related to
the transition from the phase of whole room burning to the phaseof whole house burning.
Generally, rhether a fire stops in originated room or sPreadout of the roorn depends on the relation between the severity of theflre and the fire protection perfornance of the room partitions.The fact lg, the relation is very complicated, but if hte practical-ly substitute fire duration t for severity of fire, and fire spreaddelay tine by partition p for fire protection perfornance of parti-tion, ttre relation can be reduced to the simple comparison bet*eenthe two times t and p.
(2, Procedure(i) Numbering to rooms and partitions
I-64(4つ )
As shown in Fig. {, all the chief rooms in the object at"gl-ing are hurnbered, and the partition between any two rooms is identified by a pair of the numbers of the two rooms.
ex. t I : fire duration of room 1prz! fire spread delay tine of the partition between
room I and 2
(ii) Selection of fire origin room
As a room of fire origin, the nain rooms of the object dwel-ling such as living room, bed roonr, dining- room, kitchen roomr bathroori or closet room must be selected but the sPaces such aa entrancehall, cooridor or staircase etc. should be excluded since the fireoccurrence rate in those space is expected loro
(iii) Estimation of fire durationThe fire duration time t of any room is estimated as follotrs.
R - 5.5fn1A1 1rL.......o.............................(3-2)W = wF + L + N ..................................(3-3)
where R: burning rate kg/slt{: standard fire load (kg)A: opening area (m)
H: opening height (m)
n: reduction rate of opening area due to fire protec-tion performance of oPening
w: standard live fire load per unit floor area kg/n2lF: floor area of the room (m2)
L(= [rll,i): effecfiy" weight of combustible interior finishmaterials
N(= [6iNi): effectlve weight of combustible backing
In the above expressiOn, r1r Y and 6 are reduction rate that arespecified as shown in Tab. 8.
(iv) Fire spread detay tirne by a partion P
For a partition made of simple cpmponent P- is given as shownin Tab . 9, irnife for a partition made"of more than two conponents,P is evaluated in regard to the weakest component.
.. - (v) Eatimation of behavior of fire in whole room
As illustrated in Fig. 5, starting from the room of fire ori-gin and regarding the rooi to which fiie spread as notf fire roon,i,he betravi5r of iire over whole house is estimated. Needless tositr when p of the partition between fire room and a neighbor rmissn;iter than 11 firl spread out to the room but conversely when P
ie larger than t1 the iire doesnrt spread to the room.
After examining the fire behavior, we calculated C using thefinal burnt'area and the time taken to the end of the fire.
A= the final burnt area burnt area ratiototal area of the dwelling
B: time taken to the end of fire spread(3) Grading
The prevention performance against interroom fire spread isgraded as- shown in tlU. 10 r according to the value of C. As the C
ior the evaluation, the small,est among those Cs for all roong isadopted.3.{ prevention Performance against Interhouse Fire Spread
(1) Viewpointt{hat is called here "ilterhouse fire spread" means the fire
spread from a house of fire origin to the neighbor houses. Sincetf,e most controlling cause of interhouse fire spread is thermalradiation, the hazaia of the fire spread depends on the amount ofthe heat radiated by the burning house, fire protection performanceof the external eleinents of the house that receive the radiation andthe geometrical relations between the two houses. However amongtheml what we should regard as a performance of a house is only fireprotection performance of external elements.
l2l ProcedureHere, each of external elements is directfy graded based on
regulations of Building Stqndard Law, and a configulation factorwhich is required for the materials of each grade so as notto catch firL from burning house is attached to every grade.
The required configulation factor can be translated to therequired diitance between a burning house and the object house toprevent fire spread by the following manner-
(f) Shape and dirnension of fireFor the sinplicity of calculation, the shape of fire of a burrr
ing house is assumed rectangular.The breadth of fire a is taken the breath of the house normal-
ly projected to the target surface and the height of fire H is takenas
a)When a bare wooden house is buェ lling
H = ―書― x (projected area of the burning house)
b)The other cagearea of openings and roofs of the)
distance ・く ・~
::.' Sl.nce the shape and the dimension of the fire when a neighborhouse burns can be estimated as above, the iequired distance corres-ponding to a requLred configulation factor added to each grade caneasily be obtained rnaking use of some proper expressions or graphsfor configulation factor.
(3) GradingThe grading for each'component of external elements of dwellings
(■ ) ViewpointFron Statistical data′ we can ■earn sevlral aspects on human
:こi』』]t::ncま:u:]T『 efiliSiof:rf:I:習pt: [R:せ
'。
1131。ltebilillng fII:s■s remはrkably high′ that a large :][t。
lftitevi:Ell:tニユ:::こ:leitnight and that the considerable pby the o■ d and childreno Ther● fore it is a matter of course thatin dealing ■ith the problem′ we cannOt diSregard these charaCter―
(2) ProcedureThe evacuation performance is evaluated fOr every important
[::翼 視 it::le[ia:Willl:呈
`uittt:1'al[:llil:ll:ya°
fet:[IIILd° fr:と::ua_
dure′ while this "reliability: dOeS ntO wear a strict sense concer貯ing with probabi■ ity′ but is Only a expedient to Classify the extentof safety ■ikeliness of a S■mple routes or a network of the routes.
The reliability of the netWork of the evacuation routes ■S
calculated according to the folloW■ng procedure.
(1) regarding any of the roolsafety area as a nod and then conlof the evacuation routes in the o】provュded that the network S,artS ithe safety area.
A Safety area etcc are defined ashazardous space8 eVery space ■n a dwelling
as a rule, ground surface but can be proper-ly interpreled according to actual circum-gtances
space between hazardous sPace and safetyarea ex. a balconY
(ii) quote the reliability of the rqrte between every pair ofnods from Tab. L2.
(iii) eolve the reliability of the network of the evacuationroutes.
The solution can be performed by iterating the following pro-cedure successively, that is
b) if paths I and j are Parallel*iJ t 1- (1 - f1)(1 - Yjl ..................o..o.'...(4-21
rhere y1 and y1 are the reliabilities of individual routes t and jrespectiiely and n11 le tlre reeultant reliability of the two routes.
(31 GradingIthe reltablllty of network of evacuation routes is graded as
in Tab. 13.. Usually, the value of the reliability- is consideraply high for
tfie roons oi the first floor and low for the rooms on the uPPerfloor.I. CONCLUDING REUARKS
The present evaluation'system is.based on the viewpoint thatrire ilie;e-ia- the iiansition-of burning phase.of fire and accord-ingiy i:tre fire protection performance of dwelling houses is under-stood as the syiihesis of Lne performances to prevent the transi-tions of the fire Phases. .
As a natural consequence of the viewpoint, some value whichhas a sense concerning iitn probability oi critical condition ofthe transition of the fire phase becomes essentially desirable asttre basis of the evaluationl However at present, the studies andthe data for this purpose.are so insufficient that the evaluationsystem now stays aL the stage in nhich we can only evaluate aboutririctr is likely to be superior on fire protection performance.
Therefore it is unquestionable that for a future evaluationsystem, the efforts to iccurnulate fundamental studies and data areviry ilnportant. Ttle followings are brief introductions of suchstuiies-perfonned within the present study for this aim.
REFERENCES
tfl BRI, "Development of Evaluating System for Total Performanceof Houseg - Fire Protection Performance', 1978.3
tzl y. Morishita, 'statistical Analysis of Fire Spread Process inHouses", 3rd Joint Panel !{eeting on the U..J.N.R. Panel on Fire
. Research and Safety, llarch 1978, Washington, D.C.
t3l T. Tanaka, 'A llodel on Fire Spread in Small Scale Buildings",'B.R.I. Research Paper No. 79, Sept. 1978
3 .'・ .
=-68(“
1974(B.R.I.
1975(BoC.J。 )
1976(B.C.」 。)
1977(BoC.J。 )
Fig。 l cOurse of the Study
Elementary experiments on behavlor
8f ill:ial stage(temperat`re・ dis‐tribution in room,ignition of inte-rior finishmaterial )
Revievr of the characteristics of drelllngfire and evaluatlon mthod of flreprotection perforrnance for drell lngs
Investigatlons intofactors, terms andsystem of the eval ua
tion system
Collection of flrecases and investiga-tion tnto method ofthe analysis of thecases
lntroduction of thefire phase, statis-tical analysis offire spread processein houses and model-ling of dwelling fi
velopment of thevaluation system onire protection per-
formance for dr*el-I ings
Collection and analysis of fire cases
(continued)
I-69 に5,
[:こ]]曇]EI}―――
盤量五[}― 盤_|…
睡 1亜l亜■―
師手ご予3眠11司
墨躍輩:三∫~~
囲警野―椰 轟l―
each type houSe draft planni
fire prot,ectionprrformance of
「‐~~~‐ ~~~‐ ―o¬ ●
卜vertical shaftトーーーL`_________‐」
fire protectionperformance at
vertical shaft
fire protectionperformance oflnterior finish
隈 橘 子F研~7~1
ロニ望1聖」l劇
Flg. 2 Structure of the Evaluatlon System (r for tall house)
ire protectionperformance ofdwellinq houses
fire protecticnperfonmnce atdetailed olanni
fire-using
PreventiOn againstfire outbreak
systen for detectiand exti ngui shn:ent prevention against
initial spread ofof fi re
live combustibles
prevention againstinter roomspread of fi re
openi ngs. dimension. flre protection
performance
fire protectionperformance of
rti ti ons
fire protectionperformance ofexternal elements prevention against
inter housespread of fi re
distance fromneighbor houses
evacuatlon
netvork o
I-70 にの
kind and amuntof conbustlbl es
arrangement conditionln room
fi re-us ing' appl i ancesih room of dnelling
cosbustible materialfactor Y
arrangernnt factor gfire-using appliancesscore o
K"aCy
50 くKく 100
100 くK( 150
150 くKく 200
VES(lst)
YEs, (zno)
(3rd)
YES,(qtn )
(5th)
SAFE
|
for average dwelllng
K ・ 100
|
HAZAR00US
Fiq. 3 Evaluation Procedure for Prevention Performance of 0utbreak of Fire
START
i = assumed room of fire oriqin
j=摯 ll. 9R賢黒 a rOundburnlng rate
Ri=5.5Aυ ‖
total effectivefuel load
l{i Calculation of delay tin€of fi re spread of componentsof parti tions
f1;1 (k=all components)fi re durationr i =t{i /R;
:,1:re lT,t♀ [ifir:r::I:adfire room
Plj・min。 (Pljk) 。
I <- every j
!ij-'iroom J'room to which fire
does not reachfrom iioom j=7eom to rhich fire reached fqom i
fire spread tinn: B burned areat: A
I-71 (4T
訓∃劇
cl測J3‐到「ョJЫ引劃―馴d目
〕一っ0こ コこう0一””一〓” 」0 ”””“”一こ一“〕“
ヨ
「
釧
ヨ
ゴ
ョ
∃
劇
司
到
コ
釧
司
J
コ
司
1‐
目
“0
〓‐ ご ‐
▲■
ュ
■
▲
I
■
/1
1酵代
〕”つ0こ 〓0一いくつ0“”〕
」0 】“0コ↑〕〓
ユ』」」” 一
(∠■′I-72
Tab. l Phase of Flrl
l Flre prigin burning = A heat source whichflre is burning.
potentially outbreaks
= 0utbreak of fire2. A part of furniture
finish materials inwhile large part Ofyet ignitёd.
Partiallng ・
room burn- 一一一 or corbustible interior
fire room is burning,the cor$ustible has not
= Initial fire spread
3. llhole room burning =
Large part of live and dead fuel in fireignited and has burnt up whole room.
rQom
晨Lロメニ = Inter r。。m fire Spread
4. Whole house burnlng tt Fire spread out of the room of origin and whole
]:lЯ l‖:rhouse to a neighbor house and it has一一
一5。 Fi re spread
burnt up.
Tab. 2. Si, l{i of House for Exclusive Use of Living
Fire-using appl iance Si
Wi
:}Υ↓::e Ki tchen ba tn roomlミ十Arレ hnlJ
Gas
Gas rangeGas stove
Gas water heaterGas rice cooker
0.8
2.5
1.8
4.1
0.008
0.034
0.001
0.755
0。 009
0.014
0。 016
KeroseneKerosene stoveKerosene range
1。 0
]10
0.151
0.004
0。 035
0。 008
Elect‐
rici ty
Electric range
Electric stove280
52
0.016
0.028
0.001
CharcoalCharcoal range
Charcoal brazier240
54
0。 006
0.004
0。 003
Kerosene bath furnacr
Gas bath furnaceCoal bath furnaceUood bath furnace0il boiler
2221
4
1
5
3
2
5
0.347
0.405
0.005
0。 0580。 010
Tabacco 0.5, 1.0, 1.5 0.220 0.016
0thers 1.0 0.529 0.144 0.174
Π-73 に1,
lrade pf in‐
|四‖::l::}1句1■ ■Ch
Combustible materialfactor
I
ヽ
′
t
r
l
′
0。 95
1.0
1。 05
Tab. 3 Combustible Material Factor
Tab. 5 qi for Each Interior Element
re
Value of K Grade
below 50
50 ‐ 100
100 ‐ 150
150 ‐ 200
abOve 200
Element Kind of Perfomtance ql
Interior finishmateri al
Inconbustibl e nrateri al
Quasi -incorbustibl e material
Fi re-retardant materi al
Others
8
6
4
0
Opening, etc.
Flre door (Class A,
Steel door
l{ooden doorFusuma, Shoji
B)8
6
2
0
Tab. 6 Flre tion Per ce of Interior FjniS!
use Of roomIncombustibllityntmber
Grade
Roorm ln rhich flre-usingappllances are usuallyused
more than 8
7-86-74-6below 4
1
2
3
4
Other roms
6‐ 8
5‐ 6
4‐ 5
3‐ 4
below 3
1
2
3
4
5
I-74(ぎ o′
Tab. 7 Flre Protection Performance of Flre Preventlon lhvlces
Instal I ati on Klnd of equlpment Grade
Detector
Heat detector and smoke detectorSmoke detectorHeat detectortlo lnstallatlon
1
2
3
4
Extlngui shi n9
devices
Sprinklerr ':
Flre Hydrant
lfater for fire fightingEiti ngui sher
tlo instal I atlon
1
2
3
4
5
Tab. 8 Reductloh Rate
Kind Val ue
Due to fire protectionperformance of interiorfini sh
Incombustible 0
Quasi-incontustible 0.3
Fire retardant 0:8
combustibre tlood ' I
Plastics 2
Due to fire protectionperformance of backing
Y
Inconbustible 0.1
Quasi'inconbustible 0.4
Others - I
Due to flre protectionperformance of opening
structure6
Fire door (Class A) 0.1
Flre door (Class 8) . 0.3
Glass door 0.6
tfooden door 0.7
Fusuma 0.8
0thers I
=-75 (F/)
Tab. 9 Delay Tine of Fire Spread of Partitlons
Tested ornot Kind Delay tlme (min.)
t{i thFi reProtect-IngTest
PaFtit10n
Fall
Flrc prrof constructlon(F.P.T. . Flre Proof Tlme)
above 30 min. F.PoT. +20
“ 15 mlne 50
" 10 min. 30
Flrne protection performance 40
Same to loam coated wall 20
Door
Fire door (Class A) 120
Flre door (Clast B, 30
Glass or wood 10
Fusr,ma 3
UentSel f-closlng
Stee1 200
Almi‐ 140n l uln
Others 40
Usually opened
‖ith combustibles within
llュtan,9 1n
t{i thout combusti bl es wi thin distanceI m.......c
‖ithoutFi reprotect‐
ingTest
rlaster board
Thickness above 15nn..3012mm ‐ 1 5rrrn..20
9rrEn‐ 1 2rlnl.15
below 9 mm... 5;iber reinforced plaster boardlsbesto slatelsbesto cennnted calcir.m si I icate
hn
P of plaster board +5
'lywoodP of plaster board -5
ilass wool, Rock wool lhickness above 50trn..l0
25mm ‐ 50mm.。 5
tlastlc foan 0
Tab. lo preventlon perfomance of lnter Room Flre Spread `・ ・ ・ 、
Ualue of C Grade
belo" 2。 5
2.5‐ 5。 0
5。 0‐ 10。 0
10。 0‐ 20.0
more than 20
2
3
4
5
I-76 (S2)
Tab. ll Preventlon Performance of Inter House Fire Spread
I Elementi
Performance ofelement
Grade App.l icationto partsliable tocatch fire
Requi red configulation i
factor i
@house I bear woodeniI house i
xterna Iall
Superior to fireprotection structure
OK ≦ 0.50 ≦ 0.55
Fire protectionstructure 2 OK
` 0。
30 ξ O。 35
Similar.to loamcoa ted, wa I I 3
OK(in part) ≦0.10 JO.15
'Inferior toloam coatedwal I
4 N0 ≦ 0.05‘
8 0.10
gnnep
―
1
0
,―
―
―
…
…
Superior to firedoor of class A
OK ≦ 0.50 ≦ 0.55
Fi recl ass
door ofA
2 OK ≦ 0。 30 ≦ 0.35
Fi reclass
door ofB
3 OK ≦0・ 10 ≦ 0.15
Inferi orfi re doorclass B
tO
Of 4 OK(i n part) ≦ 0。 05 ≦ 0・ 10
Roof
Superior to fireprotection structure
OK ≦ o.5o ≦ 0.55
Fire protectionstructure 2 OK ≦0.30 ≦ 0.35
Incombustibl eroof coverauthorized bylaw
3 OK
( in part)≦0010 ≦ 0。 15
Other than ln‐
combustible roofcover author12edby law
4 NO ≦ 0。 05 ≦0.10
I-77 ・ (ご 3'
0
“.一
一.0
”.0
).●
一.0
一.一”
8‥〓罠ab『8‐「8‥掛司」■α日「
〔‥「ヨ
,
|■ ■
1
・
・
一0ぃC
C●こ
劃
副
引
∃
ヨ
ヨ
劃
コ
ヨ
コ
割
刊
J
J
到
一〓
)
‘
●
一
C
C
|
|
´一
一”一4
●
二
●Ь
0
一下”)・∝・0
.
」O L00L● ●二“
一
^“,”)・・・o
.
一0 ●●●L●” ●こ“
L卜翼F妻
●
。● 一』●6●
gO中″●ち●一●゛ ‘“一3
(¨3”)
¨一.・
「
い0 」●一●●” ●二“
一。●
一●●一●●〓 “っ●´“゛
,
0。”
¨一●●一●●“・‘“一3
(“・・)
一゛・・ ゛
いO Lo一●●一 ●‘“
し
一一
●三
●
‘
0′ 個
由
デC●一一●一一」0゛■●¨0
0こョOLO ●‘“
0一 、一一0●L一●
.”
、●●υ”●● EOS』
妻鷲e露κ
』興」c墨
バ
OCっOL一 ●‘“ ●“
、”“●●L一
, 一●oL ・
一
“●”」
●0こ」
一L一●“゛ ‘一っ0し〓“
輩
群
ぶ
一●0つ●”‘ぃっOL‘“
ピヨ0し一 ●〓“ 0一 .一
いoOL¨●一い●OL一
ごョご0
,3
o“
【¨“●●』゛一 いoOL。ヽ
Yりc二
0●
Eo●い
〕つV■ 5 1lJJ00 ●0● J]JVdS u0001::| 。1
0“OC
0い L0
C,
●“L”C r“一3
」,
C●ン●
C
”一一〕一一C●C
ぃL00つ OL一
●〓“ OLOE一 」0
0い0〕
●こ“ C
,
0一うに0一一Ca●
●C0 0C一つ
,”゛
LO L0〇一
0いC一‘ 0 」0」
●”Oc〕二
〇0●
⌒C)〓
(E)〓
(F)〓
(G)2
(oし)0
●”ぃC●¨一
、一一一一0●一”●L
∞.0
一.0
一.0
”.一
(・”)(一一)“∞・・“
」O L●ぃLO” ●〓“
(‐”)(一一)】∞一〇“
」0 ●00LO一 ●〓“
卜.0 礼
″“
』O L●ぃLO″ ●〓“
Ce・”一(ぬ一び「″)“
PにLぼFF腔膨ト
』
・”
、
0 」00」●戸 0〓“
‘■
,C CC
一一、Gψ
O¨い“〓ロエ0い
1
1●
‥
一●8
ト
ト
|ギ 卜十1
「日
両
却
一o両
¨
蔓 J::::]:ll日
一ノ
一F「=手ヨ「
マ1ヨ
●0一一●0口』一゛゛●¨0
一C
,C●0● ・
”
L00〇 一一一一一一 。N
』000 ●FOC
,い 。
一
゛』一●“一 .一
●.
00,
080つヨ ロ0-つ ゛
F中OL
,C●‘ ●
‘“口〓』●●0 ぃC
, o”
・一学”゛ OFO●●一
LOoO ●Pぃ●一● .一
、C00”●一 .¨
」00L 。い
い00L 一●”」 。“
ヨ3VdS uOoo‖ l ol
30VdS uooo‖ : 口OJ`
llvds u3JJn8Slvds u000rn0 0lSlvds u000ilI IroJJ
ヨ,VdS uヨ JJng ЧrlnOJι :1
(ξ 4I-78
I I-7
A Methodology for Evaluating Fire/Life Safety Planning of Tall Buildings
1. fntroduction' In order to prevent fire disasters in tall buildings, manykinds of qquipments and facilities such as fire detecti5n-deviies,sprinklers, fire doors, smoke control system and escape stairs hadbeen devised and are installed in many of the modern Luildings inJapan. But because of the lacking of iommon laguage that canevaluate and compare the effectiveness of each corfntermeasure.from the standpoint of life safety assurance, it is still a pendingissue to decide to what extent a Luilding should be equipped'fotpreventing fire disasters. And, as the r6sult, some Uriifhings areshort-equipped and, on the other hand, s.gme are over-equipp6a.
- To-coPe with this problem, T. Terair)had proposed dn Lvafuationmethod in which a seriei of countermeasures to pievent firedisasters ttere dealt with as systems and their affectiveness wereexPressel bV systems' reliability that vrere given in probability.Basically, !tt. evaluation method we've iroposed irere is alinostthe same as Terai's on its way of thinking fir ivaluation. Butamong a serles of fire disastirs prevention systemsr w€ have par-ticularlY- taken up a system that acts to secure occupantsr iafetyafter a fire broke out in a building. And we have expriined theconstruction of the system and the rnethod of evaluation of itseffectiveness concret-ly. And it is a characteristic of ourevaluation method that a f*e/Life safety planrting of a building isevaluated_by- the casualties of expectati6n-that i6 obtainea Uy lfreproduct of the probability that lite safety securing system o? abuilding will fail and thE nurnber of occupints who inarr encounterltfe dangerous sLtuatLons.
2. construction of ELre/Life safety securing systemIn order to secure evacuation safety of occupants, it Ls alsopossible to construct a considerably comilex sy"iE, that, forinstanee, informs occupants of the 6utbreak of fire as soon aspossible and leads to Lvacuate them to a safe pfa-e before smokespreads throughout a building. But in fire sitiations, as we have
MASAMI KOBAYASHI, Dr.Eng.
Assistant of Architectural EngineeringI(yoto UniversityI(yoto-City, Japan
I-93 6s,
to control human behavior in emergencies that is very instable andhas many indeterminant elementsr w€ think it is first of allnecessary to secure minimum saf,ety by sirnple and steady systemthat consists of steady countermeasures whose functionirreliability are confirned to be very high and stable.rn a building, if an igniting fire was.,.extinguished right atonce, its occupantsr safety is of course gu?anteed. And even thoughtltq extinguishment of fire was failed and its smoke is spreadingalf over the buildingr the occupantsr safety can be also secured ifthere remained a safe escape route in every-floor. Therefore, $reconstruct fLre/life safety securing system by these two fundamentalstrategies and as *to* in Fig.r, they form a.parallel system.
Fig.l Construction of Elte/Life Safety Securing SystemAutomatic fire extinguishers as sprinklers are very effectiveto put out flre in its early stage and they are installad in manyof the recent tarl buildings in Japan. And lately, most of the
dead in building fires are those who were suffoclted by smoke orcombustion gas and to avoid this tragedy in tall buitdings, itcould be a steady and fundamental countermeasure to secuie at leastone smoke safe stair in each floor- Therefore, we assign ,'FireExtinguishment by Automatic sprinkrer system" and ,'seciring atleast one Smoke Safe Stair in each Flooi" for the objectives of twosubsyetems Ln Flg.I (see Fig.2).
Fig.2 Object of Fire/LLfe Safety Securing SystemWe- express the effectiveness of firerllife safety securing
system by "Reli.ability" that is defined as probabilily a givensystem will perform as antieipated. And if we denote Lhereliability of fire extinguishment system by RS and the reliabilityof_safe escape route securing system by REr-the probability thatlife dangerous situation will happen to the occupants is given by:
・F = (■ ― RS)。 (■ ― RE)
And' moreoverr if we denote the nunber of occupants who shallcome to encounter this life dangerous situation by e, thecasualties of expectation by a fire is expressed by:
D=Q・ F
= 0・ (■ ― RS)o(■ ― RE)
(ご6)
FIRE EXTINGUISHMENTIN EARLY STAGE
SECURING OF SAFEESCAPE ROUTE
FIRE EXTINGUISHMENT BYAUTOMATIC SPRINKLER SYSTEM
SECURINC AT LEAST ONE SMOESAIIn STAIR IN EACH FL00R
Π-94
Therefore, it can be said that the final target of fire/lifesafety planning of buildings is to minimize this D as possible. Butas thi value oi D varies Uy tfre change of the floor where a firebreaks out and also by the-number of floor where the occupants 9f 0are stayihgr w€ explain the process of calculation of this D valuein the following chapters.
3. Evaluation of Fire/Llfe Safety Securing System
3-1 Prior Conditions for EvaluationWe evaluate fire/Iife safety securing system of a building
under the following conditions.a) The fite/Life safety securing system we evaluate here is theone that acts after a fire had broken out in a building. And sorthose countermeasures that are effective to prevent fire outbreakitself are not evaluated.b) Among the countermeasures to extinguish fires, only the in-stallation of automatic sprinkler system is evaluated here. Andfire extinguishments by occupants or fire brigades are not evalu-ated.c) And if a fire eould not be extinguished by autornatic sprin-kler system, its smoke is assumed to spread aII upper floorsthrough vertical shafts such as elevators, stairs without doorsand piping shaftsr €tc.d) But occupants in and below theassumed to be able to escape safelYe) In a building, exterior stairsfire doors are the only routes that,
floor of fire origin arefrom the building.and interior stairs which havewe regard, can be safe escaPe
routes in fire situations.f) And sor life dangerous situation for occupants is defined asthe state that there remains no smoke safe stairs in a floor whichis also contaminated by smoke.
3-2 Fire Extinguishment in a FloorWe denote the probability that a fire which broke out in the
i-th floor is extinguished in its early stage by si. And if thei-th floor has automatic sprinkler systemr w€ give si= c, (c:constant value of probability), and if it doesn't, w;l.give si= 0.0.
According to the recent research by A. Watanabe", the relia-bility of sprinkler system is estimated at about c = 0.98 which wasobtained by- the analysis of records of performance of sprinklers inreal fires. Therefore, the probability that the smoke of a firewhich broke out in the t-th floor will spread to all upper floorsis given by:
FSι = ■ ― sι
And this is the unreliability of fire extinguishing system inthe i-th floor.3-3 Evacuation Safety Ln StaLrs
We evaluate the safety of evacuationfire doors for its escape routes.
that uses stairs with
In order to be of evacuation use, a staircase musted from the spread. of smoke and must be entered easilyas well. And this is determined by the states of doorsconnecting a staircase with each floor.
be protect-by occupantsthat are
I-95 (r,
a) Evacuation Safety in an Interior StairAs smoke goes upward, if lt got in an interl.or staircase at
one of lts int6rmediite fioors, occupants in the upper floors canrt
";; the staircase any more. So, it is necessary for-interior stairsifrit their doors havi been closed when smoke is coming from floors.'gut as is often the case with stairs whi.ch are frequently usedevery day thdt their dbors have been kept open and sor it ie rec-omrneirded that interior stairs should have fire doors that are coD-nected wlth smoke detectors and are able to be cloEed automatically.
The followings are of a building which ls h stories high andhas n stairs for escaPe use.
We denotE-tfre prlUafitity that the door of the j-t! interiorstair at the .-ttt fioor is cl6sed when smoke is coming from t!e. m-thfLoor by x;,,;. And we call it the reliability of evacuation safetyof an interior stairfs door.
Wtren a fi-re broke out in the i-th floor and its smoke is
"pt"uaittf ittto-ifi ,tpp"r floors, the j-th interior staircase, inorder to be of use f;i the escape from the k-th floor (k > i), mustue protected from the spread of- smoke at all fLoors from i to k-1'So,- the probabiLity thai the j-th interior staircase can be usedfor the escape froir the k-th iLoor when smok-e is spreading from thei-th floor is given bY:
r b-lr;,i = ^tu*
-,,So, the probabilitY that
to be used for the evacuation
€L - 1 - -ithti - r 'hrih-r
= I.- E.xmrimrb
b) Evacuation SafetY
(4)
the j-th interior staircase is unablefrom-the k-th floor is written bY:
(s)
An exterior stair facing to the open air is assumed to be asafe escape route even thougf, smoke.is-spreading into it from some
floors. But instead, doors 5f exterior slairs have been often keptunder lock "rrd
f"y ior security and, as the result, it sometimeshappen that ".""p"""i" canrt usi thern in emefgencies' So, it isdesirable that exterior stairs should have fire doors that can be
unlocked electrically by the remote operation from a building'scontrol room.
we assume that though smoke is spreading into a floor, itsoccupants ."tt-.""ipe-sat6fy_if there -is an eiterior stair in thefloor whose door is unlocfla. And we denote the Probability-that-il;-a";;-;; in" j-th exterior stair at the k-th iloor is unlockedin fire situalioi"-Uy-i[,;. And we call it the re]iability ofevacuation safety of- an--6kterior stair' s door ' '
Foi exterior stairs, this xth,i Uecomes, "" it is, r-o. withoutregarding to the t-th efooi--wtr.i."t fire broke out. ' Q't
in an Exterior Stair
,ii,t = x, E,iAnd eor ff.i becornes as shown below.
-i.fi,i= t - x't'i
tr-96 (J2
3-4 Evacuation Safety in a FloorWe-'ve guranteed the success of evacuation from a floor by thestate that there remains at least one smoke safe stair into
"ririctpeople can enter from the floor. So, the failure of evacuation isdefined as the state that all stairs in a floor are contaminatedby smoke or blocked by locking. Therefore, in a h-stories uuiiaingwith n stairs for escape, when smoke generated in the i-th flooris spreading -int9 all upper fLoors, tf,e probability that all stairsin the k-th (k > i) can not be used for elacuation is given bvi ----
FE: =コ E fι′
=fr" e
J=l (r t n';-1
And among n stairs, if the first to the nr-th are interiorand the rest are exterior, FEi is written by:.' ''t' .' n
rhis FEf is the unreliability of the safe escape routesecuring system in the k-th floor when a fire broke but in thei-th floor.3-5 casualities of Expectation in a Building by a Fire
Ae lndlcated ao far, in a h-storied building that has n stairsfor escape, the probability that fire which Urot6 out in the i-thfroor was not extinguished-and its smoke spreads into arl upperfJ-oors is-given by rsd . And the probability tnat iir stairs inthe k-th floor are unable to be uled for gvicuation when smoke iscoming from the i-th floor is given by rnf . Therefore, theprobability.that life danger witt happen.€o the qccupants of thek-th floor is given by the product oi- rs,i and FEI . 'L'rui = rsa. re f (7)
- Then, if we denote the nurnber of occupants in the k-th floorbI g?, lhg casualties of expectation in thl k-th floor caused bythe i-th floor's fire is given by:
di = qた。FLE= qた 。FSι・FE.
As the life danger caused by the i-th fl6or's fire willhappen to all occupants in upper- floors, the casualties -of expec-tation in a wlorp buirding ii- given uy itre sum of all df. , wtrliek equals i+I to h.
Dピ = 査dだた‐
`,1た qた FSt FI=Σ Eた 。
た・ ι・ l
I-97 Cll
And′ moreover′ if we denOte the probabi■ ity that a fire breaks
So′ the value Of D is determined by the variables shOwn be■ ow.
Tabe■ List of Var■ab■es of Expectation casualtiesin a Building by a Fire
h
n
n・
pι
q々
St
XmP′
Xふ′′
nunlcer of stories.number of stairs for escape.number of interior stairsr sorexterior stairs.probability that a fire breakssor the sum of all p; becomesone to h.number of occupants in the k-th floorprobability that fire which broke out in the i-thfloor is extinguishedoprobability that the flre door of the j-th interiorstair at the m-th floor is closed when smoke iscomingr.probability that the fire door of the j-th exteriorstair at the m-th floor is unlocked in-fire s'ituations.
is the number of
out in the i-th floor,1.0, where i equals
4. ApplicationTo show the application of our evaluation methodr w€ calcu-late the casuarties of_expectation by a fire in the folrowingbuilding, for an example.It is a ten-storied building with four stairs for escape. Afire is assumed to break out in itr floors by the same prouiuiti-
!y. so, L/LO is given to all p; where i equars one to ten. Eachfloor has fifty persons and so; five hundfeds persons are stayingin the. buitding.At first, we calculate the casualties in the case that thereLs no automatic sprinkler system in the building. And all fourstairs are interior and all-their doors are giv6n the reliabilityof x = 0.5, (see Tab.2).
Tab.2 1nitia■ Conditions
n. pι qι Sι Xι ,′
0.■ 50 Oe0 0。 5
* i equals one to ten, and j equals one to four.10
Ⅱ-98 \QJ I
The results of calculations are as follows. The change of Ff,i,which is the probability that life danger will happen in i,ne j-thtfloor when a.fire broke out in the i-th floor, beiomes as shown inTab.3, and Ds , which is the sum of casualties of all floors when afire broke out in the i-!h floor, becomes as shown in Tab.4. And inthis case, the value of 5, which is the casualties of expectationin the building by one fire, becomes L29.3 persons.
Tab.3 Ft: : Probability that life danger wilr happen in the j-tha flooT when a fire broke out in the i-th floor underEhe initial conditions of Tab.2.
\ 1 2 4 5 6 7 9 9 10
ユ
2
3
4
5
6
7
8
9
10
0.0
0。 0
0。 0
0。 0
0。 0
0.0
0.0
0.0
0.0
0.0
0.0625
0。 0
0。 0
0.0
0。 0
0.0
0。 0
0.0
0。 0
0.0
0.3164
0.0625
0。 0
0。 0
0.0
0。 0
0.0
0.0
0。 0
0.0
0.5862
0.3164
0.0625
0。 0
0。 0
0。 0
0。 0
0.0
0.0
0。 0
0.7725
0.5862
0.3164
0。 0625
0。 0
0.0
0。 0
0。 0
0。 0
0.0
0.0807
0.7725
0。 5862
0.3164
0。 0625
0。 0
0。 0
0。 0
0。 0
0。 0
0.9390
0.0007
0。 7725
0。 5862
0。 3164
0。 0625
0.0
0.0
0。 0
0.0
0.9691
0.9390
0.3007
0。 7725
0.5862
0.3164
0。 0625
0。 0
0.0
0。 0
0.9845
0.9691
0。 9390
0.8807
0.7725
0.5862
0.3164
0。 1625
0.0
0.0
0。 9922
0。 9845
0.9691
0。 9390
0。 8807
0.7725
0・.5862
0.3164
0。 0625
0.0
Tab.4 Pl : Casualties when aunder the initial
fire broke out in the i-th floorconditions of Tab.2.
Then, to see the effects of installation of exterior stairs,automatic sprinkler.-system and fire doors of higher reliability,we change those variables independently and caliulate 5 in eacir'case.
a) Installation of Exterior Stairsg{hen two stairs out of fogr are exterior, the change of rr,i
becomes as shown in Tab.5 and D becomes 39.9 persons. eia when itlfour stairs are-exterior, ELi becomes 0.0625 ior all i and j, andthe value of D becomes l4.l lersons (see Tab.G).b) Installation of Automatic Sprinkler System
l{e give si = 0.9 ifas the nunber of floorsvalue decreases as shown
tliti:::;:き::Inli:r :」 :t[lein:I:よ :習 : 3nd
c) fnstallation of FLre Doors of High ReliabiliqyIn the initial conditions of Tab.2, all fire doors of stairsttere given the reliability of x = 0.5. io see the effects of in-stallation of higher reliability doors, we calculate 6 ly incieas-ing the floors gradualry in whiih arl ihe doors of stair-s aregJ.ven the reliability oi x = 0.9.The results are shown in Tab.g.
Ⅱ-99 (61)
Tab.5 FL;′:1:19[習
°stairs under the initial conditions areto exterior stairs.
tXj ■ 2 3 4 5 6 7 8 9 ■0
■
2
3
4
5
6
7
8
9
10
0.0
0。 0
0.0
0.0
0.0
0.0
0.0
0.0
0。 0
0。 0
0。 0625
0.0
0。 0
0。 0
0。 0
0。 0
0。 0
0.0
0。 0
0。 0
0。 1406
0。 0625
0。 0
0。 0
0。 0
oゝ o
O。 0
0。 0
0。 0
0。 0
0.1914
0.1406
0.0625
0。 0
0。 0
0。 0
0。 0
0。 0
0。 0
0.0
0.2197
0。 1914
0.1406
0。 0625
0.0
0。 0
0.0
0.0
0。 0
0.0
0.2346
0.2197
0。 1914
0。 1406
0。 0625
0.0
0。 0
0。 0
0。 0
0.0
0.2423
0.2346
0。 2■ 97
0。 1914
0.1406
0.0625
0。 0
0.0
0。 0
0.0
2461
2423
2346
2197
1914
1406
0625
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2481
0.2461
0.2423
0.2346
0。 2197
0.1914
0.■ 406
0.0625
0.0
0.0
0.24900。 248■
0.2461
0。 2423
0.2346
0.2■ 97
0.1914
0。 1406
0。 0625
0。 0
NUMBER OF STAIRSINTERIOR EXTERIOR
2
0
Tab.5 Installation of Exterior Stairs
CASE 6 (persons)
a-1
a-2
39.9
14。 ■
Tab.7 Installation of Automatic Sprinkler System
CASE 6 (persons)
Tab。 8 1nsta■ ■ation of Fire Doors of High Re■iability
CASE 6 (persons)
b-1
b-2
b-3
b-4
b-5
C― ■
c-2
c-3
c-4
c-5
54.9
27.1
13。 2
12.9
115。 2
■04。 3
69。 1
■2.6
5。 2
88.2
(2)
FL00R WITH AUTOMATICSPRINKLER SYSTEM (Sι = 0.9)
■′2′ 3f
■′2′ 3′ 4′ 5f
■′2′ 3′ 4′ 5′ 6′ 7′ 8f
■′2′ 3′ 4′ 5′ 6′ 7′ 8′ 9′ 10f
5′ 6′ 7′ 8′ 9′ ■Of
FL00R WITH HIGHER (x = 0.9)REL工ABILITY STAIR D00RS
■′2′ 3f
■′2′ 3′ 4′ 5f
■′2′ 3′ 4′ 5′ 6′ 7′ 8f
■′2′ 3′ 4′ 5′ 6′ 7′ 8′ 9′ ■Of
5′ 6′ 7′ 8′ 9′ ■Of
Ⅱ-100
d) Alteration of tl7 P; and qiIf there are only two interior stairs in the building, that is
n=2, and other conditi.ons gre the same as Tab.2, the casualties ofexpectation becomes uP to D = 159.4 persons.- If we could redule the rate of Pi, which is the probabilitythat a. fire breaks out in the i-th floorr ES the floor goes down asshown in Tab.9, and other conditions are the same as Tab.2, D valuebecomes 65.3 persons
And if we- could reduce Ai t which is the number of occuPants inthe i-th floorr ds the floor goes gp as shown in Tab.IO' and otherconditions are the same as Tab.2, D value becomes, also, 66.3persons.
Tab.9 Alteration of FireOutbreak Rates of Floors
Tab.10 Alteration ofOccupants DistributionFL00R
ユ
PERSONSFL00R■
■0
9
8
7
6
5
4
3
2
■
PROBAB工 L工TYpι
0.19
0。 ■7
0.■ 5
0。 ■3
0。 11
0。 09
0。 07
0.05
0.03
0。 0■
■0
9
8
7
6
5
4
3
2
1
5
15
25
35
45
55
65
75
85
95
二 pι =1。 0
D=66。 3
conditiong of Tab.2.Tab.l2 shows the
combinations, and sorcountermeasures to thereduce the casualtiesoccupants.
二 qι =500D=66。 3
results of calculations of 6 values of majorl.f we give more than case €-1, e-3 or e-6building of initial conditions, we can
of expectation less than I t of the total
e) Effects of Combining Different CountermeasuresBy aombining three types of six different countermeasures
which lre shown in Tab.ll, we try to see the effects when more thantwo kLnds of countermeasures are given to the bqilding of initial
I-101 (`二 〕
Tab.ll Varieties of Countermeasures
BB
B
CC
AA
A
e-1
e-2
e-3
e-4
e-5
e-6
COUNTERMEASURE
A11 floors have automatic sprinkler system.Half lower floors (If -5f) have automatic sprinkrersystem.All four gtairs are exterior.faro stairs out of four are exterior.All stair {oors ofreliability of x =
all floors are given the
All stair dodrs of halfgiven the reliability of
■9wer f■ oors (■ f´ν5f)arex= 0.9。
5. Conclusionrf we could obtain fire freguency rate of occurrence per yearon each rvpe of buildlnss, and w6 shoi it by i (ii;;7y;iil.-iii.--casualties ir 3 buirding per year can be exlressed bi M = D x yr (per-son,/year). And, moreover, if we could sel ttre standard value ofM through the comparisons with other casualties of expectation insuch accidents as aircrafts, and trains' , we can ctrect trre .p-propriAteness of a building,s fire,/life iafety pfi""i"g ina iaiustit by operating the variabies in fiU.t.rn the evaluation method werve proposed here, the fire/lifesafety-of a building is evaluated by-thl probauirity that at leastone safe escape route can be secured in a- floor wneir a fire brokeout in a builiing. And to secure the above stater w€,ve constructedtlt" firerllife safetlt-securing system by physical countermeasureswhose f'unctional reliability-cai ue meisured as that of the in-dustrial products.- This is 6eciu".-r.-tiy to secure the minimumor fundamenral fire/Life safety of i-uuiiaing-;y;;"tirr.ry systemand have eliminated those countermeasures which'p"irit the inter-vention of human factors and,as the resurt, wtrosl r"ii.uiiiat-;;.difficult to be predicted quantitativeiy. However, for the successof evacuation under the coridition trrai-it least one smoke safestair is secured in a floor, it is neces""iv-tn"t""ri stairs inthe floor are recognizible and accessible f6r the occupants. so,the evaluation method we've- proposed here can be apprila ;;i"it fortalr office buildings that hive-simple irchitecturii prans.
Tab.12 Effects of Combining DifferentCountermeasures
D (persons)
4。 0
6。 9
4.2
10。 9
8.2
5.2
COMBINAT10N OF COUNTERMEASURES
AA + B
AA+C
BB +A
BB + C
A+B+C
CC′ (See case c-4)
I-102餡 〕
And WithOut the aid of proper human behavior in fire
ln tllCb Ullited states― ¨ are abOut E剛 疇 ‐ 鳳 d“ b」故裏IS品 lptthihedupOnfing hOmes(hOmes for elderly)and 7,000 hOspitals.Some ve10ping absOIЧ te values Of pЮ bability,which are belol thじ sc icnitLs do not meet current standards for Are_ yond the state Of the al at thぉ timei therefOre,hこsafc・りand thereお“require upgnang.In an erott t。
置:rd Ⅲhg VJucs ofrelative nsl,Or equaani
achieve economicJ retЮ nt systems and minimize dis‐
ぷ:盤鮮i∬
=翼
∬tl糖11[機器Stt nequ∝ぉn Jnsk LvdホO mse m ttepЮЫemご
designing buildings for ea出 quake loads.J.Ho wiggins2
tions.These、vaivers have been based On the use of al_ proposed a,yStem in which the risk level for seismic
temate protection means tO provide equivalency to he loading was deterrnined by an``importance FactOr."The
regulatiOns.The dlncult decisiOn as to what constimtes “importance factOr''was based on the nature ofthe build_equivalency has been leFt tO the various local jurisdic‐ ing occupancy,the tomi number Of occupants,and thetions,and,therefOre,there has been a lack Ofunifo品 ity amount oftime that the occupants spend in the building.
across the country in terms of、 vhat lnay be waived and
器i謝i盤lЪl諾望9よettγ9=b pЮ彎山el薔督ぶ蹴富庶■∫縄iti器置思鶏尉
As a restllt of this problem, the National Bureau of issued by the ltalian Ministtγ of lnteHOr3 deVe10ped aShndards(NBS)waS requested to develop a system that cOn9ept Of Classitting buildings by using an evaluatiOn
、vOuld provide a unifbrm method of evaluating healtt index.The ltalian regulatlons were designed to assist the
care lldh●es b detemme what Aresarety measures designer Of tte bundmび h sdecthg cHteHa br proper、vould prOvide a level of saFety equivalent tO ttat pro‐ lre prOtection.The building is described in tenns Ofa宙ded by NFPA 101,the Lル Sψ
`y Cοル,the 1973 6re load times an evJuatわ n index∞ emcient.The coe「
Editioni ncient is determined iom the evaluation index,which
more easily solved if there were an absolute degree Of tem,sprinkler systems,and other LctOrs that are ac‐
risk that could be determined,and ifthe risk level could ∝pted pararneters in Aresafety. The evaluation index
be calculated.The question ofabsolllte level ofrisk was provided a relativO risk level lbr the bullding;and thお
discussed by Chauncey Starr in his paper,3θ ■び"Cο
St index made possible the selection of the nre pЮ tec● onS,“″θs:73 SOCiorachn“
`:sys″ms.l He presented his design required fOr that designated risk.
data l,te.“ Is ofぬe pЮbability Of“ fatalities per person The Ce.1“ an lnsdtutioh for standardizatiOn't recOm_per hour exposed"in rehtiOn to a pammeter designated mendation DIN 182304is aSO∞ ncemed宙trdetemm‐as the“average annual benent per person involved。 '° He ing structural nre protectiOn, and its deterlninatiOn ismade the interesting observation that the public's ac‐ alsO based On a rik analysis evaluation。 ■ e predictedceptance ofrisk level can be related tO the pЮ bability of level for Are resistance is deterrnined■ om a series of
缶 tOrs that renect the pOssible degree of hv01vement,
Mr. Benjamin is Chief of the Fire Safcty Enginecring Division,Ccnter for Fire Research, Nationd Bureeu of Strnderds. Wrshington,D.C.
An address presented by the euthor et the Intcmrtiond Firc Srfetyin Llospikrls Symposium, held .f,pril 4-5, 197E, in Peris, Frencrt. Itctork on the firesa.fety evaluation system was sponsod by the USDepartment of llcalth, Education, and Welhre, undcr a HEIV-NBSLilL/Firesal'ety Project.
I Chauncey Strr, Beneft-Cort Sfudirr in Sociotechnical Syslerns,Public Safety: A Crouing Foctor in lloclqn Design. Nationd Academyof Engineers, Washington, D.C.. 1970.
5 2 . FrRE JOURNAL - MARCH 1979
I J. H. Wiggint, "Code Chengc Pmgosd No. ZXl.- EsildangStand-crdr, publishcd by thc Internatbnal Confcrenoe of Euilding Ofiicidr,Part III, MapJuno 193.
t Circular 91, ltalian Spccificatuncfr thc Ftc hotcction of Struc-nral Stccl Euildingc Dcdicatcd to Nor-.Militurl, Non-Industrial Ilu.Itelian lllinistry of Interior, Ililan, Italy, 1963.
' "Determination of the Fire Protection Class of Industrid Struc-tures," draft document DIN 18230. German Institution br Stand-ardization, Berlin, Germany, fune 1968 (trurslated by American lronand Steel Institute).
t92)
.,r,r5lblc fire loacls,.the nature and type of material to be
::,riuipatetl, :,nil tlre probable effectiveness of the ffre-
::rlitinq eflbrt. .ilthough this procedure does not present
!rotcr.rll fire risk concept, it does indicate a procedure
:or conrbiniru a number of probable levels of hazard.
Recently, Union Technique Interprofessionel de laf cderirtion du Batiment et des Travaux Publics (UTI) inFr.rnces published a draft document presenting a system
:ur evaluating ftre risk for both people and property.Using an extension of the system initially proposed by
Itar Gretener for insurance use, LI'II suggested egeneralized system for evaluating the safety of people in.r building, based on four components: evacuation Ume,
inroke development, toxicity, and tlae of ocrupancy.f'rorn these components, they arrived at e recpm-rnended desirable factor for the building, an approxima-rion of a risk level, without indication of the base fromrrhich the factor was derived.
The present evaluation concept at NBS was presentedinitially by the author at a Symposium in Ottawa,Cunula, in 1975, at a meeting of the NFPA Life SafetyCode Committee. Following that presentation, Harold\elson and Jeftey Shibe at NBS developed the conceptinto its present workable state. The cpncept presenteds'as a sy'stem that would provide for evaluating a build-ing aqainst theLife SofttV Code, endwould use conceptstrrken from the decision tree system that has been de-veloped over the last ten years and from the quantitativer:rting svstem, initially developed in the United Statesbv A. F. Dean in 1908. This system would be easilyrsorkirble, would present useful information for theiunount of effort expended, and would be one in whichthe authority having jurisdiction could place his conft-dence.
The evaluation system provides a tool for evaluating abuilding to determine its level of safety, either in rela-tiorr to the Life Safety Code or in relation to any othercode or regulation. The system provides a mechanismtlrat e.nables a designer to evaluate a range of alternativesolutions to determine those that can most ecunomicdlyupgrlde a facility so that the facility will end up with arerluired level of safety. And finally, using the systemsnggests alternative designs to the designer - designsthat may not have occurred to him or her, and that pro-vide new approaches for achieving the required level ofsafety.
The system is based upon the equivalency option thatis included in all the US codes. The speciffc rvording inthe NFPA Life Safety Code is as follows:
\othing in this Code is intended to prevent the use ofr!\r(.nrs, methods, or devices of equivalent quality, strength,
S E、aluatiOn du Risquc lncendie,'・ draFe d∝ument, unbn Tech‐
出‖穐ll蹴鮮l肥lふ夕
Fede面bn du B面 mett d“s TttШ=
ffre resistance, effectiveRess, durability, and safety to thoseprescribed by this Code, providing technical data is submittedto the authority havingjurisdiction to demonstrate equivalencyand the iystem, method, or devicc is approved for the in-tended purpos€.
This statement has been taken by the enforcing au-tborities in the United States to mean that one can pro-vide alternative designs to satis$ the regulations, if theddsigns provide a level of safety equivalent to that whichis called for in the existing regulations.
The system that we are describing is one that attemptsto calculate the risk to which the occupants of the build-ing are exposed and to provide a safety level in thebuilding equd to that level of risk.
From an analysis of the Ltfe Safag Co&, NBS re-searehers determined that there were l3 elements ofdesign, crnstruction, or eguipment that are specified atvarious levels. For a given type ofoccupancy, dependingupon the size and height of the building, these add to ameasure of the overall safety required for that type ofbuilding. In addition, our analysis of the Code indicatedthat the requirements could be broken down into threespecial subsystems: one having to do with containmentof ftres, a second having to do with extinguishment, anda third having to do with movement of people. To meetthe intent of the Code, one would have to meet the levelof safety of the total overall set of requirements plus thelevel for each of the three subsystems just mentioned.
The methodology used for developing this system was toevaluate in detail all the elements of risk and elements ofconstruction, and to assign relative values to each of theelements. The initial rvork was done in a Delphi exer-cise,o using ftre protection engineers and ffre researchersat.NBS. Following this initid exercise, a group of con-sultants was assembled consisting of designers, enforcingauthorities, and other experts in the ffeld to review andmalie changes in the initial evaluation. [n addition, rep-resentatives of the US Department of Health, Educa-tion, and Welfare were brought in to evaluate the ease ofuse of the system. The system was then fteld-tested onexisting buildings by various people to determine what,if *y, problems might develop.' To understand how the system works, we shall lookseparately at the risk evaluation part of the system, andthen look at the safety parameters or construction ele-ments. The risk evaluation for any building is calculated&om the product of ffve factors, after assigning weiglrting.6ctors to each of the ffve categories, depending upon thenature and type of the facility (see Table l).
.ltc Delphi technique t onc in whk*r crpcrts in r given ffcld srcasscmbled and their opinions recorded. The divergence in opiirion isthen rcpeatedly discussed until there is a generd ctmmon agieementon the result. The process is described in ur article by Nturray Turoff,*The Desigr of a Policy Delphi,- in Tehaobgicol Forecosting andSorial Changes 2, No. 2 (f9/0).
Patient mobilit7 is probably the single most importantfactor controlling risk in a health care facility. Mobilityrepresents the degree to rvhich patients must be assistedand/or can take actions necessary for their safety.
Patient density refers to the number of peopl6 at riskin a given ffre zone. The building is divided into ftrezones, and the risk is increased as the number ofpatientsin a given fire zone increases.
The fre zone location reflects the ac.cessibility of thezone to the fire department for suppression and rescue.The system recognizes the inherent advantage of a ftrst-floor fire zone, and imposes penalties for zones in theupper floors of a building.
The attendantlpatient rafio takes into account the im-pact on the safety of the patients of the number of sta.ffavailable to respond in an emergency. Emergency ac-tions that may be undertaken by the staff include detec-tion, alarm, ftre extinguishment, conffnement of the fire,establishing barriers between the patients and the ftre,rescue, emergency medical aid, and other related func-tions. The staffing ratio considered is based on the staffimmbdiately available at the minimum staftng level,which normally occuns at night.
We have added one other factor, to reflect the age ofthe patients. This rating takes into account the suscepti-bility to harm by smoke of people less than one year oldand over 65 years old. The imposition of this chargeprovides an additional risk factor in nursing homes andpediatric hursery rvards.
The total occupancy risk is therefore the product ofthe five factors, and this calculated risk must be olfset bythe building design or salety hctors (see Table 2).
The safety pararneters are the components of thebuilding constrrrction and are divided into l3 elements(see Table 3). These are a nleasure of those building and
5 { . FIrtE IOURNAL - MARCH 1979
“"“mm□ xttx由 :山 x由 8」
Trbl€ 2. Occ,upucT ri* 6c.tor celculetbo.
ftre-protection factors that afiect the safety of the perso6p.who may be in a parti,cular ftre zone et the time of a ftrelEach of the hctors is assigred a value, depending upspits presence or absence in the building and the level of.sophistication at which it is employed. The values trrg*from a negative value when a required component ir-missing to a positive value when more protection is pro..vided than the Code requires. The vdues assigned re. '
flect the relative merit ofeach of the items to each other,in relation to their effectiveness in life safety.
Values are assigned to evaluate the quality of the con-struction of the building. The values vary from *4 to .
-13, depending upon the number of floon in the build-ing, whether it is of combustible or noncombustible con-struction, and also whether the construction is "not pro-tected," "protected," or "ftre resistive" in accordancewith deffnitions available in the US codes.
The values assigned to each of the elements or cpm-ponents for the given building are additive. The totalvalue of the 13 elernents determines rvhat general levelof safety is provided in the building. The safety levelmust equal or exceed the value cdculated for risk.
An analysis of the 1976 Life SafetV Coda shorvs thatthe additive value of the components required in theCode tor a health care facility will vary, depending uponwhether the building is sprinklered or not sprinklered,and whether it is a nerv or existing building (see Table 4).
It should be noted that the Co& allows a lower level ofsafety for existing buildings, because of the problemsand costs of retrofttting to upgrade the existing build-ings. On an overall composite evaluation basis, the sunrrequired by the Life Safety Code for an existing build-ing, based on the addition of the sum of the values of therequired components in the 19/6 Edition, rvas approxi-mately one-half that of a new building.
001:ハ l‖‖[曽 :
Sa
EET1lGり iS‖ HCIISL
r:口〔 201〔 L0017:01 11● lo● [list lo, EEiSt
rilSI FL001
100V〔 F:lST FL001 5.0
Table 4. Ilandatory safety reguirements.
As mentioned previously, the total safety of the build-ing is evaluated in terms of the general safety, rvhich isthe sum of 13 components, and in terms of three subsys-terrts designated as containrnent safety, extinguishrnentsafety, and people movement (see Table 5).
Several methods has been used in the world to determine the level of firesafety in buildings- In Europe, the first useful method was initially pro-posed by M. GRETENER ( 1 ) for insurance use. This evaluarion concepE wasbased on the fundamental retation :
II‐4
MCthode E.8.1.C・ Evaluation du Risque Incendie par le Calcul
De CLUZEL P. SARRAT
Inglnieurs a la Direction de la Recherche del・ Un■on Technique lnterprofesslo■ nelle des
Fこ dこ rations Nationales du Batinent et des Travaux Pub■ icST REMY LES CHEVREUSE, FRANCE
where : R probable level of flre safetyP overall fire hazards
l{ protectlve and prevention measures
Since 1975, this approach led to further development in different Etropean
:il;11:蟹 :i:Iealddi[::[i::iptation from the、°riginal cRETENER Method
I-37(913)
In France, the same approach is developed by Union Technique Interprofession-nelle du Bitinent et des Travaux Publics (U.T.I.).
This body \ras requested by the French Government (Direction de la Securit€Civil.e) to develop a system which would provide a universal and sinple methodof evaluating the fire safety level in different types of existing buildingsor construction projects. Basic users are potentially : fire safety engineersor consulcants, fire authorities, fire figthers
The principal application for this trethod is to suggest economical retrofitsystems specially when decisions cannot be based on codes or current srandards
tt '.for fire safety engineering.
Ttre way that U.T.I.has chosen is to use an extension of the Gretener System,where the fire safety is evaluated for proper.ties protection and also forpeople safeguardlgenerally there is no conflict between life safety and pro-perties safety, but we consider that one protective or prevention measurehas not the same weight when we consider the first or the second fire safetyevaluation.
The system is based on the equivalence between all protection and preventionmeasures, that are nontral1y included in all codes or auEhorities juridiction.ERIC .Method (Evaluation du Risque Incendie par le Calcul) is principally fordesigners, or fire authorities use. We hoped Ehat the ERIC Method which purPu-sely is not scientific will be used in the future io determine the level offire protection and also to propose alternative protection measures.
To use this nethod suggest two statements :
must choosegoal
- secondty, to have a "fire safety barometer'r directly linked withfire prevention and proEection effort by two different calculaEions,one to evaluate people safetyranoEher to evaluate the level of pro-' tection for goods.
The evaluation of fire hazards for any building is calculated from Ehe pro-duct of different factors. The first factors(in relation with people safety)are evaluated with a large group of fire protection engineers, fire authori-ties and also representatives of laboratories ard French Government (Educa-tion Department, Health Departmenc ...). The second factors(in relation withproperties safety) are given by the GRETENER Method generally use for firesafety evaluation of industrial buildings
- first, to give different equivalent solutions ; youbetween all of them in relacion with your own safety
Ⅱ-38 (97)
l{e have also checked with the group every fire prevention and protectionmeasure.
Different values, one for people safety another for properties protectionare atlocated to evaluate the performance of usual meaaures adopted in firebuilding protection.
We have guccessively etudied the following topics 3
t. E!:e-E*ergg :
- for people protection- for goods protection
2. Eleygggigg-eg9_Eg,g9gglgg_gggggEeg, weight of measures relarively ro :
- people safety,- proPerties protection. '
The mechanism of the ERIC Method is sirnple, each fire compartment or firezone must be individually calculated.
2 rr^, HAzARDs FoR pEopLE pl
The level of safety depends on different principal factgrs :
- evacuation time, (T)
- smoke density (visibility decreases) (f)- toxicity of sraoke (I)- probability that the fire will start and the physical fitness and
environmental knowledge of the people (R)
- combqsri^bility of materials (C)tk-rlrTtre potential danger for people is defined by :
Pt - (t) (f) (i) (r) (c) t'
iwhere the value of the coefficients. ( . ) depends on the eetinated influencefor people safety of each five precedent factore.
I-39 (`?8ノ
_ (t〕 EVACuAT10N
evacuation tiue ia calculated or estiDated by( I ) provide good information when real data
relation between (t) and the total evacuation
drills. Eupirical forurlasare not available.
tiure is given by the figure
雅的〕 The
■VACUAT10NCO■ FFttC:■NT
R■FUG■
when the occupancyof building includesmore Ehan l0 Z ofhandicppped peopleand when the peopleare sleeping(t) - 2 (oinutes)
inute{25time
■■■■mum
ElgttF9_1 :
8 10
T Total evacuatlon(t) as a functlon of
tlme (mln)
the total evacuatlon tlme T
I-40 9:
(t) SilgKE DENSITY
The smoke production depends on the ventilation and the nature of naterialeused in building fabrics and furniture.
The snoke density is classified in four classes :
F = 0 no smoke hazard (good visibility)p r t small smoke hazard (visibility reduced)
E - 2 reat smoke hazard (visibility < t5 (nl)F = 3 very hlgh hazard (visibiliry ( a (m))
The value F is given in tables established for different uses by the fire Bri-gade; the F value includes all the materials generatry in place and also anaverage of associated smoke production.
(i) SmoKE ToXtCITy
I{e have no real simple data on smoke toxicity during fires. Wtren lre are pre-dieting any danger the value (il is equal to 1.2, when there is no real dan-ger (i) = I
This approach is purposely very simple, but at the presen! time it is irnpos-sible to include fallacious data, the problem is too complex for this type ofapproach.
_ (c) CoMBUSTIBILtTY
The fire propagation depends on the nature of the fuel. The European InsuranceComittee (C.E.A.) has claseified all the materials in six categories.
M. GREIENER give some correlations between building dses and cmbustibilityclasses. Examples are given in appendix.
TABLE I
(C.E.A.) 6
l,0
2
1,4
3
1,2
4
:,0
51C。
(c) 1,6 l'0
I-41 ( {orr)
(T] RISK DUE TO THE OccUPANTS AND THE NORMAL BEHAvIOUR OF PEOPLE
This Jactor includes the hunan influence on fire ignition and also the averagealertness of people (nental, physical, and environmental knowledge).
The value (r) is given by trro other values A and P..
FIRST A ig the probability that the fire will start (risk of acciva-tion). It depends on the normal activity inside the buil-ding.
'l
SECOND P depends on the ability of people
These two values A and P are given for each building use (see appendix I)
TABLE II PROBABILIttY OF IGNIT10N
A
Potentlalhazards'
for people
Pl 2 3 4 5 6
0 0,85 1 l,20 :,45 1,85 2,60
l 0,95 1,15 1,40 :,75 2,35 3,70
2 : l,25 l,55 2 2,90 5,20
3 l,10 :,40 1,75 2,35 3,70 8,70
COEFFICIENT 〔r〕
To sum up we can say that the potential fire hazard for people Pl is :
tg*t f o,7zs 0,750 o,77s 0,800 0,g25 0,875 0,900 o,g25 0,950 o,gls I
TABLE V
I-44υ鉾 )
r vrhen the surface of
. if the comparcment
A-b21. if the compartment
A - \,fi5TABLE VI
the compartment is greater
is acceeeible with the fire
ie not accessible
than l
eng■ne
zo0 (n2)
Жl。6A 0,04 0,1 0,2 0,4 0,6 0,8 1,5 2,5 3 :0
〔g+〕 l l,2 1,5 :,95 1,35 2,7 3,6 4,2 4,4 4,75
E匹」望L望坐ェ璽La聾■9■TABLE VIIb.1(.2〕
400 1000 2000 3000 4000 5000 6000 7000
〔g 〕 : 1,08 1,21 1,34 1,47 1,6 1,73 :,86
〔f〕 SMOKE DENSITY
The s■oke deボsity is divided intOcient 〔f〕 depends oi the estimated
TABLE VIII
four classes ;initial smoke
the variation ofproduction.
the coeffi-
F
F
F
F
eua.
一
・ V
a Q no smoke hazardr I small smoke hazard
- / real smoke hazardr J very hlgh smoke hazard
F is given in appendix f for different
i,
building user.
F 0 1 2 3
〔f〕 l Ⅲ l,l 1,2 1,3
I-45 (tu-')
(K) RISK OF CORROSION BY SMOKE FOR GOODS
Generally there is no risk for properties by aooke corrogion but yhen thisevent is possible, the potential fire hazard P2 is considerably increased :
K-0 nohazard. K - | normal hazard (ln lndustry)
K - 2 :;il"!:t:.:;.""0
(corputer room, elEctrontc equtpment
TABLE IX 1,
K 1 0 1 2
〔k〕 1 1。 0 1。 2 :。 4
'. Examples for the nornal value of K is given in apendlx for different buildinguses.
(A) COEFFICIENT DEPENDENT ON THE PROBABILTTY OF TGNTTTON DUE TO
THE NORMAL ACTIVITY OF PEOPLE AND BUTLDING MECHANICAL SERVICES
A i.s the probability that the fire will start (risk of activation). Itdepends on the'nornal activity inside the building and also on thecurrent building mechanical services (boiler-roorl, electricity, heatingand ventilation circuits).TABLE X
A
(a) 0 r85 1,2 I,45 | ,85 2 160
hazard very small smal I medlum htgh very htgh spectal
To ― up, we can say that the potential fire hazard for 800dS P2 188
P2 - (q).(e).(g).(f ).(k).(a).(c)
Π-46 (toi )
4 r*otrcrtvE AND 'REVENTToN
MEASURE'
The potential fire hazard must be reduced by the appropriate choice of protec-tive and prevention measures. To understand how the system workg, ne ehatl lookqeparately at the five following factors 3
. location oF the building S
. detection and alarm T
. extingubhnent rrEf, E
. smoke control DF
. fire partition and fire resistance F
Each factor affects the safety of the people (Ul) and the eafety of goode (lO)
Ml ‐ Sl Ж TI x El x DFl Ж Fl
M2 ‐ S2 Ж T2 Ж E2 x DF2 Ж F2
See attndIЖ II the ccmplete chart of ERIC fire evaluation system.
LOCAT10N OF THE BUILDING SA/SBTABLE XI
SA ― Availability of extinguishing water
SA
flow rate and Ipressure sufficient
not sufficient
Public fire hydrantPrivate fire hydrant connected withpublic water eupply
Permanent reseryoir' ; 'Non pernaoent resenroir
Electric fire engine + resenroirIdeo with power supplyWater resenre (pond, pool , dam. . . )No water
0。 7
9
B
l
・
S
O
SB
:
0。 6J′
0.9
1
0.65
0.5
Location coefficient :
Sl-S2-SAxSBl{tren the evaeuation time ie greater than 26 minutes, S I - |
fr- 17 (16,,
l_ DETECT10N AND ALARM
l′‐ T l Petecti°n I Man:il:計 re・
tr」董■sion We.8hting
Tl ‐ DT: Ж A x TR + PI (people)
T2 ・ DT2 x A Ж TR + P2 (800d3) . ´
HUMAN DETECT10NTABLE Xll
AUTOMATIC DETECT10N丁ABLE XIII I
S■oke detection■hとn I 口 : Orwhen I チ 1 0r
Flame detection
3
3
一 ′田′
F
F
21。 9
1.52
1。 040。 88
2
l。 9
1。 60
:.3:。 1
. Sprinkler or themalh く 6〔■〕h>6〔 m〕
detectOrs
The'decection coefficient is the greatest of all then we have a cmbination ofdetection systeus
Detection by occupants :
- continuous
- discontinuous':
Presence of safetyguards
- perDanent
- during rhe day
- during rhe nighr
劉
o2
o2
D
I ‥
DT2
1。 2
:。 0
1.3
1。 2
】.3
:.l
].:
:
I-48 t107;
MANUAL FIRE ALARM
TABLE XIV
Not any alarm 1.00
One per compartment I . l0Insurance regulation 1.20
ALARM TRANSMISS10N
TABLE XV
_ FIRE BRIGADE
A
TR
丁ABLE XVI
Arrival time t A B C
tく 10・・
e
fg
c
d
e
f
a
b
c
,d.'e
lo (t 1ts Dn
15 (t <zO un
20 (t 4fo on
t )so un
A: voluntary flre flghtersB : voluntary and professlonal fire flghters
:,t t professtonal flre flghters
i [-ls ( lot ,
Alarm on the totalarea of buildirg(klaxon, bell ...)Just for the fire area
Cal1 internalfire fighters (industry)Call the fire station.automatic call. human call by telephone
Close fire resistive doorsOperate smoke controlOperate extinguishers. autonatic for the fire area. automatic for the total area
t. r0
t.20
r.20
r .401.25
| .25|.r0
_ EXTINGUISHMENT
E: ‐ EXl
E2 ‐ EX2
PEI (people)
PED + PU (goods)
E
x
x
EXTINGUISHERo
INTERNAL FIRE F:GHTERSTABLE XVIII
TABLE XVII
Aglgee!is
. Sprinklersone sourcetl,o sources
. Halons type l30tother
. coz
. Powder
Portable
. l{ater
. Powder
. Noc any, insufficient
Er9'esg
. Normal arrangement
. One per level
. Not any or insufficient
EXl
l.:5:.15
:
0。 8
0。 8
:.00
:.0
1。 0
0。 9
EX2
:.60:.90
1.201.20
1。 20
:。 20
:.2
1。 2
0。 9
:。 2
1。 1
0。 9
:.2
1.2
0。 9
. A11 the tiue
. All the time(elept week-end)
. Ihrring the day
. Ihrring the night
PEl
le5
1。 5
1.25
1。 5
PE2
1。 6
1。 5
1。 25
1。 2
I-50 (げ )
EttigL二主ユE (Only
TABLE
for
XIX
goods)
SMOKE CONTROL
,DFl
DF2
(people)
(goods)
0.26 0。 25 0.25 0。 25 0。 25
Fire fightersclass
TABLE XX
INDUSTRIAL BUILDINGS DFl
l.15
0。 9
DF2
1.:0
0。 9
In agreement withNot in agreement
insurance regulation
PUBLIC BUILDING
R00MS Not any smoke controlNatural ventSmoke ventltechanical ventilationnot any smoke controlNatural ventMechanical ventilationOrtside corridorsNoE any smoke venESmoke ventPiessurized or exterior
0。 9
0.951.051。 05
0。 80
0。 9
0。 951.051.05
0。 901。 001。 10
1。 10
0。 901。 001。 10
CORRIDOR。05。15
。20
STAIRCASE 0。 801。 001。 10
RESIDENTIAL
CORRIDOR
BUILDING
. NoB anY smoke control
. Natural vent
. Uechanical smoke control
. Ortside corridors
. Not anY suoke vent
. Smoke vent
. Exterior or Pressurized
0。 7
0。 951.:51。 20
0.91。 001。 10
0.90。 951.101。 10
0。 9:.001。 10
STAIRCASE
(||〔I-51
_ FIRE RESISTANCE
TABLE XXI sTABILITY OF STRuCTURE
RFA :ノ4h。 1ノ 2 h. l h. :h. :ノ 2 2h.
withoutopen■ng セ
opening betweent":?irttnis
iイ(Sく15〔 m2〕
Sン5(■2〕
1
0。 95
0。 9
0.75
l。 3 :。 6
1。 5
1.4
:.25
:。 9
:。 7
1。 6
1.4
2。 2
1。 9
:.7
:。 5
:。 25
1.2
:.:
The RF coefficient ie the srnallest one between RFA and RFB.
t{eighting PFl PF2
stability
stability
l〔 h〕 0。 9 0.8
1〔 h〕 0。 7 0.9
TABLE XXII INTEGRITY OF FIRE WALLS
Fl ‐ RF x PFl (people) :
F2 ‐ RF Ж PF2 (800d3)
:/4 h. :ノ2 h. l h. lh. :ノ2 2 h.
without openingexcePt front of I I :。 3 1。 6 :.9 2.2the buildingopening betweencomPartments(except fireresistive doorg)
s<0.5 fr2-ll 0.95 t.2 t.5 t.7 t.80,5<S < t (n2) | O.SO t. I t.3 1.4 t.6
PRESS10N ET EIEIIIT INSUFFISANttS A l_A ItOuCHF OU AU F・ 00TEAU D'INCENDIE
AuCUNE IIISPOSItt10N II'ALIMENTAT10N ET IE RESERυ E EN EAUDETECT10N CONTINUE FAR LES OCCUPANTSAEISENCE El'ALARMEPAS IIE CENttRALE EIE TRANSMISS10NEXTINCIEURS INEXISTANTS OU INAIIAPTESROBINETS D'INCENEIIE ARMES INEXISTANTS Ou INEFFICACESAISENCE DE F00MPIERS Il'ENTREF・ RISEABSENCE EIE EIESENFUMAGE DANS LE l.OCAL LABSENCE IIE EIESENFUMAGE EIES CIRCULAT10NS HORIZONTALESArSENCE D'EXuTOIRE DハNS L'ESCALIEROUυERTURES ENTRE NIυ EAUX: S TOTALE 》OU = 5 M20UυERttURES DANS DES CL01SONS AUTRES QUE FACADES CF OU PF 》 OU = l M2
■■■■■■■■■■
RISaUE POuR LES PERSONNES = 姜 2.フ 1 姜
III姜 着姜贅姜・II
姜姜■螢螢贅姜螢贅贅
RISQUE FOUR LES IIIENS = ■ 3.01 ■
■姜姜贅贅贅姜螢姜贅
(ノ ノイ ノI-55
7.0*.'-rs IoN
ERIC Method is an empirical method but based on the practice and experienceof a large group of people engaged in fire protection engineering. Usersare generally non scientific people, fire protection engineers, fire-fighters,fire-authoricies .
this approach is a very sinple tool for conposing different fire rieke orlevel of proEection,r lt ie possible to set values of mininum safety leve1.After having composeh the calculated level R nith such adnissible values,Ire can decide further suitable protection or prevention meaaures.
This method is usefut for fire authorities specially to suggest economicalretrofi t systems when it is not possible to base the fire protection on co-des or existing regulacions.
In the future attempt should be made Eo prove the validity of these enpiri-cal faccors in building fire safety using the already defined nathenacicaland srarisrical merhods ( 5)( g).
APPENDIX エエ FLOW CHART ERIC SYSttEM
EVACUAT10N T■ HE
SmKE TOX■ CITY
PEOPLE BEHAnO▼R
ACHVAT10N
SHЮКE OPACITY
COIBUSTIBILITY
FIRE ■OAD
LEVEL OF FL00R
"IDTl1 0「 ,■RE 20NE
smlc coRRos■ ON
LOCAT10N OF THE BU■LDINC
DETECT10N
…
TRANSlaSSIoH
EXTINGUISIn・ cNT
INTERNAL F=REFICITRS
SraKE CONTROL
F工RE RESISTANCE
∽●ECNC〓
∽日〓コ∽C口〓
Ml".or.tM2oooor
I-56 C'5ブ
8 tTBLT'GRAPH rE
(1) DETERMINAT工 ON DES MESURES DE PROTECT10N decoulant de L'EVALUAT■ ON DU DANCERPOTENTIEL D。 工NCENDIE
selon M6thode de M. GRETENER
A● sociation des Etabliss・ ments cantonaux d'assurance contre l.incendieservice de Prlvention d.Incendie pour l.industrie et l'artisanat (1973)
(2〕 HETHODE D.EVALUAT10N bU DANGER D.INCENDIE DANS L.INDUSTRIE ET AUTRES OB」 ETS
4ё口に Slminaire lnternatioial pOur la protection contre 101ncendieZURICH (1973)
(3〕 EVALUAT■ ON DU RISQUE INCENDIE
H. RAES
Revue belge du feu (oct. 1975)
〔4〕 HETHODE EURALARM― CHOIX DES MESURES DE PROTECT10N
Dr. PURT (197:)
〔5〕 THE EFFECTS OF DIFFERENT PROTECT10N MEASIIRES WITH RECARD TO FIRE― DAMACEAND PERSONAL SAFETY
Staffan BEN6STON
FOUBRAND (Suこde) (:978)
(6〕 A SYSTEM FOR FIRE SAFETY EVALUAT10N OF HEALTH CARE FACIL工 TIES
l. Conalruction oヽtl Combl13tible. Wood Frrd OrdiD.ry
n∞r“
ζL IU"TF“ IRT“薇bITl"| 二1:
Ungrotcqad Prdcded Protected i Fire Re● lot
2 1 2
:| :-7 : |
““02
,
9
゛ 一 一 一
″
2:llterior Filli3b Cia30 C(Corr &Exit)
3. Intcrior Fini.b Ch.. C(R∞ m● )
Clar B I
C:●
“B
C:●
“A
' ff[i1""'..r*.',. "tllfi.tく:/3H7 >:/3く 10 Hr
:(0)・
ン1.O H7
5. Doora toCorr idor
No D●●7 く"Min fr
>"Min frl(0)・・・
>"Min fr aAuto Clo3
2(0,・・・
I *". ,'. I
6. Zon. Dio.nrion. I Mo.c Th.n l(P i
-6(0)・・ | ¨oNo Dead EndB>"'こ Zone Len“ h:●
:m.-150・
EncI●Bed With indicated Fire Re● |“
|°
"卜野yO"|。TI婦 3
く:Hr0
>2 Hr>:Hrく 2 Hr2(0)・
t.Hurrdosa Ara.a
Corri`or Only R∞ m● On17 12 i 3 1 紹r卜l T“ Jξ…
か否かについては可観測で且つ影響の
大きい属性を見落としていないかなど
の吟味が必要となる。例えば,管理状
=を表わす日性として建物用途を用い
てある設備の作助薔率を統計から推定
し,事務的用途の建物について,土`,
という作動率を得たとしても経営が零
細で小規模な建物については同一設備
について全く作勁率が異る場合も予想
され慎重な吟味が必要とされる。
第二に,本質的に不薔定要因ではな
い (建物完成後は確定 している)が設
計段籠では決つていない要因があると
いう点である。鮨エミスにより防火区
面のパイプロ通部の埋め戻 しがなされ
ていないといつた場合がこれに当る。
この種の要因は施工段階でチェックさ
れねば,火災が起るまでわからないと
というこになりかねない.
以上のことから。設計段階の呼価は
勿鍮,建物が崚工したのちの評価にお
いても,諄●された値にはある程度の
ばらつきが存在せぎるを得ないことに
なる。評
“
が有効なものとなるか否か
は,こ のばらつきをどこまで小さく抑
えることができるかに一つの重要なポ
イン トがあるといつて良いであろう。
くE菫省菫員研究所
“
五研究
“
妨火研究
=>
,■ 文 麟12Hコn日 子.「
"笑餃EIを どう●めるか」. 腱
■知=V●
:18.ヽ02.19お=2月
oNel●●●.HE ・・Dir‐l― to:●●r●ve
Application or S,3tem Appro● cb toFire Protectio■ ReOire“ llt●
`Or Bllild‐ing3・・.じ 工N R Panelo口 Fire Saretァ
綺d Joint Mcetinc.19Ъ●C SA.. “interin Culde for CoalOrlel13ea Sソ .ten3 Appr● ●cb to Build‐
:口
`Fire Saretソ
・・. 19″
{|lN.l6, H. E rnd Sbib., A. J., 'ASyrtcn for FirG S.tctt Ev.lortion olll..lth C.r. F.ciliti6", U. J. N. R.P.*l on Fir. sat.rf atb ,oirt Met.i.a, l9!t
No Dericlenc‐
'Sm。““ntrol l NlttF ISm・ 恥
:a““姉
|
M.cb Ar.itt.d Syl.nrBフ Z●¬r
3Bソ COrridor
Multiplc Rour.r10. Effrgcrcy
MoraDantRoutci
ll.M.null FircAl.rm
<! Rout€.-E
¨¨
・2
腱●
IW°ま
lliげ
tJ H°・ 80nttt E=i(→ Dire“ Exit(3)
5
No M.rurl Firc Allrn-1
Maru.l Firc Ahan
IW/O FP Conn W/F B Coin
12 Shot. Dctdion& Alrrm ¨
0
13. Autofi.ticSDrirklcn
mo ち計
71鳳 i:llI:農T●tal S,3Ce
oヽte: ●U8e(0)When item 5 i● -10
“U“ (0)then item 10 i3-0
となるよう考慮されている.
以上の評価では家具や敷物,=は運
用的な事項(避雌誘導.最難訓練など)
は対象となつていないが,こ れらにつ
いては別途義務的に守らね`【
ならない
事項として評
“
システムに添えるとい
う形式をとつている。
ところで,表 1,表 2に示されたパ
ラメータの値は客観的データに基づい
て得られたものではなく,デルファイ
法を用いて専門家のコンセンサスを得
ることにより定められた値である。即
ち,NBSの 火災研究所のグループと
保健医療施設の法律や設計に携わる専
門拿から構成される外部ヨンサルタン
トグループの二つのグループカ1結 成さ
れ,tず火災研究所のグループがデル
ファイ法により防火安全に関係する因
子に対 して暫定的な饉を定め,次 いで
コンサルタン トグループがその値の批
判検討を行うという方式がとられた。
更に,こ うして定められた饉を現実の
.Ue (0) vhan itcm I it bard oa fitrt tl@r &il or onrn unglotactcd typa o( conrtruction...U8 (0) in ronc *itb lera !h.n 3l prtiartr ..U* (0) ?hcn itcm I ia brEd oa an uDprotactcd tyta
in Cti.ling buildin3r. of con3trudioo・●Use(0),ben lten 4:● -10
surancl do coincide or fO■ ■OW para■■e■ ■ines with the work of Other
星[1:RSoEhils:111::t[]tthEns:li[i:ス::i[:L:員:]:]i:ibli二:::ili:ili::[riP_Partialo This statement wi■ ■ be fu】
2. CALCULATION SYSTEMS FOR FIRE HAZARD AND FIRE SAFETY。 。
ur-l
Risk Improvernent by Active Systerns
Relative Protection values for Insurance considerations
1・ EVERT Co WESSELS,Dre
Director,TBBSTechnical Bureau for Loss PreventionPost OffiCe Box 54, 3740 AB Baarn
The Netherlands
Historically, fire hazard evaluation has always manifested-t-
it―
υηl
Self mgre aS an "art" than as a "SCienCe". In the paSt, numerous SySt-ems and schedules have been proposed and tried out (and mostly reject-ed for practical reasons!) in an effort to erect an objective, neutraland scientiflc basls for the assessment of fire hazards and the ca1cul-ation of the effects of preventive measures on fire safety.
It was not before'the l970ts that an approach was developed, tlatls still c.onsidered to be appllcable ln practice. The ploneerlng work,of M.?tGretener has become the source for many very practlcal develop-'ments ln. the last flve years (2). One example ls the EURALARM-study byDr. G. Purt, where a procedure is given to determine what kind ofprotection - automatic extlnguishment, automatic detectlonr none orLottr - is necessary in certain buildings from the viewpoint of equip-'ment, manufacturers (3). More recently, the French development of ERICmay be mentioned (4) also presented at this slrmposium -, while adap!-ations of the r'Gretener-method'r r^rere also published in Austrla (5) andBelgium (6). The "Gretener-philosophy"has in a way - become theEuropean counterpart, of t,he American "decision-tree-approach" (7) .
It has been a loglcal development, that the Worklng Group "Fire"of the European Insurance Commlttee - (Cne, Comit6 Europ€en desAssurances) - i where L7 European countries cooperate in matters pert-ai.ning to, fire insurance, has also adopted the Gretener studies as abasis'for its further work on the objectivation of flre risk assessment.fn the annual CEA "Fire"-meeting of f970 ln Rome it was decided, that atechnj.cal sub-committee should study the relative flre-protection valuesof automatic sprinklersystems and automatlc fire-detection- and -alarm-installations. The sub-committee under the chairmanship of Switzerland,with members from Austria, France, Germany, The Netherlands and theUnited Kingdom, has conducted intensive studies and elaborate discuss-ions with technical representatives of all CEA-member-countries. Aconfj.dential interim-report hras circulated in April L976 and it is tobe expected, that a final report will be presented in the annual meet-ing of CEA "Fire" in 1980.
. In order to understand the lmplications of the CEA-results correct-Iy it is most necessary to have an idea of the lnsurance-part in thetotal safety picture.
3. RISKIMPROVEITENT, PREII{IUMS Al{D COSTS.
Of course, rates and premiums for fire lnsurance play an lmportantpart'in economic consideralions of fire safety. It Ls often assumed,Itrat there should be a direct mat,hematical relationship between thecosts and effects of ri.skimprovement and rebates on the lnsurance-premiums. However, this is usually not correct as is shown in theialculatory examples in appendix A. Although every measure to lowerthe fire rLsk for insurerl- should be and ls taken into account ln therate- and premium-setting of flre insurers two very significant con-clusj.ons can be drawn from the examples given:
a. Because of the nature of'the insurance operation an lnvestment in. fire protectj.on measures can never be fully flnanced out of'savinls on the premi.um. Thls could only be sor !f _and when-avery low lnv'estment ln flre protection would yletd a very largeprol,ectlve effect. However, such measures are almost alvJayg aI-ieady lncorporated ln the offlclal bulldlng regulations. tl:o,a delrease ln the flrethreat to llfe cannot be compensated forin the premiums, as usually this risk is not part of the normalcover, that is provided in the fire insurance contract. The In-
-2- (t.p)
sured, however, - and the authoritles - must indeed be concern-' ed with the protect j.on of life but the coii -of --such meaiuitJ-that are designed for life-safety should not and cannot be borneby insurers. _They can onJy offer-.financj_al assistance againstthe costs of measures trrlt (also) affect ttre aciuei-u"iiiii-costs
b' There is always a certain amount of monetary loss potentiar forthe insured that cannot and is not included in the i.nsurancecover. Premium rebates can again only retfect In"t part of risk-. improvement through fire pro[ectloD, that is subject of the-in-: surance contract. The costs of reducing the incidence of uncover-ed conseguenll?I and- emotional losses must be borne fully uy-tt"insured himsdlf, or by society.' So, alY guantitative values of fire protection neasures that in-surers aPPry must be based on their way oi operation, their statisticsand experiences-. Any application of thl aata given in appendix B musttheref,ore only be seen- in the described context and any conclusj.onsdrawn for purloses oud"ia" insurance should ue aaiftea-accoraingiv.
4. .RELATTVE PROTECTTON VALUES OF ACTIVE SYSTEMS.
The CEA-subcommi.ttee mentioned before has used the ,,Gretener-formula" for the risk in buirdings ;"-ia; starting-point:
R =
where R is a measure of the risk, P is the inherent hazard of the build-ing, A is the hazard of ignition, while N refers to,,standard,, prevent-l3tt,-s-99'special" prote-tive measures and F to the fire resistance ofthe building. .
. e, A and F are not discussed in the context of this paper andbuildings are further assurned to have ',standard,, pioiection, i.e. fireextinguishers, hydrants, watersupply, trained personnel and public fireservice, to make N = l.'Then tte-ieiative risk! in two identicaJ, build-_ings gith identicar use and oper.ai;;.;;n be expressed as R, = ! andR2 = ;, where C is a constant factor.- -2 :ant factor - r ol
. In appendix B S-values_are given for actlve protection measures,as-they have been determined by [ne-suucommittee ind have been checkedand corrected after consultations with all cEA-member countries. Thef igures have no absolute signif icance i"-tt.-".;;;;"t they are nowarready-fully tncorporated Ln any rnsuii.J;- ;;il;g-system in Europe.They only represent at this time a weigtrted averag6 o'r present Europeanthinklng .rmong flre i,nsurance technical specialists on the rerativeprotection value" of fire-protection syst-ms as far as they relate tothe specific operatlons of-flre rnsureis. However, ln The Netherlandsand ln other countries the cEA-s:;;il;"-h..r" been'testea against exist-ing- rating- and tartfication-schedules for commercial and industrialrisks and it has been found, that there usuarry Ls- a distj.nct correl-ation between these "objective' reraiive protection varues of activesystems and the calculation-factors curreirtly used in- the commercialnarket for setting rates for risks to be lnsured.
-3-(13/)
5. CONCLUSIONS TO BE DRAWN FROM S-VAI.UES
As an example a "standard" risk (N =. l) wif,hout any extra pro-tection (S = f .0O) can be represented by Rn = #n'= C- tfhen a guardservice, making its rounds every two hours, is'6ifanized,-one findsin appendix B an S-value of I.05 to apply: R guarded = fii<, andR unguarded : R guarded = lr05 : 1.00 3 an unguarded rlSR-Is consi,der-ed to be 5t less "attractlve" for the insurer than a guarded rlskSimilarly, a fully sprinklered risk with two lndependent water sources,completely installed according to the rules (S = 2.401 ls evaluated astvrice as "good" for insurers as a risk equiped with automatLc flre de-tection (S is I.20) r.,again complefely according to the exJ.sting lnstall-i:}3",11;;.=*i.;.fr and Rd = ft' : lt forlows that Rs : Rd =
. It should be noted - as has been explained ln appendlx A - that.this does not necessarily- imply that the actual insurance rates forsprinklered risks must be half of the rates for risks with automaticfire detectioni_ nor may be concluded, that in any given buildlng
independent of any other factors that do influence the risk -, equip-ed wilf a sprinkler installation the "sprinklered rate" should alwaysbe a-Ifr = 4L.67 t of the "unsprinklered raterr .. An other complicatlngfac€6?"is, that a rat,e of f OO applies to a "standard" rl'sk (N = l),while, of course, in actual practice there does not exist the theoret-ical i'standard" building. The "base-rate" of I00 ls usually alreadymodified by considerations of structure, building materials used,special processes.being carried out, etc. So, again_S-values may not
.be seen ls absoluie, but they do serve a very useful PurPose ln com-paring the protective value in the eyes of insurers - of a varietyirt fiie prolection rtteasures. Finally, the aspect of life safety, thatof course does play a very important role in any fire Protection cons-iddration, is not Lafen tnto iccount in the approach described here.
For general orientational purposes the risk synbol R used heremight be inserted for the burning costs BC in the formula of appendix A:P+IR=BC+OC+RF+TR+IR.The maximum S-value in an optinally protected building is I 4-00.. Abuilding equiped with simullaneous alarm (I.05), fire detection (f'20)'automatic €ransmission of the alarm signal (f.04), a professj-onal firebrigade (f.28) and an automatic sprinkler installation (2.40) can beconiidered to have a maximum protLction by active systens. The totalS-value ls 1.05 x 1.20 x I.04 x 1.28 x 2.40 = 4.02. Rn:Rmax = 4:I :BCmin = t x BCn = * 15. Calculatlng Pmin for the lnstallation of activesystems one finds E maximum rebate of + 50t for maximur Protection.
6. SUMMARY
European insurers have provlsionally reached a degree of _agreementon the reiative Protection values of active systems, expres:9d il "table of ,,S-vaIuLs". However, in the actual application of "S-values"j.n any country modifications tg the table rnight be necessary becauseof naiional insurance pract.ices and differences in national standardsand regulatlons pertainlng'fo the installation of active systems
-4-(ノ32)
REFERENCES.
l. Fire Safety ln Buildings : Needs and Crlterla; ClB-Proceedings,Publication 48, pages 55-50: "The insurance interest in identlf-ying and integrating the various needs for fire safety'1E.C. Wessels - L977.
2. Bewertung der Brandgef5hrdung und Ableitung von Schutzmassnahmen;nach Methode M. Gretener; Brand-Vertrtitungs-Dienst ffir Industrieurld. Gewerbe, Ziirich, L973 - plus earller and subseguent public-ations.
3. Die Beurteilung des Brandrisikos als Grundlage fiir die ProjektierungYg"
automatischen Brandschutzeinrichtungen, G. Purt, EURALARM, L97L.
4. Methode E.R.I.C., Evaluation du Risque Incendie par le Calcul,P. Sarrat and D. Cluzel, Revue Technique du Feur W, lrlay 1979,pages 82-85
5. Rechenmethode der 6sterr. Brandverhiitungsstellen zur Ermlttlung derBrandschutzmassnahmen, Brandverhiitung, Sonderheft, Apr5.I 1973.
6. Evaluation du Rlsque Incendle, H. Raes, Revue Belge du Feu, October1975.
7. see' for. instance: Flresafety Systems Analysis for ResldentlalOccupancies, NFPA/HUD, llarch L977
8. European fire loss statistics compared, Tom Wilmbt, Fire fnternational53, March L979, pages 79-82, table 3 ptus text.
■osses in the future (PRF: premium reserve funds)FTR .= teChnica■ resu■ ts3 the profit (or loss)on the insurance
operation,IR tt TR= total result = RP + IR = turnoverRF = PRF + BCRF′OC = Cc + Ic,CC = commissiOn costs3 payments to brokers′ etc. to compensate
for their acquisitiOn′ consu■ tation and inspection servicesFIC = ■nterna■ costs: the office― and inspection― costS of insurers
themse■ vese
b. SPECIFIC
To demonstrate the mutua■ re■atlon between the monetaty sums ュnthe foェ `1:ula an example has been drafted′ based on offic■ a■ figurestaken from the annua■ report of The Nether■ ands. Insurers Associationof 1978。 These figures re■ ate to the total non― life insurance marketin The Netherlands′ ュncluding fir:こ
1 [illillte習 :t:II′ thと91。
.1:尋II:nCe.The effects of reinsurance have b(equation can then be given in m■ ■■■ons of Dutch gu■ ■ders: ・
ors r00 + $ = 65 + 29 (r3 + 16) + tt G + 7l + 3 {-5 + g) MuPIRBCOCRFR,where llU stands for any rmonetary unit,'.
c. INE.LUENCE OF PREVENTTON oN PREMIUM.
If lt is assumed, that a certaln fire protection measure lowersthe burning costs (BC) wlth 20t - to 52 - the turnover (p + IR),the commission cost,s (CC) and the premium reserve funds -(pRF) willbe lowered by a factor x. The j.nternal costs (IC) witl be somewhathigher, because of'the need to control and lnspect the flre protect-lon measure: an lncrease of IC rnlght for lnstance be 6t. ff lhe
-6-(なの
burning costs reserve funds (BCRF) are assumed to remain the sameand if the insurer wants to maj.ntain hls total result, the follow-ing equation results:r00.x + 8.x = s2 + l3.x + L7 + 4.x + 7 + 3; it follows that x is1 0,87 and this brings the equation to:85 + J = 32 + 2g + f0 + 3, wherethe insured has been given the beneflt of the difference in round-ing offSo, if BC aoes from 65 to 52 (2Ot risk improvement), p goes fromf00 to 86 (f 4t rebate). It i.s obvj.ous, thit only a part (7Ot) ofthe risk improvement can be returned to the inslrred- in premium-'savings
If we assume a rj-sk improvement of 40t (instead of zot in theexample above), the eguation becomes:
72rS + 5,8 = 39 + 26,4 + 9r9 + 3.
Hgre P goes from 100 to 7215 (Z7 rlt rebate) for a risk improvementof 40t; only 59s of the risk improvement shows up in.,the lowerpremium.
' In. the very theoretical case, that the fire hazard is completelyelj-minated, the rebate can only be 65t; of course this is an irre-alistic sj-tuation, as insurance has lost his meaning long Ueforethis "absolute safety level".d. COST OF PREVENTION AND BENEFITS TO THE TNSURED.
rn order to construct a calculation-model for the evaluation of-th" cost,/benef it-aspects of active f ire protection systems, Iet itbe assumed rhar rheie are rooo idenri.car-;;ii;i;;;, each varued at1000 Mu- The totar value of this class of buildiigi is then onemirlion MU (and this should arso be the insured varue).
Furthermore, i.t is assumed that in a given year the followinglosses occur:
I total loss = 1000 !,tU4 losses of L2\t = 500 MU20 losses of 5t = f0O0 MU
25 1o,seS = 2500 Mu: the average ■oss = loo Mu.
T:[ ::I]li:jeT:ulh:n ::甘::臓38=ユ:号85 oテ ::: MU = 3846 MU′
the prOmium
The insured wants to lnvest. lt of the buildlng-value in a fire pro-tection system: this would te an investment oi l0 MUi the annuarcost fqr the insured wourd be: depreciatior, (r03) = i r,ru pruslnt€rest (l x fot) = 0r5 tr{U, totai t.5 !,tU. .
The fire protection system might ylerd a zot or 40t reduction inthe l0ss-hazard. The indlviauil ri=i l"i u,rirding is eguar to thetotar burning costs divtded by the nriu.r of uuiidings.
-7-(/」えf「シ
The followj-ng sohedule results:
Risk ■mpr ovement Prernlum Cost Risk
:Q t20r40r
3r853r3l2 r79
0lr5r.r 5
2r52tOlr5
If decisions about fire protectlon were only to be based on thepremium-, cost- and risk-relation, these figures lmply, thatthere rrrould be very little incentive to take any measures. However,it should be remembered, that consideratlons of life safety are'eli-minated from this approach, while in "real llfe" they certainly.are a very i.mportant factor
However, apart from life safety, there is one more loss-aspectthat is usually not or insufficiently taken into account in thlstype of calculation. The monetary value of direct losses can usu-aIIy be insured and so can consequential losses, such as a decreasein earningiir reconstitution of files, etc. There are also lndirectlosses, bofh material and lmmaterial, that are associated with anyfj.re and that are not subject to insurance-cover. Estimates of the-se indirect losses have never been very accurate, but the most re-cent publicatj.on on the problem citesan average - all over Europeof indirect losses to be 50t of the direct losses (8). If thisfactor is taken into account a different picture emerges:
40t2 r79I ,500r75
These figures show, that the investment does pay off in real moneyif the risk improvement would be 40t. Again, life safety aspectsare not included but they will certainly be influenqed very favour-ably by the installation of the fi.re protection system.
Note: These calculatj.ons are only meant as numerical examples ofEf,-e main considqrations, that should play a part in any decislonabout fire protection. Most figures assumed here can be replacedby practical data and analogous conclusions can be drawn. The larg-est problem, of course, ls still the estlmation of the "Ioss Potent-ial" of a class of buildings. Statistlcs on loss results ln sprink-Iered and unsprinklered buitdingsr,in premises.equiped or notequiped with fj.re ddtection'installationsr €tc. r are bei-ng compiledby the insurance and the manufacturing industry. These statisticscould very well provide factual support to future calculation-models.
20r3r3trr50lr00
0r3, 850L r25
Risk improvementPremiumCost of protectionIndirect loss
The efficiency of autornaticdetection and. e:<tinction isconeidered to be a firnction ofthe operation of the public fireservice. S-values have there-fore been determined iIl relatioawith t:he I'E-factoril for'thepublic fire service. r' (t.j7 )
Centra■ station in bui■ dingAutomatic transm■ ss■onsimu■ taneous a■ armSecur■ty line
INHOUSE FIRE FIGHTING
Fire fighting teamsVo■ untary fire serv■ ceProfess■ ona■ fire serv■ce
AUTOMATIC EXTINCT工 ON
sprink■ erinsta■lation c■ ass I
ュn re■ ation w■thpub■ic fire service
Sprink■erinsta■■ation c■ass II
in re■ ation withpub■ic fire service
::1:1111:t:11ま :1。n l
FIRE VENTILAT工 ON )
1
2
3
4
5
6
1
2
3
4
5
6
E
E
E
E
E
E
E
E
E
E
E
E
C FIRE SERVICE
E time (min) distance (km)1
2
3
4
5
6
4 10
'10 - 15ン15 - 20
'20 - 30,30 - 40>40
1
3
6
01
>>ン>
(■- 3- 6- 10- 15ン15
-9-
FrRE September,1982 157
o Over the past two years a "Fire safety evaluation scheme" has been developed in the UK,essenrially for patient areas in hospitals. This background to the scheme has been prepared byPaul Stollard, a research associate with the Oepartment of Fire Safety Engineering, Edinburgh
University.
Developing an easy―tO口use systematiё「
technique of determinino■ re safetyLJOSPITALS. by their vcry nature :rcI I occupied by the ill. thc aged, the infirm.thc handicapped and thc immobile. Therefore,they pr6cnt spccial problcms whcn consideringfrresafety. In addition, they normally conoin
. potentiatly hazardous cquipment or materialswhich are rhemsclves a fire risk. ln psychiatrichosgitals thc problem is funhcr acccntuated asthcy are often housc patienls who may starthres deliber-ately.
Many hospitals still in usc datc from thc lastccn(ury, and were not designed with frre safetyin ,:rind. As the economic constraints on thchcalth scrvice increase, less money is beingmade availablc for new buildings and so thcre
' is an incrcasing need to upgrade thc existinglbuilding stock so tha!- it can conainuc to
- gtovide adequate and safe accommodation.Improving thc frre safety of existing
buildings does not automatically mcan thcptovision of morc hrc doors or hreixringuishcrs. but it does mean a carefulevduirion of all factors contributing to firesafcty so as to identify problem areas. Asystematic evaluation technique applicdconsistently to hcalth buildings will cnablcavailablc moncy to bc uscd to-thc gre.t6tbcncfit.- - A method of cvduating thc frre safety ofpatient areas within cxisting hospitals has bcert ..=derclcae<i by thc Depanment of Firc Safety ':-
E:rgineering at the Univcrsity of Edinburgh,under the sponsorship of thc Depanmcnt ofHcatth and Social Security. This schcmc secksto :rss6s exactly which patient arcas fall belowan acceptable standard of fire safety and toindicate how imprnverrents could be made.
Twcnty indepcndent componcnts of ftresafety have bccn identified (see Tablc l) and for
- each one a worksheet has been prepared. Thcworksheets are designed for usc by thehospital's firc prevention 9ffr631 61 worksoffrccr, and a detailed survcy should takc aboutonc hour for each particular watd or nursing
. area. Thc numcrous sub{omponen6 of eachj of thc twenty componcnts are listcd on the
worksheets along with an indication oi thcdcsircd standard so tha( the asscsor can gradecach componena bctwecn 0 and 5 for each
Table 1: complonantt o{ frrc refcty.
Ol StaffO2 Patients and visitotsO3 Factors affecting
smoke movemgntO4 Protected areasO5 Oucts, shalts and
cavitisg06 Hazard ItotectionO7 Interior frnishO8 Furnishings09 Access to protected
systemsl6 Signs and frre noticss17 Manual frre-ftghting
rquipment| 8 Escage lightinql9 Automatrc suppression2O Fire brigade
Peccentagecontribution
96
76
4756
445553
51
3534
nursing scction. A scorc of 5 indicatcs nofailings, 4 a 20 per cent defrciency and so ondown to 0 shich rcprgenas a 100 pcr centdeficiency. To achicvc a value for thc.overallfirc safety of thc survey arca th€ twcnty survcygrades are muitiplicd by thc pcrccntagc conrri-butions of thc components to frrc safety (sccTeblc l).
Thc pcrccntage clntributioru of thc twcntycomponents to hrc safety wcrc calculated byconsidering thcir relativc contributions to theachicvcment of a sct of frre safety tactics, andthrough thcsc to thc attainmcnt of threcobjectives which constitutc th€ ovcrall hrcsaiety policy.
The tacdcs considered werc ignition,prcvention. fire control, ctress, refugc andrescue. which ari thc technical measuret
availablc to mcct thc objectives of 6re safctyfiife safcty, mission continuity and prop<typrotcction). Thc interadions bctwecn thedifferent componcnc have bccn carcfullyenalyscd and takcn inlo account whcncalculating thc valuc of relative pcrccntag€contributions to fue safay.
Thc multiplication of the survey grades forcach componcnt by thcir percentagc contribu-tions gives a sct of products which, whensummcd, will givc a iinglc score betwecn 0 and5fl). This rcprerents a me:rsure of the overallfrre safcty of thc particular patient area. Ascorc of 500 indicatcs tha( th€ dcsirablestandard has been achieved. whilc 400 indicatesa 20 pcr cent residual dehciency and so on undl0 which r€p.es€nts a totally deficient situation.
Although any score of lcss than 500 indicates
Evelua tion workshee t - O3 factors a ffecting smoke movement
Oefrnition:Constructional components which affect air movemsnt in the survey volu,n6.
Make an initial assessment of the 'leakiness' of th€ building by the followingcalculation;'leakiness' - l'windows' +'doors'l x'fitting' x'position' x'exposure'
-The 'windows' factor is derived from the following table:
j:;: . :T Area ol -r!: : Percentage oF oreneole winoow=rea100‐ 80% ‐ 80‐ 6096 ~ 60‐40% 40ワ6●′:ess--windows {sqm}
up to 7575‐ 150
150‐ 225225‐300
1.5 21 1.5
0.5 10 0_5
2.5 32 2.51.5 21 1.5
-The 'doors' facior is;O.9 if O- 5 doors to the outside0.6 if 6-1O doors to the outsidsO.3 if I l-15 doors to the oirtsid€O if over 15 doors to the outsade
-The 'frtting' factor is;1.1 it the windows and doors are sealed or weather stripped1 if the windows and doors are tight fttting
O.9 if the windows and doors are badly fttting
It the survay yolume is a basement tha 'position' will be 1.1
-The 'cxposurc' factor is;1.1 if the survev volum€ has an external wall on on. sids1 if th€ survey volume has external walls on two side3
O.9 if the survev volume has external walls on three sidasO.8 if th€ survey volume has extarnal walls on four sides
N.B. It tha surv.y volume has no exte.n€l walls thcn the 'leakiness' will automaticallybe4or5.This initial assessnl€nt can ba modifred depending upon yout evaluation of lhefollowing sub-components;Al Assess any existing forms of smoke control8t Assass the number of open windows and doors at iama of survey.cl Assess the degr€€ and integrity of internal sub-divisions in th€ survey voiuma
le.g. open ward ot singlc bed roomsl.Asiess tha surrounding terrain and direction of prevailing wind.Assass tha vzntilation sYstom.
0,0
position of survey Number of storeys in buildingvolumeinbuilding 1 2 3 4 5 9 over 9
gfosnd floofft.st tloo.second floorthird lloorfourth tloorfifth floorsixth floorseventh flooreighth floorninrh floor and over
tome measurc of deficiency it is imporrant tocstablish at what lcvcl this inadequacyconstitutes an unacceptable situarion.
Through a series of freld rrials andrcpcatability tests carried out over the lasr twoyears it has been possible to scr this level ofacceptability at 350, with scores of below thisbcing considered as unacccptablc. Any score ofless than 280 must be regarded as defrnitelyunacceptable, and only scorcs of over 450 canever bc considered as good.' This scale of acceptability can be illusrraredby figurc l. The limits of accurasy of the survcytechnique employed mean that borderlinescor6 of approximately 450, 3J0, or 2EOcannot bc definitely assigned to panicularcatcgories. The evaluation scheme wiUlhcrcfore identify the worst patient arcaswithin a hospital, or a health authority, andindicatc whcther thcy are acceptablc or not.
This should enablc thc rcsources available tobe spcnt where they can be of most benefrt, andit is imponant to rccognisc that the objcctivesof all improvement work should bc to raisc lhefire safety scorc to ovcr 350.
Thc results of thc surveys will indicatc inwhich of the componenrs an individual ward ispanicularly defrcient, and cnable improve-ments to bc conccntrated in these arcas. The
Figure 1: levels of acceptability
500
450
3so
2AO
Good
ACCEPTABLE
UNACCEPTABLE
DefrnitelvUnacceciable
FIRE, September, 1982
frgurcs for the percentage conrribution of eachcomponent indicatc which will give the biggestincrease in overall fire safety for a.one pointimprovcment in thc survey scorc.
For examplc, improving thc Bradc for thestaff component from a three to a four is aseffective as jmproving lhe manual frre-fightingcomponent from a score of rwo to hve. It isthercfore possiblc to ensure thar the availablcfrnance is where it is most needed and where ityields the Ireatest improvement in safcry.
The principal objective of thc evaluationschcmc is the provision of a sysrcmatictcchniquc of determining fire safety, which isers]' to usc and repcarablc. It is hopcd that thetwcnty worksheets, along with a fullerdescription of the schcmc and derails of itsapplicarion, will be published in rhc form of aHospital Technical Memorancium, sponsordby thc Departmcnt of Health and Social
Security, and published by the HMSO later thisycar.
This will provide hcalth authorities wirh auniform mcthod of assessing frre safery andmay help to cstablish uniform srandardsinitially throughout England and Wales. lt willbc possiblc to identify wherc debcicncics dooccur and to be reasonebly sure than anrcceptablc standard has been obtained onceimprovemcnts are madc.
Acknowl.dgsmnttThe sponsorship and financial suppon of the
Dcpanment of Hcalth and Social Sccuritl'hasbeen invaluable, and thanks arc also duc ro DrEric Marchant (Senior Lectorer in the
.Departmcnt of Fire Safety Engincering,University of Edinburgh) who has direcred thcoverall project and assisred in th. pr.parationof this aniclc.
Evaluation worksheet-O2 patients and visitors
Deftnition:Patients are all people receiving trestment in the survey volume.Visirors are those people other than staff allocated to the survey volume or patients.Make an assessment based on the mobiliry of the patients as tollows:
Patients able to walk without assistancePatients able to walk il assistedPatients needing blanket or mattress evacus:ion =Parients requiring bed evacuationPatients who should not be moved =
This initial assessmont can be modified depending upon y(rur evaluation of thcf ollowing suFcompon€nts;Al Assess the psychological condition of the patientsB, Assess the ability of visitors to assist or impede evacuation of patients and
assimilate this in the assessment. How many are'likely, at whst times, of whatages end physical abilities, and how well they will know th€ escape ?outes.
Cl Assess the likely increase in tre hazard due to smoking.Dl Assess the responsiveness of patients to instructions G.g. consider sedation,