Risk tolerance thresholds

Everyone has a different pain threshold and likewise, everyone has different risk tolerance thresholds. We use the plural because each one of us has various thresholds, for example, a perceived one and a financial one.

Your risk and risk tolerance thresholds made simple

Anyone of us, every-day decides to undertake some activities and consciously or unconsciously assumes risks we consider acceptable/tolerable or sometimes intolerable.Think about various activities:

• Hunting,
• fishing,
• driving a heavy vehicle, and finally
• cooking in one’s house.

They all have possibly unpleasant consequences such as, for example and respectively:

• encountering an aggressive bear,
• capsizing the boat,
• veering off road, and finally
• starting a fire in the house.

Each activity has a probability of turning into an accident and that accident may have a range of consequences. For instance from fender-bender to totaling the vehicle in addition to other vehicles in its trajectory, burning the stove, or the kitchen, up to the all house.

We evaluate risk when we combine those probabilities of an activity accident with their consequences. We reach our tolerance threshold when we say: I cannot accept that activity with that probability paired with those consequences.

Now, you may find a risk intolerable from a financial point of view, e.g. I cannot afford to total my truck, «ever» because I can’t pay for a replacement. Or you may find a risk intolerable from a perception point of view, e.g. if I total my truck my spouse will kill me!

Operations have risks and risk tolerance thresholds

Over the life of any operation it is inevitable that some incidents will occur. Some will be benign; some might be more significant and evolve into accidents. Higher significance consequences will occur as a result of accidents.

Obviously, accidents of various types will happen during the service life of the operation. Indeed, zero risk is not achievable, not even in highly controlled industries like civil aviation, and certainly not in industrial mishaps.

So as a manager of an operation you should ask yourself:

• are the risks tolerable and acceptable without mitigation?
• Are the risks tolerable and acceptable with mitigation?
• What are the residual risks if I mitigate risks? And finally
• are the proposed mitigation appropriate and sufficient?

To answer those questions we have to select the operation risk tolerance thresholds. Note that literature does not help to complete these endeavors as public thresholds address generic “large scale” societal tolerance concerns we discussed already in the past.

Risk Tolerance thresholds are indeed always project/culture specific. Therefore we need to define specific thresholds  by probing the perception and the financial resilience of various stakeholders through appropriate questions.

In the example below, after asking those appropriate questions to the managers of a mining operation, we developed their reasonable negotiated acceptability threshold model. They can be use it to define which risks are tolerable, intolerable, manageable and strategic in view of tactical and strategic planning of the operations.

Example of Risk Tolerance thresholds

As one can see in the figure below we prepare a quantitative probability-consequences quadrant BEFORE plotting the various «risks», i.e. probabilities, consequences pairs of various scenarios. It is paramount to prepare the tolerance beforehand in order to avoid biases, and it is even better to avoid showing any risk result until we plot them on the graph, for the same reason.

The following lines/curves are depicted on the graph:

• Yellow depicts the operations perceived risk tolerance threshold obtained through a facilitated process of Q/A, using client’s answers and data.
• Orange is the so-called «iso risk» threshold or a line with identical risk tolerance for any probability. It is also based on client’s data. It would correspond to a «soul-less» acceptance where there is no «shy-away» factor in front of even the highest consequences, provided their probabilities are very low.
• Purple is a benchmark obtained by reviewing historical performances in the industry. Indeed, anything above it is riskier than the world-wide portfolio of similar operations. Anything below is less risky than the portfolio. Finally,
• burgundy defines the ALARP concept expressed in terms of Willingness To Pay (WTP=2.5MUS$) (Glossary), based on literature. NB: for higher values of consequences ALARP generally displays a limit value (vertical line), as shown in N-P curves.

Both tolerances (corporate and Iso-risk) correspond to the top horizontal axis (consequences in US$). The ALARP tolerance corresponds to the bottom axis, number of fatalities, but can also be read on the top consequence due to the equivalence of WTP=2.5MUS$ adopted in this example). Note the vertical line placed at one fatality in the graph.

Now, those different lines/curves define specific areas numbered from 1 to 7, which denote specific nature of the risks they may include.

Here is a list of the seven areas (1 to 7) with a verbiage describing their “properties”.          

1. Below ALARP (fatalities expressed in WTP) and tolerable.
2. Tolerable for the considered mining operation and below world-benchmark.
3. Tolerable for the considered mining operation, but above world-benchmark.
4. Corporately intolerable: high-probability low-consequence scenarios which creates “fatigue”. For example, having a piece of equipment which breaks down often, but does not really have any significant consequence. Despite the corresponding risks are lower than the constant risk curve, clients will often decide to replace the equipment, either because the impact is overestimated and/or by “fatigue” process. This is often voiced as “that piece of equipment was a pain in the neck, so we changed it!”
5. “False comfort”, tendency to “play” because overestimating tolerance.
6. Iso-tolerable, corporately intolerable, ALARP intolerable.
7. Entirely intolerable.

Examples of risks in different areas of the risk tolerance thresholds of the example 

• A haul truck accident (not a crash) would likely plot left of the vertical “one fatality” line. Thus, it would land in Areas 1,2,3 depending on the level of training, road maintenance, truck maintenance, etc. for those influence the probability. Truck damages, if they are common, would land in Area 4.
• Accidents that have even very significant financial consequences but remain ALARP compatible would land in Area 1. For instance, this could be the case of multi-bench failures where monitoring and alert systems allow personnel to evacuate the danger area in time.
• Area 2 would receive all those accidents that are fine with respect to world-benchmark and are corporately tolerable. Their expected fatalities consequences make them H&S intolerable.
• Area 3 receives similar risks to those of Area 2, but with probabilities higher than the benchmark.
• A failure could plot in Area 2, 3, 5 or 6 depending on
  • the number of people and equipment in the area (including access control to these areas),
  • the business interruption it will generate; and finally
  • extant mitigation and monitoring, defenses.

Note that Area 6 also includes extremely high consequences, extremely low probabilities events.

• A full collapse of a pit face would certainly fall in Area 7 if the probability is above world-benchmark, and its cost of consequences extremely high (no double access road, for example).

Closing remarks

 Contact Riskope if you want to develop your own risk tolerance thresholds and use them to guide your decision-making and to establish mitigative road-maps.

Tagged with: ALARA, ALARP, Operational risks, risk acceptability, risk tolerance, tolerance thresholds

Category: Consequences, Hazard, Probabilities, Risk analysis, Risk management, Tolerance/Acceptability