Hazard probability levels

The hazard probability levels listed in Table 2.2 (MlL-STD-882) represent a qualitative judgment on the relative likelihood of occurrence of a mishap caused by the uncorrected or uncontrolled hazard. Here again, assuming a high probability that a situation will occur, a judgment can be made as to the importance of addressing one specific concern over another. 

Previous issues of MIL-STD-882 used the term hazard (defined as a condition that is a prerequisite to a mishap , and went on to define hazard severity and hazard probability levels (see Tables B.13 and B.14). This was definitively incorrect, as these categories described accidents (or mishaps) - not hazards. 

Hazard Probability Levels are qualitatively ranked from A to E. An example description from the Federal Transportation Administration follows ... 

Failure frequencies of structures, equipment, and piping are related to their acceleration which is related to the ground-motion of the plant s foundation (e.g., the peak ground acceleration). For PSA, it is useful to present the seismic hazard at the site as a family of hazard curves with different nonexceedence-probability levels (Figure 5.1-3). By selecting various values of the peak ground acceleration, the acceleration and forces on the plant components may be obtained as described in the following. 

Ref. 39 suggests an initial qualitative hazards analysis early in systems design, with only general levels of hazard probabilities identified, in addition to severity categories. An example of such a qualitative ranking from Ref. 39 appears in Table V. 

Likelihood of Hazard. The NEC recognizes two distinct levels of hazard probability. Division 1 denotes an environment in which the probability exists that sufficient levels of the hazardous element may always exist, under normal operating condition, as to warrant extreme protections. Whereas, Division 2 denotes an environment where the probability for sufficient levels of the hazardous element to exist, under normal operating conditions, is less likely, and therefore, the extreme protection is not justifiable. Further areas adjacent to Division 1 areas can often constitute classification as Division 2 environments. 

The level corresponding to a 50% probability of initiation is the value reported by most in the field of impact sensitivity. However, as indicated in the introduction and in the previous paragraph, for greater assurance with respect to hazards, the 10% probability or, as close as possible, the zero probability levels are more appropriate. 

It addresses only functional hazards. The determination of the hazard severity level does not attempt to account for the system failures necessary for its occurrence it only seeks to determine the appropriate limits for probability of occurrence for a given hazard. 

Table 2.3 shows the hazard risk matrix, which incorporates the elements of the hazard severity table and the hazard probability table to provide an effective tool for approximating acceptable and unacceptable levels or degrees of risk. By establishing an alphanumeric weighting system for risk occurrence in each severity category and level of probability, one can further classify and assess risk by degree of acceptance. Obviously, from a systems standpoint, use of such a matrix facilitates the risk assessment process. 

Design for Minimum Risk The system safety order of precedence dictates that, from the first stages of product or system design, the system should be designed for the elimination of hazards, if possible. Unfortunately, in the real world, this is not always practical or feasible. If an identified hazard cannot be eliminated, then the risk associated with it should be reduced to an acceptable level of hazard probability through design selection. 

The information provided above on this proposed system reveals many serious or potentially serious hazard risk levels. When asking the basic questions associated with the identification of system risk, the analyst can begin to categorize the severity of a potential mishap, and evaluate the probability of a possible occurrence... 

Therefore, when using the severity and probability techniques simultaneously, hazards can be examined, qualified, addressed, and resolved based upon the hazardous severity of a potential outcome and the likelihood that such an outcome will occur. For example, while an aircraft collision in midair would unarguably be classified as a Category I mishap (catastrophic), the hazard probability would fall into the Level D (remote) classification based upon statistical history of midair collision occurrence. The system safety effort in this case would require specific, but relatively minimal... 

The information provided above on this proposed system reveals many serious or potentially serious hazard risk levels. When asking the basic questions associated with the identification of system risk, the analyst can begin to categorize the severity of a potential mishap and evaluate the probability of a possible occurrence (refer Tables 2.1 and 2.2). The following is an itemized listing of a few of the initial safety concerns which should be resolved prior to proceeding into the design phase of this project s life cycle. The identification of these potential hazard risks is the result of proper utilization of the basic system safety tools discussed thus far. 

The ERDEC Safety Office has since performed risk assessments on various detectors brought to government facilities for testing. Such procedures, when followed, permit the return of contaminated items to manufacturers. In the following, assessments of two detectors, the MiniRae and the M43A1 upgrade, are used as examples of how risks were assessed. Tables 3.16 and 3.17 define hazard severity and probability levels used as the criteria for assessing risk. 

Again, a PHL is developed (use Appendix C as a starting point). The PHL is divided into hazard categories. The functional tree is created. Then the actual facility hazard analysis is started. Each hazard is assigned a severity and probability level, and the other portions of the hazard analysis worksheet are completed. Then a system safety assessment is performed and the worksheet results are analyzed. 

The combination of the hazard severity and the hazard probability defines the hazard risk classes. These classes are listed in Table 4 with different levels of tolerability Class A forms the intolerable area of the risk matrix, Class B and C the tolerable area and class D means acceptable risk. 

The first factor contributing to seismic risk, such as from an earthquake event with a PGA of 0.7 g, is the occurrence rate of such earthquake which is often measured by annual exceedance probability (AEP). The other factor is the craise-quence as a result of an earthquake, which is ultimately measured in terms of loss of life and sometimes economic loss. At a conceptual level, seismic risk can be expressed as the product of seismic hazard probability and consequence, and it represents the probabilistic expectation of the consequence. 

In Fig. 8, a set of hazard curves is shown, each curve being associated with a probability level (50 %, 5 %, and 95 %). This characterization of the site hazard in terms of a population of hazard curves, rather than a single hazard curve, reflects the uncertainties in the hazard estimation procedures. 

This example demonstrates that the Level B hazards would probably (but not necessarily) be the appropriate hazard level to manage, because Level A might be too vague by not focusing on any specific system, and Level C is designated as contributing causes/failures to the level of hazard, whereas Level B directly leads ... 

Table 3.1 is a representative matrix that combines hazard severity and hazard probability, so that risk assessment criteria may be applied to determine the acceptance of the risk, or to identify the need for corrective action by means of redesign, procedural instructions, or other means, to eliminate or reduce the risk to an acceptable level. 

---