Near-miss metrics

The process safety pyramid described in Figure 1.2 provides a useful concept for categorizing metrics by severity. Lagging, leading, and near-miss metrics are associated with the different levels of the safety pyramid. Figure 3.1 illustrates how each of these four areas fit into that categorization. ... 

Near-miss metrics include actual process safety incidents that do not meet the threshold defining a lagging metric as well as system failures that could have, but did not, lead to an incident such as an instrumentation failure or vessel corrosion. Near misses are often less obvious than accidents and are defined as having little if any immediate impact on individuals, processes, or the environment. Despite their limited impact, near misses provide insight into accidents that could happen. 

Commonly reported near misses include such events as exceeding operating limits, a release of a chemical or other hazardous substance that does not meet the threshold for a process safety incident metric, activation of relief valves, interlocks, or ruptured disks. Companies may establish near-miss metrics based on the specifics of their operation, based on their observations of frequent upsets or failures, or to track and correct observed unsafe practices or behaviors. 

Since a near miss is an actual event or discovery of a potentially unsafe situation, this metric could be defined as a lagging metric. A large number or increasing trend in such events could be viewed as an indicator of a higher potential for a more significant event therefore, many companies use Near Miss metrics as a surrogate for a Leading metric. 

A surrogate is a substitute, replacement, stand-in, or deputy. The many companies that use near-miss metrics as a surrogate for a leading metric are obviously aware that tracking near-miss trends can produce, as the paper says, a forward looking set of metrics which indicate the performance of the key work processes, operating discipline, or layers of protection that prevent incidents. ... 

Chapter 3—Process Safety Management Metrics commonly used in process safety management systems including leading, lagging, and near miss and activity and outcome, external and internal metrics as well as characteristics of successful metrics... 

A metrics program that identified and implemented appropriate metrics would allow the facility to track whether PHAs and MOCs are completed prior to work or a change and whether near misses had been documented and investigated. 

Serious process safety incidents involving loss of life, serious injury, or property damage are often preceded by multiple less-serious incidents or near misses as well as recurring unsafe behaviors or insufficient operating discipline. Process safety metrics are established, collected, and acted upon to help companies and... 

Lagging Metrics—process safety incidents that meet the threshold of severity and shouid be reported as part of the process safety metric Near-Miss Incidents—incidents that did not meet the definition of process safety incident metric... 

Regardless of whether a particular metric is classified as lagging, leading, or a near miss, the purpose of the metric is to establish good indicators of conditions that could ultimately lead to a severe event... 

But they should not worry The important point is not to develop a crisp definition to delineate the differences among the terms, but to capture and document that a failure of some sort occurred (Hopkins, 2009). Regardless of whether a particular metric is classified as a lagging indicator, a near miss, or a leading indicator, the purpose of the metric is to establish good indicators of conditions that could ultimately lead to a severe event. 

Because they are a retrospective measure, learning from such metrics can come only after an accident or failure. A simple accounting of catastrophic events alone does not provide sufficient information since the incidents are infrequent and with disparate causes. While it is important to track such incidents, it is also important to track other, less-severe incidents and near misses. 

In addition to the continuum of lagging-near-miss-leading, metrics may be described by additional dimensions based upon what is actually measured (an activity or quality of performance) and who will use the information collected by the metric (internal or external audience or both). 

In addition to differentiating among the severity of a metric (incident, near miss, unsafe behavior), a metric can track activities (whether an action occurs) as well as the outcome (the quality or performance) of the action. 

Both activity and outcome metrics are needed to understand fully the performance of the system—that a particular activity is being conducted and the outcome or quality of the activity. For example, the system may measure the activity of carrying out MOC—how many are conducted and their frequency. The audit of the MOC system coupled with a change in the number of incidents or near misses associated with the MOC system describes the outcome or quality of the MOC program. 

Metrics should be periodically audited to ensure they meet all the above expectations. For example, the near-miss data should be reviewed to ensure the appropriate investigations were done. Operating logs should be reviewed to ensure all near misses were reported, investigated, and recorded. 

The ultimate goal of the process safety system is to prevent process safety incidents. The Center for Chemical Process Safety s Process Safety Leading and Lagging Metrics Report (CCPS, 2007b) defines a consensus from several chemical and allied processing industries for definitions of process safety incidents and process safety near misses. If an organization adopts these definitions, a... 

The initial results reported from a metrics system will often indicate performance below expectations. Initial data may be skewed until metrics data collection is reliably implemented, and performance may continue to drop as improvement opportunities are defined and implemented. This sometimes occurs when the historical perspective of performance is too optimistic because objective measures had not been used in the past. Implementing near-miss data collection is one example The number of reported near misses will probably increase as reliability of reporting improves, so the number of reported near misses is likely to climb during the initial implementation phases of near-miss data collection. Audiences of metrics reports should be prepared for such situations, so the result is expected and potential negative reactions are managed. 

Metrics on process safety events and near misses that have occurred... 

In an organization that has not historically evaluated performance using objective data, the initial reported results may appear to indicate that performance has declined. This sometimes occurs when the historical perspective is too optimistic. For example, the number of near misses often increases as reliability of reporting improves. The recipients for metrics reports should be advised what to expect and provided explanation as to why the results may not meet expectations. The potential cultural impact, if left unaddressed, may raise doubt about the accuracy and value of the metrics reports, and these potential negative reactions need to be managed. 

INCIDENT REPORTING METRICS Process Safety Incidents Process Safety Near Misses Incident/ Near-Miss Subcategories and Cause Types... 

With these activity-based performance measures, there is also a variety of safety metrics that can be used to assess program performance. As identified by OSHA in studies ofVPP organizations, OSHA injury incidence rates, lost work day rates, and workers compensation losses are a few safety metrics that have been correlated to the performance of the VPP criteria activities. These measures can easily be expanded to include unsafe behaviors, accident trends, and near misses. 

Three types of process safety performance metrics are described in the document and the text on their selection and application is extensive. Reference is made to lagging metrics leading metrics, and near miss and other internal lagging metrics. Note that near misses are a metric separate from lagging indicators. This is the only publication found that makes that disfinction. As will be seen, there is a purpose for doing so in the CCPS document. The metrics pertain only to chemical process incidents and near misses, to the exclusion of types of incidents that are not process related. 

There is a separate section in the CCPS document on Near Miss Reporting and Other Lagging Metrics. This is said about near misses. 

---