1.1 本指南提供了统计评估被测报警阈值的具体要求，这些阈值被称为报警限，因为它们适用于从在役油分析中收集的数据。这些报警限值通常用于状态监测，以产生与机械磨损、油质和系统污染状态相关的严重程度指示。报警限值区分或分离不同级别的报警。四个级别是常见的，将在本指南中使用，但也可以使用三个级别或五个级别。 1.2 当被测数据集可以被表征为参数和受控时，建议使用本文描述的基本统计过程控制技术来评估报警限。 这种参数数据集的频率分布符合表现良好的具有“钟形”曲线外观的双尾正态分布。使用该技术计算统计控制限。在选定的置信水平上，这些控制限将控制内数据集的信噪比与具有显著、可分配原因的变化区分开来。操作员可以使用它们客观地创建、评估和调整报警限值。 1.3 还建议使用本文描述的统计累积分布技术来创建、评估和调整报警限值。 这种特殊技术采用排序数据集值的百分比累积分布。该技术基于实际数据集分布，因此不依赖于假定的统计特征。当数据集为参数或非参数时，可以使用该技术，如果频率分布出现倾斜或只有一条尾巴，则可以使用该技术。此外，当数据集除常见原因变化外还包括特殊原因变化时，可以使用该技术，尽管当特殊原因发生显著变化或消除时，应重复该技术。 该技术的输出是对应于排序数据集累积分布图中选定百分比水平的特定测量值。这些基于百分比的被测值用于创建、评估和调整报警限值。 1.4 本指南可适用于从大型车队或单个工业应用中采集的机器（例如，柴油机、泵、燃气轮机、工业涡轮机、液压装置）在役润滑油样本测试的样本数据。 1.5 本指南也适用于- 从其他设备应用中收集的服务油样本，其中磨损、油状况或系统污染的监测很重要。例如，它可以应用于充油变压器和断路器应用的数据集。 1.6 本指南不包括不基于统计的报警限评估技术。此外，本标准的技术可能与以下报警限选择技术不一致:“变化率”、绝对报警、多参数报警和经验推导的报警限。 1.7 本指南中的技术提供的输出可与其他报警限选择技术进行比较。本指南中的技术不排除或取代由原始设备制造商（OEM）或其他责任方制定和验证的限制。 1.8 本标准并非旨在解决与其使用相关的所有安全问题（如有）。本标准的用户有责任在使用前制定适当的安全和健康实践，并确定监管限制的适用性。 1.9 本国际标准是根据世界贸易组织技术性贸易壁垒（TBT）委员会发布的《关于制定国际标准、指南和建议的原则的决定》中确立的国际公认标准化原则制定的。 ====意义和用途====== 5.1 报警限值广泛用于使用在役润滑剂样品测试结果的数据进行状态监测。最初选择这些报警限值有很多依据。有许多问题需要解决。 其中包括: 这些限制是对的还是错的？ 是否有太多的假阳性或假阴性结果？ 它们实用吗？ 5.2 本指南介绍了统计技术，用于评估报警限值是否有意义，以及它们对于需要立即或未来采取行动的标记问题是否合理。 5.3 本指南旨在通过为机械维护和监测人员提供一种有意义和实用的方法来评估报警限值，以帮助解释监测机械和机油状况以及润滑油系统污染数据，从而提高基于状态的行动建议的一致性、有用性和可靠性。

1.1 This guide provides specific requirements to statistically evaluate measurand alarm thresholds, which are called alarm limits, as they are applied to data collected from in-service oil analysis. These alarm limits are typically used for condition monitoring to produce severity indications relating to states of machinery wear, oil quality, and system contamination. Alarm limits distinguish or separate various levels of alarm. Four levels are common and will be used in this guide, though three levels or five levels can also be used. 1.2 A basic statistical process control technique described herein is recommended to evaluate alarm limits when measurand data sets may be characterized as both parametric and in control. A frequency distribution for this kind of parametric data set fits a well-behaved two-tail normal distribution having a “bell” curve appearance. Statistical control limits are calculated using this technique. These control limits distinguish, at a chosen level of confidence, signal-to-noise ratio for an in-control data set from variation that has significant, assignable causes. The operator can use them to objectively create, evaluate, and adjust alarm limits. 1.3 A statistical cumulative distribution technique described herein is also recommended to create, evaluate, and adjust alarm limits. This particular technique employs a percent cumulative distribution of sorted data set values. The technique is based on an actual data set distribution and therefore is not dependent on a presumed statistical profile. The technique may be used when the data set is either parametric or nonparametric, and it may be used if a frequency distribution appears skewed or has only a single tail. Also, this technique may be used when the data set includes special cause variation in addition to common cause variation, although the technique should be repeated when a special cause changes significantly or is eliminated. Outputs of this technique are specific measurand values corresponding to selected percentage levels in a cumulative distribution plot of the sorted data set. These percent-based measurand values are used to create, evaluate and adjust alarm limits. 1.4 This guide may be applied to sample data from testing of in-service lubricating oil samples collected from machinery (for example, diesel, pumps, gas turbines, industrial turbines, hydraulics) whether from large fleets or individual industrial applications. 1.5 This guide may also be applied to sample data from testing in-service oil samples collected from other equipment applications where monitoring for wear, oil condition, or system contamination are important. For example, it may be applied to data sets from oil filled transformer and circuit breaker applications. 1.6 Alarm limit evaluating techniques, which are not statistically based are not covered by this guide. Also, the techniques of this standard may be inconsistent with the following alarm limit selection techniques: “rate-of-change,” absolute alarming, multi-parameter alarming, and empirically derived alarm limits. 1.7 The techniques in this guide deliver outputs that may be compared with other alarm limit selection techniques. The techniques in this guide do not preclude or supersede limits that have been established and validated by an Original Equipment Manufacturer (OEM) or another responsible party. 1.8 This standard does not purport to address all of the safety concerns, if any, associated with its use. It is the responsibility of the user of this standard to establish appropriate safety and health practices and determine the applicability of regulatory limitations prior to use. 1.9 This international standard was developed in accordance with internationally recognized principles on standardization established in the Decision on Principles for the Development of International Standards, Guides and Recommendations issued by the World Trade Organization Technical Barriers to Trade (TBT) Committee. ====== Significance And Use ====== 5.1 Alarm limits are used extensively for condition monitoring using data from in-service lubricant sample test results. There are many bases for initially choosing values for these alarm limits. There are many questions that should be addressed. These include: Are those limits right or wrong? Are there too many false positive or false negative results? Are they practical? 5.2 This guide teaches statistical techniques for evaluating whether alarm limits are meaningful and if they are reasonable for flagging problems requiring immediate or future action. 5.3 This guide is intended to increase the consistency, usefulness, and dependability of condition based action recommendations by providing machinery maintenance and monitoring personnel with a meaningful and practical way to evaluate alarm limits to aid the interpretation of monitoring machinery and oil condition as well as lubricant system contamination data.