- 527.54 KB
- 26页
- 1、本文档共5页,可阅读全部内容。
- 2、本文档内容版权归属内容提供方,所产生的收益全部归内容提供方所有。如果您对本文有版权争议,可选择认领,认领后既往收益都归您。
- 3、本文档由用户上传,本站不保证质量和数量令人满意,可能有诸多瑕疵,付费之前,请仔细先通过免费阅读内容等途径辨别内容交易风险。如存在严重挂羊头卖狗肉之情形,可联系本站下载客服投诉处理。
- 文档侵权举报电话:19940600175。
'毕业设计(论文)外文文献翻译文献、资料中文题目:消防系统运行可靠性的估计文献、资料英文题目:文献、资料来源:文献、资料发表(出版)日期:院(部):专业:班級:姓名:学号:指导教师:翻译日期:2017.02.14
附录一:中文译文消防系统运行可靠性的估计在过去的三年屮,美国国家标准技术研宂所(NIST)已经在研宂开发一种新的加密标准,以确保政府的信息安全。该组织0前正处于为新的先进加密标准(AES)选择一个或几个算法或数据打乱公式的开放过程的最后阶段,并计划在夏末或秋初作出决定。此标准内定明年实施。RichardW.Bukowski:体育,高级工程师,瑟斯堡建筑及消防研宄实验室的MST,美国医学博士20899-8642;EdwardK.Budnick:体育,巴尔的摩休斯联合公司副总裁,美国医学博士21227-1652;ChristopherF.Scheme1,克里斯托弗计划1,巴尔的摩休斯联合公司化学工程师、美国医学博士21227-1652;前言背景资料:为执行特定功能而设计和安装的美国消防计划。例如,自动喷水灭火系统目的在于控制或扑灭火灾。为此:自动灭火系统必须长开,即能满足火灾地所需水量达到控制或消灭火灾,火灾探测系统是为了尽早提供火灾预警通报来通知楼内人员安全逃生,并提供消防通知,使其他的消防组成部分开启(例如,特殊灭火系统、排烟系统)。两种消防系统启动(检测)和(警报)必须达到尽早报警。建筑防火墙的一般设计目的为:限制火灾蔓延的程度和保持建筑物的结构的完整,以及在火灾发生时保护逃生路线的安全性。为了做到这一点,特殊的消防系统必须按标准测试及保持特殊消防系统完整性的特点.。消防系统的组成部分如探测系统、自动灭火系统、防火墙的可靠性,在于提高基T设计基础上的联合演习的细节分析的投入。在安全系统方面,有几个可靠性要素包括有效和能使用的可靠性,运行可靠性能提供一定程度的概率,即消防系统在需要吋运行。运行可靠性能在特定的火灾情况下利用起特点成功完成起任务的一种检测手段。前者是系统组成和可靠性的评估,而后者是系统设计适宜性的评估。这项研宄的范围仅限于运行可靠性的评估,其主耍原因是在于来自文献资料内容的可靠性。除了这项业务区分可靠性和性能,无条件评估的可靠性和故障估计的研宂范围也会在失控的火灾中列出。在该文件的后面将会提供这些条款的讨论。
研究范围:这份文件中提供了关于(1)火灾探测(2)有限范围内的自动灭火(3)防火墙的运行可靠性和执行可靠性的一些观点。一般而言,火灾检测的可靠性大都在于烟气检测或火灾报警系统。自动喷头构成了大部分的自动灭火的数据,防火墙包括分区防火和围墙的完整性。应当指出,在某些情况I,该文献不会超出一般"火灾探测"或"自动灭火"的范畴和要求假设具体类型消防系统.几项研宄报告估计了火灾探测的可靠性和自动灭火系统计划。然而,对被动防火系统如防火分区的详细评估很少被发现,如根据有限的统计资料经分析后,被用来归纳包括评估和不确定的关联性等信息。后者的作用仅限于文献资料在检测和灭火时的评估。防火分区的可靠性也包括与之关联的不可靠数据。这份报告列出了与放火系统相关的可靠性原理。为了冋顾分析和重要发展以及数据概括,在文献检索时被完成。该文献中适用于喷头、烟雾侦测系统可靠性的数据已经被分析筛选。这些数据是描述防火系统运行可靠性在均值和95%的置信区间时的可靠性。可靠性分析的原理在文献屮的数据可靠性和相关分析上有很大的变化。棊本上,可靠性是一种概率的估计,即一个系统或其组成部分在一定吋间内按照设计正常运行,其组成部分在正常运行或预期寿命的时间中。这一时期是“改写”的一个组成部分,是毎次测试都发现是运行正常的一个时。因此,系统及其组成部件越经常测试和维修保养,他们就越为可靠。这种形式的可靠性就叫做无条件。系统正常运行的可靠性是无条件的概率的估计。有条件的可靠性是对所提及的两件事情的估计,即发生火灾和消防系统成功运行在同一个吋间内发生。可靠性估计并不认为火灾发生的几率是无条件的估计。涉及到运行可靠性的其他两个重要概念是安全故障和危险故障。无火灾发生时,消防系统却运行叫做安全故障。一个常见的例子就是一个烟雾探测器的假报警现象。发生火灾时而消防系统却不起作用,这叫做危险故障。在这项研宂屮不能有效使用的概率(1-可靠性估计)称为危险故障。火灾期间自喷系统不能运行或者运行系统不能控制或扑火火灾都是这种类型的失误。整个系统的可靠性取决于各个组成部分的可靠性及其和应的失败率,系统组成部分的相互依存性,安装后系统及其组成部分在维修和测试时所出拒的评估。考虑到关键的可靠性时也涉及到消防系统的性能。系统性能被定义为某一特定系统的能力,为完成其
设计安装的任务。例如:被评估为性能分离的系统,是基于在火灾期间各个组成部分在保持建筑物的构造和防止火灾蔓延时的作用。系统性能根据其各个组件控制火灾蔓延的程度来界定。性能可靠性评估所需耍的数据在于,消防系统在一般和大规模火灾情况下完成设计A的的程度,性能可靠性的数据通过复检这些数据的来源。因为这些作用取决于显示数据的内容,因此,这不是某单方面的作用。各种类型系统失败的原因通常分为几大类:安装错误,设计错误,制造/设备缺陷,缺乏保养,超过设计限额和环境因素,有几种方法可以利用以减少失败的概率,这些方法包括:(1)冗余设计,(2)积极监测故障,(3)提供最简单的系统(即最少的部件)为解决危险,以及(4)一个设计检验、测试、维修计划。这些运行可靠性的概念都是重要的,当运行可靠性评估在温宪忠报道吋,因为在某一分析中用到的资料,可靠性评估可能用到一个或多个上述概念,在这一范围内阅读这一文献时可酌情处理,大部分数据是从支持这份论文的文献中获取得,这些文献却符合在无条件运行可靠性!文献检索文献检索是搜集各种类型消防系统可能性的数据,这些数据被认为与安全计划冇关:自动灭火,自动检测,和消防隔离。文献检索的目的是获得特殊系统的运行可靠性评估,这些特殊系统中每一种类型的消防系统都为一般的居住物(如住宅,商业建筑和公用建筑)。信息来源包括全国火灾事故的数据资料,美国国防部安全记泶工业和住房的特殊研究,工业保险历史记录和检査报告的公开文献和试验数据。试点工作和火灾测试结果的报告只有在火灾探测、自动灭火或者防火隔离计划时被明确评价是被利用,测试系统用于资格核准或列表,并且用于审查失效方式的资料,英国公布的数据也包括日本、澳大利亚和新四兰在内。常识多个基础广泛的研究报告指出,这份调查是关于火灾探测和火火系统还有防火分区的可靠性。这些包括(1)火灾研究[1996]托比在英国(2)澳大利亚消防工程索引[消防法改革中心、1996](3)日本东京火灾统计汇编[东京消防处、1997](四)日本研宄消防系统根源的成果[渡边1979]。
托比消防研宂所致力于解决消防系统的可靠性和各组成部分的相互作用。德尔菲方法是一种用来揭示各个组成部分单独使用吋的可靠性估计。组成部分包括:火灾探测、报警系统、灭火系统、自动排烟系统和被动防火(如防火隔离)。澳大利亚消防工程指导守则提出了工程法规依据了新的工作标准,即澳大利亚消防工程法规。在这个方法的指导下,为燃烟、燃烧但无火花的火焰、和燃烧又有火焰建立防火安全性能评估。消防系统的工作情况(即探测概率、灭火或控制火灾)完全根据各个特殊系统运行可靠性来预测。在这份指导手册中可靠性评估来自一个专家小组而不是来自实际数据。最后,运行可靠性的数据分别在H木被两个不同的研究小组公布,一个研宄小组涉及东京从1990-1997年间的火灾事故评估[东京消防处1997]。另一个研究小组涉及F1本全国从早期到1978年为止的火灾事故报告评估研宄[渡边1979]。表1概述Y这些研宂提供Y可靠的估计。单独的可靠性估计存在个别差异取决于这些估计所用的参数。因为消防系统需要准确预测未来的运行性能,从这些研究上导致的可靠性变化,将引起结果的显著改变。此外,不确定性伴随着一种单一的可靠性评估或者在这些推导可靠性的方法中存在某种潜在的偏见,可能限制它们在消防系统中研究运行可靠性或可靠性性能的指导作用。表1:消防系统运行可靠性评佔的公告(成功率)ProtectionSystemWarringtonDelphiUK(DelphiCroup)FireEngGuidelinesAustralia(ExpertSurvey)JapaneseStudies(IncidentDiila)SmolderingFlamingSmolderingFlaming/FlashOverTokyoFDWatanabeheatdetector089090/959489homesmokealarm76796575/74NANAsystemsmokedetector869070SO/659489beamsmokedetectors86887080/859489aspiratedsmokedel.86NAW95/95NANAsprinklersoperate955095/9997NAsprinklerscontrolbutdonotextinguish64NARaNAsprinklersextinguish48NA96Samasonryconstruction8129%probabilityanopeningwillbefixedopen95ifnoopening90ifopeningwithautocloserNANAgypsumpaniiions6929%probabilityanopeningwillbefixedopen95ifnoopening90ifopeningwithautocloserNANA由于在一般的文献中可靠性估计的使用性有限,审查文献是扩展了它的作用在(1)建立一个完善的原理,该原理是关于被认为能影响可靠性的三种策略,并且(二)确定并评价关系到单独系统可操作性和故障率的一定数据。自动灭火系统(即洒水系统)表2概述了一些研宄报告估计,评价实际火灾事故中自动洒水系统灭火的运行可靠
性。作为一个群体,这些研宄报告差异很大,在时间周期、房屋类型和详细程度关系到火灾的类型和洒水系统设汁。表2所显示的自喷系统的运行可靠性估计一般相对较高,而一些研究提出把火灾控制或火灾失效,作为可靠性评估的一部分,但该报告的数据却并不一致。因此,运行可靠性假定为限喷洒操作。评估也应显示价值范围,暗示不宜使用一个自喷系统可靠性而不注意数据的偏差和一般的从不同数据库不确定性数据源相结合。OccupancyReferenceReliabilitydueCommercialMilne[1959]Aut^mancSpnnklcr[J?70]KliUer0974】DOE[1982]98-9Mavbee[1988)yy.)I况61TayJnrpPPO}81JSprinklerFocus[1993]98.4-95,8Linder[1993]96GeneralBuildingResearchEst.[1973J92JMiller1974]95.8Miller1974)94.8Powers[1979]96.2Richardson(J985)96Finucanee(al.[1987J969-97.91Marryat[1988]99.5ResidtnfialMilne[1959]96.6InstitutionalMilne[1959)96.6原预算表2由可靠性估计范围由81.13%到99.5%[泰勒][maybee,marryat]081%的偏低价值与泰勒的研宄中和一些被kook估计过高的(即87.6%)的报告,这些出现重大偏差的数据在这些研宂屮使用。在这两种研宄屮,发生火灾的次数十分少,并且在数据库中不区分自动火火系统和其他的火火系统。最终maybee和marryat报告中的99.5%高估计反映了自喷系统在检查、检测和维修是严谨的和有案可稽的。在自喷系统可取得的数据中,另一个重要的限制是大部分的自喷系统包括记载喷水的事故。在这些研究中,很有限的事故数据也参考了快速反应或适宜的喷水技术。在评估适宜喷水系统的可靠性时应特别关注几个因素,包括(1)允许复盖范围内(2)供水能力较低(3)在火灾中无遥控或警报系统的潜力很大。基于此,还冇与这些技术(如维修水平)相关的其他因素可以直接影响这些类型的自喷系统的运行可靠性。另外,还需要解决这些问题时的系统数据,但基于后来的观察和一般住宅一般不太可能保持正常,一些旨在保证住宅自喷系统运行可靠性的东四可能被降低。火灾探测或警报系统表3提供了一份关于用于住宅系统运行可靠性分析的概述,评估包括平均可能性和95%的置信区间都是基于HALL[1955]提供的数据所预估的。平均可靠性估计的范围从68%
至88%不等。这些标准同托比德尔菲研究所所提供的可靠性数字相一致。然而,95%的置信区间的一般范围为66%至90%。表3:烟雾探测器的可靠性分析[HALL,1955]OccupancyProperlyUseMeanReliability(%)n■1095%tipperConfidenceInterval95%LowerConfidenceIntervalResidentialApartments69J6^.968.7Hotels/Motels77.879.376.4Dormitories86.388484.3CommercialPublicAssembly67.969.865.9Stores&Offices71:773.569,9Storage68J70.0663Industry&Manufacturing80.281.379.1InstitutionaiCareofAged84.986.6833Careofloung84.086.381.6Educational76.979.674.1Hospitals&Clinics83.385.481.2PnsonsJails84.285.982.5CareofMentallyHandicapped87.590.384.9防火分区依靠各种类型器材的功能例如:门(包括固定器材)、墙壁、地板/天花板、渗透孔、玻璃窗、防火卷帘、防烟材料和建筑物。当防火分区被认为是防火计划中的重点时,在文献中有很少的数据认为单个组成部分的运行作用于防火分区。单个为建筑的评估和运行可靠性在WARRIGT0N的研宂屮和澳大利亚消防工程索引屮被提到。但这些评估是完全基于专家的判断。因此不会提供更加深入的分析。统计数据和不可靠估计文献资料概括了先前部分提供的描述自动喷水系统和火灾探测可靠性评估的信息和数据。自动喷水系统可靠性的数据有几个出处,火灾探测的可靠性评估仅来自一个会议,IIALL[1944]o这个会议包括十年(1983-1992)的可靠性评估和列出丫在文献中搜到的综合可靠性研究。这份文件的最初一个0标是提供一个关于所研究的系统运行可靠性评估的预览。为自喷系统和火灾探测,它基于现实数据做了一个统计分析。自动喷水系统分析表2中关于自动喷水系统可靠性分析,是根据每一种居住类型来分析的。应当指出,只有一个出处[MILNE,1959]提供了关于公共建筑和居住房屋的可靠性估计,并且这些早期数据没冇提供现代住宅喷头技术的可靠性数据。图1的分布直方图列出了每一个住宅类型的可靠性估计。平均值和95%的置信区间的限制是适合一般住宅(在研究中不区分商业建筑、住宅建筑和公共建筑的类别)和商业建筑,并且适用于综合楼(商业、公共建筑、住宅类)的可靠性评估。这些结果列在表4。ResidentialInstiturionalCommercialGeneralCombinedn=IJl=1n=9n*7n=1896.69088.1<93.1<98.193.9<96.0<98192.2<94.6<97J
76543F2Q7籌100.62222Instutional•400.纹ProportionperBarRELIABILITY(%)RELIABILITYPrcponknperBar-O0.B7654.3.21OO.OO.0.O.OoO.1S4321ResidentialRELIABILITY(%)pqporvonpefBarqq.8.6.4.211—o5oo}00.76543255图1:自动喷水系统对各种住宅类型的可靠性评估关于商业建筑和公共建筑可靠性评估的平均值控制在其他住宅类型的95%置信区间内。适宜居住和公共建筑的单个点估计,增加了一些与运行可靠性有用的东西,也増加了数据库的容量。用18估计四个独立的门类。然而,关于住宅和公共建筑的点估计不应单独使用而作出任何结论。关于商业建筑、住宅和综合建筑的可靠性估计提供了一些有用的信息。基于对喷淋系统分析的可利用数据是运行的可靠性估计超过88%,如果不考虑商业建筑,喷淋系统的可靠性可达到92%以上。然而,判断这种特殊的喷淋系统与那些评估屮提到系统是否相似是十分重要的。商业建筑的的可靠性范围在80%至98%,而一般建筑的为94%至98%。
火灾探测系统分析关于火灾探测系统可靠性估计的数据是全面的。这份数据跨越了十年,并且每年都做可靠性评估报告(反映在表3),它为了各种不同用途的房屋而完成。这里的分析根据房屋的用途把它们分为几个建筑等级。每种用途的房屋得出数据后,然后计算每种房屋的可靠性估计。图2显示了所冇烟雾探测器关于全部住宅类型的可靠性估计。
30201060708C90RELIABILITY!:%}i—Ao0.0100proponisperropr图2:烟雾探测器对各种住宅类型的可靠性估计如直方图中所示,数据有一个双态分布。因此,为了进一步研宄两个平均值完全不同的数据库,进行了一个方差(变异数)分析。变异数检测了可靠性估计的平均值和对一个给定建筑类型的可靠性影响。图形代表性的变异数以最小二乘法的形式在图3屮体现。变异数影响最终结果。如图3所示,三种住宅分类分别冇不同的关于烟雾探测起的平均可靠性估计。图4中包含的直方图分别描述了每种住宅类型的可靠性估计。///OCCUPANCYLeastSquaresMeans82723AlnlHvnLUo:62
图3:烟雾探测系统对各种住宅类型的可靠性分析时变异数的影响这些住宅类型分别在平均可靠性估计和95%的置信区间估计内进行单独分析。结果列于表5,每种类型的结果明显不同。各种住宅类型的置信区间与自喷系统的可靠性估计时的置信区间不重叠。这就可能使有更多的数据用于烟雾探测器的分析,列于表5中各个住宅类型的烟雾探测器的可靠性估计完全不同,判断非相关数据差异的原因超出了这个分析的范围。ResidentialInstitutionalCommercialns30*75.1<77.8<80.fr813<83.5<84.670.2<72.0<73.7
分析中所用到的数据是在研宄中描述为典型系统的,在公开文献中关于喷淋系统和烟雾探测器可靠性的最好数据。典型数据是一种重要的依据,它用来判断某种类型的信息是否达到这种类型的统计学分析。分析的结果应该被用来做出推论,但必须在研究相关资料和测验它们对分析系统中使用的特殊安全计划的适应性以后。然而,总体的接近代表着在解决不同消防系统类别的可靠性时更高的标准,包括注意报告数据屮的不确定性和偏差。CommercialResidentialProponksnporBarInstitution70803RELIABILITY(%)CW0120.410860708090RELIABILITYW108图4:烟雾探测器对各种住宅类型的可靠性分析分配概要和结论一份详细的文献摘要和运行可靠性分析被用来关注几个消防计划的运行可靠性,消防计划包拈:火灾探测、自动喷水和防火分区。在这项研究中,运行可靠性被定义为消防系统在需耍时运行的可靠性估计。这些出版物的标准不直接在评估中叙述不确定性或偏差。关于防火分区,在表1屮的运行可靠性概述是它的唯一信息。在试图解决评估中的不确定性过程中,几个火灾的实际细节研究,烟雾探测器和自
动喷水系统的运行性能被重新分析,并II报告数据被提取为一个更加条理的评价。没冇发现防火分区的相关数据,这个评价包括利用常规统计方法来评价可靠性数据和运行可靠性的平均估计,还有运行可靠性达到95%的置信区间范围。表4和表5概括了这个分析的结果。测试结果显示,使用单一标准来评估一个消防系统的运行可靠性是不恰当的。例如,在表4中对喷淋系统的运行可靠性评估,在商业建筑中范围从88%至98%,同时平均估计为93%,人口数量太少(单值)为计算平均价值和住宅或公共建筑的置信区间的限制,但综合楼计算的平均可靠性估计为95%,同时95%的罝信区间为92%至97%,平均价值应用到可靠性上,基于这个认识即价值代表95%置信区间的平均范围,它是比较合理的与用来任意衍生的价值相比。另外,整个置信区间的使用和不是最可能的平均值相比,当比较系统吋冇更加明显详实的信息,因为所冇的相似系统的可靠性评估包括比较。这是当拿一个系统同其他许多系统相比较时一个公认的统计方法。烟雾探测器在表5中的运行可靠性值有一个与95%置信区间和关联的更为紧凑的范围。这可能是数据库的大小和质量以及通过HALL[1995]来保持结果的最初说明的一致性所导致的直接后果。基于表5屮所体现的结果,烟雾探测器的平均值为,对商业建筑为72.5%(下界70.2%,上界为73.7%),对住宅为77.8%(下界75.1%,上界为80.6%),对公共建筑为83.5%(下界82.3%,上界为84.6%)。烟雾探测器可靠性的变异结果进一步表明可靠性估计由为数据分析的住宅类型决定(见阁3),烟雾探测器的最高可靠性与公共建筑有关。这可能是许多的公共建筑需耍更多的维护和口常系统需求保证的直接后果。这一分析方法能很容易的应用到其他消防系统的运行可靠性评估。但是,应当指出文献中数据的可靠性是一个重要的因素。值得注意的是数据在内容和形式上的巨大变化,在学习报告和研究时这是努力的一部分。这项研究提供了一个十分广泛的初步尝试去描述消防系统的运行可靠性。调杳报告需耍大量的数据来改变数据库。这种努力的重心在取得更加具体的数据,系统的广泛人口能为消防系统运行可靠性的巨大改善提供基础,从而引起设计工程师的兴趣,另外这项技术对工程师基于高速发展的设计理念的性能分析也是必要的。外文出处:http://www.docin.com/p-262843630.html
附录二:外文资料原文
EstimatesoftheOperationalReliabilityofFireProtectionSystemsForthepastthreeyears,theNationalInstituteofStandardsandTechnology(NIST)hasbeenworkingtodevelopanewencryptionstandardtokeepgovernmentinformationsecure.Theorganizationisinthefinalstagesofanopenprocessofselectingoneormorealgorithms,ordata-scramblingformulas,forthenewAdvancedEncryptionStandard(AES)andplanstomakeadecisionbylatesummerorearlyfall.Thestandardisslatedtogointoeffectnextyear.RichardW.Bukowski,P.E.SeniorEngineerMSTBuildingandFireResearchLaboratoryGaithersburg,MD20899-8642USAEdwardK.Budnick,P.E.,andChristopherEScheme1VicePresidentChemicalEngineerHughesAssociates,IncHughesAssociates,Inc.Baltimore,MD21227-1652USABaltimore,MD21227-1652USAINTRODUCTIONBackgroundFireprotectionstrategiesaredesignedandinstalledtoperformspecificfunctions.Forexample,afiresprinklersystemisexpectedtocontrolorextinguishfires:Toaccomplishthis,thesystemsprinklersmustopen,andtherequiredamountofwatertoachievecontrolorextinguishmentmustbedeliveredtothefirelocation.Afiredetectionsystemisintendedtoprovidesufficientearlywarningofafiretopermitoccupantnotificationandescape,fireservicenotification,andinsomecasesactivationofotherfireprotectionfeatures(e.g.,specialextinguishingsystems,smokemanagementsystems).Bothsystemactivation(detection)andnotification(alarm)mustoccurtoachieveearlywarning.Constructioncompartmentationisgenerallydesignedtolimittheextentoffirespreadaswellastomaintain
thebuilding’sstructuralintegrityaswellastenabilityalongescaperoutesforsomespecifiedperiodoftime.Inordertoaccomplishthis,theconstructionfeaturesmustbefire“rated”(basedonstandardtests)andtheintegrityofthefeaturesmaintained.Thereliabilityofindividualfireprotectionstrategiessuchasdetection,automaticsuppression,andconstructioncompartmentationisimportantinputtodetailedengineeringanalysesassociatedwithperformancebaseddesign.Inthecontextofsafetysystems,thereareseveralelementsofreliability,includingbothoperationalandperfornzzsancereliability.Operationalreliabilityprovidesameasureoftheprobabilitythatafireprotectionsystemwilloperateasintendedwhenneeded.Performancereliabilityisameasureoftheadequacyofthefeaturetosuccessfullyperformitsintendedhnctionunderspecificfireexposureconditions.Theformerisameasureofcomponentorsystemoperabilitywhilethelatterisameasureoftheadequacyofthesystemdesign.Thescopeofthisstudywaslimitedtoevaluationofoperationalreliabilitydueprimarilytotheformofthereporteddataintheliterature.Inadditiontothisdistinctionbetweenoperationalandperformancereliability,thescopefocusedonunconditionalestimatesofreliabilityandfailureestimatesintermsoffail-dangerousoutcomes.Adiscussionofthesetermsisprovidedlaterinthepaper.ScopeThispaperprovidesareviewofreportedoperationalreliabilityandperformanceestimatesfor(1)firedetection,(2)automaticsuppression,andtoalimitedextent(3)constructioncompartmentation.Ingeneral,thereportedestimatesforfiredetectionarelargelyforsmokedetectiodfirealarmsystems;automaticsprinklerscomprisemostofthedataforautomaticsuppression,andcompartmentationincludescompartmentfireresistanceandenclosureintegrity.Itshouldbenotedthatinsomecasestheliteraturedidnotdelineatebeyondthegeneralcategoriesof“firedetection”or“automaticsuppression,’’requiringassumptionsregardingthespecifictypeoffireprotectionsystem.Severalstudiesreportedestimatesofreliabilityforbothfiredetectionandautomaticsprinklersystemstrategies.However,verylittleinformationwasfounddetailingreliability
estimatesforpassivefireprotectionstrategiessuchascompartmentation.Alimitedstatisticalbasedanalysiswasperformedtoprovidegeneralizedinformationontherangesofsuchestimatesandrelateduncertainties.Thislattereffortwaslimitedtoevaluationofreporteddataondetectionandsuppression.Insufficientdatawereidentifiedoncompartmentationreliabilitytobeincluded.Thispaperaddresseselementsofreliabilityastheyrelatetofiresafetysystems.Theliteraturesearchthatwasperformedforthisanalysisisreviewedandimportantfindingsanddatasummarized.Thedatafoundintheliteraturethatwereapplicabletosprinklerandsmokedetectionsystemsreliabilitywereanalyzed,withdescriptiveestimatesofthemeanvaluesand95percentconfidenceintervalsfortheoperationalreliabilityoftheseinsitusystemsreported.ELEMENTSOFRELIABILITYANALYSISThereisconsiderablevariationinreliabilitydataandassociatedanalysesreportedintheliterature.Basically,reliabilityisanestimateoftheprobabilitythatasystemorcomponentwilloperateasdesignedoversometimeperiod.Duringtheusefulorexpectedlifeofacomponent,thistimeperiodis“reset”eachtimeacomponentistestedandfoundtobeinworkingorder.Therefore,themoreoftensystemsandcomponentsaretestedandmaintained,themorereliabletheyare.Thisformofreliabilityisreferredtoasunconditional.Unconditionalreliabilityisanestimateoftheprobabilitythatasystemwilloperate“ondemand.”Aconditionalreliabilityisanestimatethattwoeventsofconcern,i.e.,afireandsuccessfuloperationofafiresafetysystemoccuratthesametime.Reliabilityestimatesthatdonotconsiderafireeventprobabilityareunconditionalestimates.Twootherimportantconceptsappliedtooperationalreliabilityarefuiled-safeandfailed-dangerous.whenafiresafetysystemfailssafe,itoperateswhennofireeventhasoccurred.Acommonexampleisthefalsealarmingofasmokedetector.Afiresafetysystemfailsdangerouswhenitdoesnotfunctionduringafireevent.Inthisstudy,thefailed-dangerouseventdefinestheOperationalprobabilityoffailure(1-reliabilityestimate).Asprinklersystemnotoperatingduringafireeventoranoperatingsystemthatdoesnotcontrolorextinguishafireareexamplesofthistypeoffailure.Theoverallreliabilityofasystemdependsonthereliabilityofindividualcomponentsand
theircorrespondingfailurerates,theinterdependenciesoftheindividualcomponentsthatcomposethesystem,andthemaintenanceandtestingofcomponentsandsystemsonceinstalledtoveri@operability.Allofthesefactorsareofconcerninestimatingoperationazreliability.Firesafetysystemperformanceisalsoofconcernwhendealingwiththeoverallconceptofreliability.Systemperformanceisdefinedastheabilityofaparticularsystemtoaccomplishthetaskforwhichitwasdesignedandinstalled.Forexample,theperformanceofafireratedseparationisbasedontheconstructioncomponent’sabilitytoremainintactandprovidefireseparationduringafire.Thedegreetowhichthesecomponentspreventfirespreadacrosstheirintendedboundariesdefinessystemperformance.Performancereliabilityestimatesrequiredataonhowwellsystemsaccomplishtheirdesigntaskunderactualfireeventsorfullscaletests.Informationonperformancereliabilitycouldnotbediscerneddirectlyfrommanyofthedatasourcesreviewedaspartofthiseffortduetotheformofthepresenteddata,andtherefore,itisnotaddressedasaseparateeffect.Thecauseoffailureforanytypeofsystemistypicallyclassifiedintoseveralgeneralcategories:installationerrors,designmistakes,manufacturing/equipmentdefects,lackofmaintenance,exceedingdesignlimits,andenvironmentalfactors.Thereareseveralapproachesthatcanbeutilizedtominimizetheprobabilityoffailure.Suchmethodsinclude(1)designredundancy,(2)activemonitoringforfaults,(3)providingthesimplestsystem(i.e.,theleastnumberofcomponents)toaddressthehazard,and(4)awelldesignedinspection,testing,andmaintenanceprogram.Thesereliabilityengineeringconceptsareimportantwhenevaluatingreliabilityestimatesreportedintheliterature.Dependingonthedatausedinagivenanalysis,thereliabilityestimatemayrelatetooneormoreoftheconceptspresentedabove.Theliteraturereviewconductedunderthescopeofthiseffortaddressestheseconceptswhereappropriate.MostoftheinformationthatwasobtainedfromtheliteratureinsupportofthispaperwerereportedintermsofunconditionaloperationaZreliability,i.e.,intermsoftheprobabilitythatafireprotectionstrategywillnotfaiZdangerous.
LITERATUREREVIEWAliteraturesearchwasconductedtogatherreliabilitydataofalltypesforfiresafetysystemsrelevanttotheprotectionstrategiesconsidered:automaticsuppression,automaticdetection,andcompartmentation.Theobjectiveoftheliteraturesearchwastoobtainsystem-specificreliabilityestimatesfortheperformanceofeachtypeoffiresafetysystemasafunctionofgenericoccupancytype(e.g.,residential,commercial,andinstitutional).Sourcesofinformationincludednationalfireincidentdatabasereports,USDepartmentofDefensesafetyrecords,industryandoccupancyspecificstudies,insuranceindustryhistoricalrecordsandinspectionreportsdocumentedintheopenliterature,andexperimentaldataReportsonexperimentalworkandfiretestingresultswereutilizedonlywhenfiredetection,automaticsuppression,orcompartmentationstrategieswereexplicitlyevaluated.Testsofsystemsusedforqualification,approval,orlistingwerealsoreviewedforinformationonfailuremodes.PublisheddatafromtheUnitedKingdom,Japan,Australia,andNewZealandwereincluded.GeneralStudiesSeveralbroadbasedstudieswereidentifiedthatreportedreliabilityestimatesforfiredetectionandfiresuppressionsystemsaswellasconstructioncompartmentation.Theseincluded(1)theWarringtonFireResearchstudy[1996]intheUnitedKingdom,(2)theAustralianFireEngineeringGuidelines[FireCodeReformCenter,19961,(3)acompilationoffirestatisticsforTokyo,Japan[TokyoFireDepartment,19971,and(4)resultsfromastudyofinsituperformanceoffireprotectionsystemsinJapan[Watanabe,19791.TheWarringtonFireResearchstudyaddressedthereliabilityoffiresafetysystemsandtheinteractionoftheircomponents.ADelphimethodologywasusedtodevelopdiscreteestimatesofthereliabilityofdetectionandalarmsystems,firesuppressionsystems,automaticsmokecontrolsystems,andpassivefireprotection(e.g.,compartmentation).TheAustralianFireEngineeringGuidelinesweredevelopedastheengineeringcodeofpracticesupportingthenewperformance-basedBuildingCodeofAustralia.Followingthemethodsinthisguide,buildingfiresafetyperformanceisevaluatedforsmouldering,flaming
non-flashover,andflamingflashoverfires.Theperformance(ie,probabilityofdetecting,extinguishingorcontrollingafireevent)offiresafetysystemsispredicted,accountingexplicitlyfortheoperationalreliabilityoftheparticularsystem.ReliabilityestimatesfromanexpertpanelratherthanfromactualdataareprovidedintheGuidelineforthispurpose.Finally,operationalreliabilitydatawerereportedintwoseparatestudiesinJapan.OnestudyinvolvedevaluationoffireincidentreportsfromthecityofTokyoduringtheperiodfrom1990to1997[TokyoFireDepartment19971.TheotherstudyinvolvedreviewoffireincidentreportsthroughoutJapanduringanearliertimeperiodendingin1978[Watanabe19791.Table1providesasummaryofthereliabilityestimatesprovidedinthesestudies.Significantdifferencesexistintheindividualreliabilityestimatesdependingontheparametersusedtodeveloptheseestimates.Dependingontherequiredaccuracyinpredictingfutureoperationalperformanceoffireprotectionsystems,dependenceontherangeofestimatesfromthesestudiescouldsignificantlyaltertheresults.Inaddition,theuncertaintyassociatedwithasingleestimateofreliabilityortheexistenceofpotentiallyimportantbiasesinthemethodsusedtoderivetheseestimatesmaylimittheirdirectusefulnessinaddressingeitheroperationalorperformancereliabilityoffireprotectionsystems.NA=NotAddressedTable1.PublishedEstimatesforFireProtectionSystemsOperationalReliability(ProbabilityofSuccess(YO))ProtectionSystemWarringtonDelphiUK(DelphiCroup)FireEngGuidelinesAustralia(ExpertSurvey)JapaneseStudies(IncidentData)SmolderingFlamingSmolderingFlaming/FlashOverTokyoFDWatanabeheatdetector0890907959489homesmokealarm76796575/74NANAsyscemsmokedetector86907080/fi59489beamsmokedetectors86887080/859489aspirated&mokcAct.86NA9095/95NANAsprinklersoperate955095/9997NAsprinklerscontrolbutdonotextinguish64NARaNAsprinklersextinguishNA96bfAmasonryconstruction8129%probabilityanopeningwillbefixedopen95ifnoopening90ifopeningwithautocloserNANAgyp$umpartitions6929%probabilityanopeningwillbefixedopen95ifnoopening90ifopeningwithautocloserNANAReviewofAvailableReliabilityDataDuetothelimitedapplicabilityofthereliabilityestimatespublishedinthegeneralliterature,theliteraturereviewwasextendedinaneffortto(1)developanimproved
understandingoftheelementsofeachofthethreestrategiesunderconsiderationthatinfluencereliability,and(2)identifyandevaluatequantitativedataregardingindividualsystemoperabilityandfailurerates.AutomaticSuppressionSystems(i.e.,sprinklersystems)Table2providesasummaryofreportedoperationalreliabilityestimatesfromseveralstudiesthatevaluatedactualfireincidentsinwhichautomaticsprinklerswerepresent.Asagroup,thesestudiesvarysignificantlyintermsofthereportingtimeperiods,thetypesofoccupancies,andthelevelofdetailregardingthetypesoffiresandthesprinklersystemdesign.TheestimatespresentedinTable2generallyindicaterelativelyhighoperationalreliabilityforautomaticsprinklersystems.Whilesomeofthereferencesincludefire“control”or“extinguishment”aspartofthereliabilityassessment,thereporteddatawerenotconsistent.Therefore,operationalreliabilitywasassumedtobelimitedtosprinkleroperation.Theestimatesalsoindicatearangeofvalues,suggestingthatitwouldbeinapprotasourcesandgeneraluncertaintyassociatedwithcombiningdatafromdifferentdatabases.Automaticfireextinguishingsystem(sprinklersystem)Table2summarizessomestudiesestimatethattheevaluationoftheoperationalreliabilityofautomaticsprinklersystem,firefightingactualfireincidents.Asagroup,thesestudiesreportverydifferentintimeperiod,typeofhousingandlevelofdetailrelatedtothetypeoffireandsprinklersystemdesign.Table2showsthespraysystemoperationalreliabilityestimatesaregenerallyrelativelyhigh,whilesomestudiesthefailureofthefirecontrolorfire,aspartoftheassessmentofreliability,butthereport’sdataisnotconsistent.Therefore,theoperationalreliabilityassumethattheoperationislimitedtospraying,assessmentshouldalsoshowthevaluerange,suggestingnottousethedeviationofaspraysystemreliabilitywithoutattentiontodataandcombiningdatafromdifferentdatabasesuncertaintysources.OccupancyReferenceRdiabilitv"alueCommercialMilne[1959JAutwnatKSpnokkrf^O]MiUerD974j86DOE[1982]989Maybee(1988J~993~I况61tiyJnrpPPO}813SprinklerFocus(1993J98.4-95.8Linder(1993]96GeneralI3r.i-.mgResearchEst1197JJ92.1Miller[1974F95.^Miller(1974]94.8Powers[1970)962RKhardson(1985)96Finucaneelal.[1987]96.9-97力
Manyatll^ST993-ResidentialMilne[1959]966InstitutionalMilne[1959J96.6Table2oftheoriginalbudgetbythereliabilityestimatesrangefrom81.13%to99.5%[Taylor]thelowvalueofmTaylor’sstudyandsomekookoverestimated(87.6o/o)reportthatthesesignificantdeviationthedatausedinthesestudies.Inbothstudies,thenumberoffiresisverysmall,andinthedatabasedoesnotdistinguishbetweenautomaticfireextinguishingsystemsandotherfire-fightingsystem.Thefinal99.5%ofthemaybeeandmarryatreportedhighestimatesreflectaspraysystemintheinspection,testingandmaintenanceofarigorousandwelldocumented.Thedataobtainedfromthespraysystem,anotherimportantlimitationisthatmostoftheAutomaticSprinklerSystemrecordssprinkleraccident.Inthesestudies,verylimitedaccidentdatawithreferencetotherapidresponseorasuitablewaterjettechnology.Assessthereliabilityoftheappropriatesprinklersystemshouldbeparticularlyconcernedaboutseveralfactors,including(1)allowscoveragewithin,(2)lowerwatersupplycapacity,(3)remotecontroloralarmsystemshavegreatpotentialinthefire.Basedonthis,thereareotherfactorsrelatedtothesetechnologies(suchasmaintenancelevel)candirectlyaffecttheoperationalreliabilityofthesetypesofAutomaticSprinklerSystem.Inaddition,youalsoneedtoresolvetheseproblems,thesystemdata,butbasedonlaterobservationsandgeneralhousingisgenerallylesslikelytomaintainnormal,somedesignedtoensuretheresidencewhatspraysystemreliabilitymaybereduced..FiredetectionoralarmsystemsTable3providesanoverviewofareportontheoperationalreliabilityforresidentialsystemsanalysistoassesstheaverageprobabilityand95%confidenceintervalsareestimatedbasedondataprovidedbyHALL[1955].Thescopeoftheaveragereliabilityestimatesrangingfrom68-88%.FiguresareconsistentwiththoseprovidedbythestandardswithTobyDelphy.Institutereliability.However,thegeneralscopeofthe95%confidenceinterval66-90%.Table3:Analysisofthereliabilityofsmokedetectors[HALL,19551OccupancyPropertyVscMeanReliability(%)n-1095%UpperConfidenceInterval95%LowerConfidenceIntervalResidentialApartments69.36068.7Hotek/MolcU77.679J76.4Dormitories86388484.3CommercialPublicAssembly67.969.865.9Stores&Offices71.773.569.9
Storage68J70.0Industry&Manufacturing80.279.1InstitutlonolCareofAged84.986.683.3CareofYoung84.086.381.6Educational76.979.674.1Hospitals&Clinics83.385.481.2Pnsons&Jails84285.9823CareofMentallyHandicapped87.590.384.8Thefiredistricttorelyonvarioustypesofequipmentsuchas:doors(includingfixedequipment),wall,floor/ceilingpenetrationholes,windows,fireshutter,smokematerialsandbuildings.Whenthefiredistrictisconsideredtobethefocusinthefireplan,intheliterature,thereislittledatathattheindividualcomponentsoftheoperatingroleinthefiredistrict.ThesingleismentionedforthebuildingassessmentandoperationalreliabilityWARRIGTONresearchandtheAustralianFireEngineeringIndex.Theseassessmentsarebasedentirelyonexpertjudgment.Thereforedoesnotprovidemorein-depthanalysis.AutomaticsprinklersystemsanalysisOnTable2,thesprinklersystemreliabilityanalysisistoanalyzeaccordingtothetypeofeachlive.ItshouldbenotedthatonlyonesourceMILNE,1959]onthereliabilityofpublicbuildingsandresidentialhousingestimates,andtheseearlydatadonotprovidethereliabilityofthedataofmodernresidentialsprinklers.ThedistributionhistogramofFigure1liststhereliabilityestimatesforeachhousingtype.Averageand95%confidenceintervallimitissuitableforgeneralresidential(inthestudydoesnotdistinguishbetweencommercialbuildings,thecategoryofresidentialbuildingsandpublicbuildings)andcommercialbuildings,andisapplicabletothebuilding(commercial,publicbuildings,residential)reliabilityassessment.TheseresultsareshowninTable4.ResidentialInstiturioDalCommercialGeneralCombinedn=1jl=1n=9n«7n=IS96.696.688.1<93.1<98193.9<96.0<98192.2<94.6<97.1CommercialGeneralAaiABUTYWInstutionalResidentialR6UABIUTY(*»RELIABIUTY(%>
Figure1:TheautomaticsprinklersystemreliabilityassessmentofavarietyofhousingtypesCommercialbuildingsandpublicbuildings,reliabilityassessment,theaverageofcontrolinotherresidentialtype95%confidenceinterval.Livableandpublicbuildingsinasinglepointestimate,anincreaseofsomeusefulthingsandoperationalreliability,andalsoincreasedthecapacityofthedatabase.18estimatedfourseparatecategories.However,thepointonresidentialandpublicbuildingsoftheestimatesshouldnotbeusedalonetodrawanyconclusions.Thereliabilityofcommercialbuildings,residentialandconstructionisestimatedto
providesomeusefulinformation.Basedontheanalysisofthesprinklersystemcanmakeuseofthedataisthereliabilityofoperationisestimatedatmorethan88%,ifyoudonotconsidercommercialbuildings,thereliabilityofthesprinklersystemcanreachmorethan92%.However,thejudgethisparticularsprinklersystemwiththosementionedintheassessmentsystemsimilarityisveryimportant.Thereliabilityofthecommercialbuildingsisintherange80-98%,94-98%inthegeneralconstruction.FiredetectionsystemanalysisThedataonestimatesofthereliabilityoffiredetectionsystemiscomprehensive.Thisdataspansadecade,andeveryyeartheReliabilityAssessmentReport(reflectedinTable3),itisdoneforthehousingofavarietyofdifferentuses.Wheretheanalysisisbasedontheuseofthehousedividedintoseveralbuildinglevel.Aftereachusehousingderiveddataandthencalculatethereliabilityofeachhousingestimates.Figure2showsthereliabilityestimatesofallsmokedetectorsonallresidentialtypes.30i607080分0RELIABILITY(%〉Propocionperroar2010snoo0.(100Figure2:ThesmokedetectorforavarietyofresidentialtypesofreliabilityestimatesAsshowninthehistogram,thedatahaveatwo-statedistribution.Therefore,inordertofurtherstudythetwoaverageacompletelydifferentdatabase,avariance(ANOVA)analysis.Variancetodetectthereliabilityoftheestimatedmeanandreliabilityofagiventypeofconstruction.ThegraphicalrepresentationofthevarianceintheformofleastsquaresmethodisreflectedinFigure3.Varianceaffectthefinalresult.ShowninFigure3,threeresidentialclassificationoftherespectiveestimatedaveragereliabilityofsmokedetectionfrom.HistogramcontainedinFigure4describestheestimatedreliabilityofeachhousingtype.LeastSquaresMeans
92■iS272WAmlsynUJc62111•/XXOCCUPANCYFigure3:ThesmokedetectionsystemvarianceanalysisofthereliabilityofavarietyofhousingtypesThesehousingtypesintheaveragereliabilityestimatesand95%confidenceintervalestimateforseparateanalysis.TheresultsarelistedinTable5,foreachtypeofresultsaresignificantlydifferent.Confidenceintervalfromthespraysystemreliabilityforavarietyofresidentialtypesofestimatedconfidenceintervalsdonotoverlap.Thismaymakemoredatafortheanalysisofthesmokedetectors,arelistedinTable5,thereliabilityofvariousresidentialtypesofsmokedetectorsisestimatedtodeterminethereasonforthedifferenceofnon-relevantdatabeyondthescopeofthisanalysis.ResidentialInstitutionalCommercialn■30a叫175.1<77.8<80.fr813<83.5<84.670.2<72.0<73.7SummaryandconclusionsTheaboveanalysiscaneasilybeappliedtootherfireoperationofthesystemreliabilityassessment.However,itshouldbenotedthatthereliabilityofdataintheliteratureisanimportantfactor.Itisnoteworthythatthedataincontentandformofthetremendouschangesinthestudyofreportsandstudies,whichispartoftheeffort.Thisstudyprovidesaverypreliminaryattempttodescribetheoperationalreliabilityoffireprotectionsystems.Thesurveyreportrequiresalotofdatatochangethedatabase.Thefocusofthisefforttoobtainmorespecificdata,population-widesystemcanprovideabasisforthegreatimprovementoftheoperationalreliabilityoffireprotectionsystem,whichledtotheinterestofdesignengineers,inadditiontheperformanceofthistechnologyforengineersbasedontherapiddevelopmentofdesignconceptstheanalysisisalsonecessary.'
您可能关注的文档
- 住宅楼电气毕业设计书_secret
- 某商住楼工程电气毕业设计论文_secret
- 建筑电气毕业设计(论文):住宅楼电气施工图的设计—某多层住宅楼电气施工图设计
- 建筑电气毕业设计(论文):逸夫教学楼电气设计
- 建筑电气毕业设计(论文):某小区电气工程设计
- 建筑电气毕业设计(论文):长春市宜家国际综合楼建筑电气设计
- 沈阳建筑电气毕业设计说明书
- 电气毕业设计下载
- 电气毕业设计330kv
- 沈阳建筑电气毕业设计论文说明书
- plc商店自动门控制电气毕业设计论文
- 220KV变电站工程电气毕业设计
- 探究某办公楼电气毕业设计论文
- 机械厂降压变电所的电气毕业设计
- kv变电所电气毕业设计.doc
- 民用建筑电气毕业设计(论
- 建筑电气毕业设计--酒店建筑弱电设计
- 建筑电气毕业设计说明书