Risk matrices

Based on the safety management principle of the critical few, the top 20 percent of near misses could hold the potential for 80 percent of the potential accidental losses. 

An invaluable safety tool is the risk matrix (Model 1.3). Remember, it s not what happened, it s what could have happened. The risk matrix is a crystal ball to predict the future or possible outcomes of near miss incidents. It can be used to forecast the probability and severity of the next loss. 

Occurrence Defined Frequency (per annum) Consequence Severity  

Note V = significantly high risk, S = medium risk, N = low risk, Z = negligible risk. 

This tool for risk assessment offers a wide scope of applications thanks to its simplicity. It is based on the estimation of probability and consequence of an identified hazard. The main aim of the tool is to provide for the risk extent estimation, or provide information necessary for the risk assessment (see Table 4.1). 

The drawbacks of this method include the fact that the parameters of the probability and consequences of a negative event are quantified by blunt nonnumeric values, that is, values defined verbally. It is up to an assessor to select criteria for dividing risk into individual groups, which renders comparability of results defined by different assessors difficult and eventually biased. 

---

Minmax Regret or Risk Rule.—In this case one forms a new matrix called the risk matrix in which the elements of each column k are defined by rik = max uik — uik (k = 1, -, n), (i.e., for each column... 

Resonance, nonlinear, 354 Resonance, subharmonic, 372,376 Rhombus rules, 85 Risk matrix, 315 Risk rule, 315... 

Using a tool such as a qualitative risk ranking matrix can be very useful in identifying low-risk buildings. For those events that have potentially major or catastrophic consequences to buildings and their occupants, however, a qualitative risk matrix may not always be an appropriate final evaluation. For events that are potentially major or catastrophic, regardless... 

Figure 5.1 Example qualitative risk matrix (adapted from Ref. 3).

Hazard assessment is a consequence analysis for a range of potential hazardous chemical releases, including the history of such releases at the facility. The releases must include the worst-case scenario and the more likely but significant accident release scenarios. A risk matrix can be used to characterize the worst-case and more likely scenarios. 

Figure 37. Risk Matrix for Swiss Risk Assessment Methods. (Ref. 42)...

The risk matrix will be used to find a possible link between Sis and accidents which have occurred recently and attempt to establish why in spite of the presence of so many Sis accidents still occur in the chemical process industry. Therefore, in the next sub-Section, the most commonly used Sis are discussed and the data (events) they are constructed from are displayed in the risk matrix. Subsequently, this leads to the first risk coverage area, that of events Sis use to indicate safety. 

These findings were obtained by discussing the different Sis with different experts from practice. Finally they are presented in the risk matrix at Figure 11 which shows the risk coverage area in terms of likelihood and consequences The data from this is used to construct pro-active Sis. Note that the results give an overview of the risk coverage area in qualitative terms in contrast to the risk coverage area of the events present in the accident trajectory described in the next Section. 

From hindsight analyses of accidents by Heinrich (Heinrich, 1959), Turner (Turner, 1978), Leplat (Leplat, 1987), Reason (Reason, 1997), etc., it is known that failures or deviations in normal operations are present prior to, and are directly related with, an accident. From hindsight analysis as reported in FACTS, the failures or deviations as well as the accident trajectory or causal path, of 70 accidents are known. To derive the risk coverage area these deviations, are placed in the risk matrix. The only deviations taken into account are those which occur in the operational process and are part of the accident trajectory or causal path prior to the critical events as described in FACTS. So the latent conditions lying behind these operational deviations as described by Reason (Reason, 1997) are not yet taken into account but will be discussed in the following Chapter. 

Some further explanation must be given here, because in practice some Sis encompass Tow consequence deviations. For the known risks type of predictive Sis it is obvious that they only cover high consequence deviations in the risk matrix. New risks type of predictive Si s are developed to encompass indirect safety related deviations (i.e. Tow consequence deviations). Though, in practice these deviations are often excluded from the predictive Sis tools as Bayutt (Bayutt, 2003) shows encompassing such indirect safety related deviations is perceived as a waste of resources and attention is distracted from more important scenarios. 

Finally, both the properties, likelihood and consequences, determine the risk. However, to prioritize between different risks, the risk has to be expressed as a onedimensional entity, as discussed in Chapter 1. To establish this one-dimensional risk, a standard risk matrix is used, as depicted on the left side in Figure 29, taken from the safety standard IEC 61508 (IEC 61508, 2000). From this standard risk matrix, the risk matrix used to prioritize the precursors is derived, as depicted on the right side in Figure 29. 

In Figure 29, the rows of both risk classification matrices show the established likelihood, while the columns show the perceived consequences. The result of both likelihood and perceived consequences is the perceived risk, which is indicated with a perceived risk class from 1 to 4 for the left risk matrix and from 1 to 3 for the right risk matrix. The interpretation of the different risk classes for the left matrix in Figure 29 is derived from the IEC 61508 (IEC 61508, 2000) and is stated as ... 

To establish the individual risks of all precursors the properties likelihood and perceived consequences are combined in the derived risk matrix, as shown in Figure 29. If there are several precursors of the same class of perceived risk the final choice of which precursor will be analysed further is made by the multi-disciplinary group of experts which established the perceived consequences. 

To draw a comparison between the previous results and the results from applying this protocol, the ten precursors already presented in Table 7 are taken for further analysis. From the derived precursors in the previous stage, which are based on analysing deviations during six months, a prioritized list of precursors is derived. The derived risk matrix for all ten identified precursors is depicted in Figure 32. Note that no lower aggregation level precursors were present amongst the ten identified precursors. 

Figure 32 Risk matrix to prioritize all ten identified precursors.

The second step, of sorting the selected precursors according to their perceived safety related consequences, is achieved by studying safety reports and confronting the precursors with multi-disciplinary experts, i.e. experts from production, maintenance and safety. The expert group provided the identified precursors with perceived safety related consequences, by formulating possible scenario s, from which the consequences could be obtained. From both the likelihood (see Table 17) and perceived consequences, the perceived risk class is obtained, as discussed in Chapter 5. Figure 42 shows the risk matrix for the precursors presented in Table 17. 

Risk indices are single numbers or a tabulation of numbers that are correlated to the magnitude of the risk to people. Some risk indices are relative values with no specific units. The limitations on the use of indices are that they may not be an absolute criteria for accepting or rejecting the risk. Risk indices also do not communicate the same information as individual or societal risk measures. An example of risk indices is a risk ranking matrix. Table 6-4 (modified from CCPS, 1992) shows how severity and likelihood are combined to obtain risk indices. An example risk matrix is shown in Figure 6-3 (RRS, 2002). 

Table 16 Financial risks matrix in pipeline projects. ...

Table 17 Political risks matrix in cross-border pipeline projects.

Figure 10.11 Risk matrix adapted from the IEC 61511 standard, indicating the accepted and non-accepted risks, as well as an intermediate field. The numbers represent the number of required IPLs together with the required SILs.

Ranking The qualitative estimation of risk from severity and likelihood levels, in order to provide a prioritizing of risk based it s magnitude (refer to corporate risk matrix for ranking based on severity and likelihood levels). 

Some companies are now using Risk-Based Inspections (RBI), especially for the costly preparation efforts for internal inspections. RBI is a method that is gaining credibility and popularity. Organizations calculate a proposed inspection frequency by considering the consequences of failure (as described previously) and the likelihood of failure based on the anticipated internal corrosion rates. The method employs a risk matrix. 

Once the specific issues and scope of the analysis are defined, a semi-quantitative risk assessment may be conducted using either risk indexing or a risk ranking matrix. The risk indexing and risk matrix techniques should be built on the information from the earlier analyses. Each level of risk analysis should not be considered a separate effoit, but a continued understanding of the transportation issue. Additionally the information gained from these activities can be used to update the qualitative analysis, especially benchmarking comparisons. 

---