How To Do A Risk Analysis

There is a difference between a risk analysis and a risk assessment. A risk assessment involves several steps and forms a platform of an overall risk management plan. A risk analysis is one of those steps and is a defining a characteristic of each risk level and is assigned a weight score as a part of the risk assessment. Generally speaking, a risk assessment includes identification of the issues that contribute to risk, analyzing their significance, identifying options to manage, or maintain oversight of the risk, determining which option is likely to be the best fit for size, complexity and scaled to an organization, and assigning recommendations to decision-makers. A risk assessment also includes one or multiple risk analyses for both pre-risk and post-risk mitigation. A risk analysis is one single justification task of likelihood and severity of a hazard and communicated as a risk level, and it may be a standalone document or a supporting document in a risk assessment. 

There are several guidance materials available on how to do a risk analysis which comes with different designs. A risk analysis may focus on likelihood of a hazard, or it may focus on the severity of a hazard as a determining factor. The combination of likelihood and severity is communicated as a risk level. A risk analysis tool is the risk matrix, which assign a reaction to the risk by colors. Red is normally an unacceptable risk level, while yellow may be acceptable with mitigation and green is acceptable without a reaction to the hazard. In a risk matrix likelihood and severity are assigned classification letters and numbers. A low number could be assigned a high severity or a low severity depending on how the risk matrix is designed. The same is true for the likelihood level where the letter “A” could also be a high severity or a low severity depending on risk matrix design.  

Level of exposure is a third component of a risk analysis and to simplify the risk analysis it is normally assigned an exposure level of 1. An assigned exposure level would be between 0 and 1, or 0% to 100% certainty. With the exposure level assigned as 1, the certainty is definite, and the hazard has appeared. As an example, birds are a hazard to aviation.

The exposure level to birds for an aircraft on an approach is quite different for an aircraft on an approach to the same runway in January or May. Due to a common cause variation, the migratory bird season increases bird activities during spring and fall months at airports. During the migratory season an airport may apply multiple mitigation processes by ATIS notification, ATC direct notification or means to scare the birds away from airport. Birds are attracted to food sources, and the black runway surface is an attraction for insects, which then again attracts birds. An exposure level for bird activities, without affecting flight operations, may be between 0.1 to 0.9, or up to 90%. An operator may decide to cancel flight with an exposure level at 90%. However, this is an extreme operational decision task since passengers and freight are dependent on the airline in support of their own integrity and on-time commitments. Most often a scheduled or on-demand flight would continue as planned and rely on other aircraft or the airport to scare birds away from the approach or departure ends. By eliminating the exposure criteria and applying an exposure level of 1, a hazard, or risk level for each segment of the flight may be applied. Another common cause variation are thunderstorms, and an expectation that they are to be mitigated at the time of exposure.   

 
When conducting a risk analysis, one of the most important factors is to reduce inputs of subjective, wishes, or biased, or opinion-based applications. A common cause variation of a risk analysis are the individual assumptions, which does not make it a faulty risk analysis, but an analysis with justification of assumptions or individual variations. One descriptor of a risk assessment is “Possible” with a definition that “It is Possible to occur at leas once a year”.  When a risk analysis is conducted of the hazard of bush flying or flying into one of the most congested airports in the world, the Likelihood of Occurrence would not reach the level of “Possible”, since human behavior is to take the path of least resistance. A likelihood of “Possible” would increase the workload dramatically and it could also restrict business, or flights, into areas of a high profit margin. If this is a new route or area of operations their justification is based on a wish or opinion and might not be a true picture of the hazard. However, if the risk analysis justification is based on prior years documented records, the risk analysis is based on data and paints a true picture. There are no one-fits-all answer in a risk assessment, and there are no correct or incorrect answers to a risk assessment, since it is the operator who accepts or rejects the risk. While this is true, it is also a customer who accepts or rejects the risk to use services provided by one or the other air carrier.

One of the principles of a Safety Management System is to operate within a Just Culture. A Just Culture is a culture where there are justifications for actions, both proactive and reactive actions. A risk analysis is just as much a part of justification as any other areas of a Just Culture operations. After a risk analysis, both likelihood and severity are processed through a justification process. The platform to build on for likelihood justification is times between intervals. The first level of times between intervals is when times between intervals are imaginary, theoretical, virtual, or fictional. This is a level with no data available and it is unreasonable to expect the likelihood to occur. An example would be the likelihood of a meteor to land in your back yard. The second level is when times between intervals are beyond factors applied for calculation of problem-solving in operation. At this level, the likelihood cannot be reasonable calculated. It is just as impossible as reaching the last number of PI. Third level is when times between intervals are separated by breaks, or spaced greater than normal operations could foresee.

Number four is when times between intervals are without definite aim, direction, rule, or method. Incidents happens, but they are random and unpredictable. Level five is when times between intervals are indefinable. This is when it is impossible to predict an incident, but most likely one or more will occur during an established timeframe. Level six is when times between intervals are inconsistent. This is when incidents occurs regularly, but they are not consistent with expectations. Level seven is when times between intervals are protracted and infrequent and may last longer than expected, but the frequency is relatively low. Level eight of the likelihood are the foothills of a systemic likelihood and when times between intervals are reliable and dependable. Levels nine and ten are the systemic levels when times between intervals are short, constant and dependable, or times between intervals are methodical, planned and dependable, without defining the operational system or processes involved.

After the likelihood level is justified, the justification process continues to the severity level. There are also ten severity levels, which are independent of the likelihood levels. The platform for classifications of the severity levels is a platform of expectations. Building on the platform is a severity level that is not compatible with another fact or claim of the hazard. The next blocks are a severity level with insignificant consequences, a severity level inferior in importance, size or degree, a severity level that would attract attention to operational process, cause operational inconvenience, or unscheduled events, a severity level large in extent or degree, a severity level involving an industry standard defined risk, or a risk significant in size, amount, or degree, a severity level having influence or effect of a noticeably or measurably large amount caused by something other than mere chance, or ignorance. Severity level eight is the foothills of the catastrophic levels, when a severity level having influence or effect of an irrevocable harm, damage, or loss. Severity levels nine and ten are the catastrophic levels with a severity level of a turning point with an abrupt change approaching a state of crisis and sufficient in size to sustain a chain reaction of undesirable events, occurrences, incidents, accidents or disaster, or a severity level where functions, movements, or operations cease to exist. 

With a justification risk analysis, the first action is defined based on the risk level. At level one the risk the initial task is to communicate. Level two is to communicate – monitor. Level three is to communicate – monitor – pause. Level four is to communicate – monitor – pause – suspend. And level five is to communicate – monitor – pause -suspend – cease. The beauty of a justification-based risk analysis is that after corrective action is implemented and during the follow up process, the tasks are to be completed in reversed order until the risk reaches the communicate task level.  

Catalina9