Risk Tolerability / Acceptability:

A risk assessment please does not really help to make any decisions on risk reduction/accident prevention and other mitigative plans. It becomes rationally operational only when its results are compared with a threshold generally called “tolerable/acceptable risk”.

In our practice the term tolerability will be used for physical, monetary risks, whereas acceptability will be used for harm to human beings and biological diversity.

When discussing human acceptability a distinction will be made between location-based risks and societal risks. Location-based risks derive from the annualized likelihood of a person being killed at a given location as a direct result of an accident associated with hazardous activities undertaken there. Societal risks represent the likelihood of a group of people who are not directly engaged in an activity that involves a hazardous substance being killed in an accident arising out of that activity. Location-based risk is an expression of the risk that someone who lives or works in a place where a hazardous activity takes place is exposed to. Societal risk is quite different: it looks at the consequences of mishaps from a very broad point of view of an entire society, possibly physically and emotionally removed from the mishap itself; as such, it is of interest mainly to public administrators.

Looking back in time, Wilson and Comar, then in the field of chemical industry Renshaw defined simple societal risk acceptability criteria as described in Table 1 and depicted in figure 5.1

Table 1

Figure 1 Comar, Renshaw and Wilson acceptability intervals

The application of these criteria is easily understood with an example. If a limiting probability of 1.00E-05 (Renshaw acceptability limit) is selected in a country with approximately 60 millions inhabitants like France, Italy or the United Kingdom, the application of the criteria leads to accepting a hazard potentially leading to 600 fatalities per year, provided fatalities occur for one to ten individuals at a time. This number is roughly ten times lower than traffic related casualties in Italy in 1994, respectively 5232 for France in 2004, or five times lower than the 2004 observed 3221 fatalities in the U.K. Thus the limiting value for Renshaw is totally passed by reality in the three countries. The traffic risk should be considered as unacceptable. Instead, if the Comar and Wilson criteria are used, then the risk would be considered acceptable. Interestingly none of the criteria cited here make a distinction between voluntary risks and involuntary risks and if on one hand governments make huge efforts to reduce the fatality count in their respective jurisdictions, the public keeps using (and abusing) cars, with no (or very little) consideration to risk. This apparent acceptance of the unacceptable and other apparent paradoxes when looking at risk tolerability constitutes the core of this section.

Obviously, qualitative risk assessments introduce more fuzziness than quantitative ones when the comparison of risks with tolerability/acceptability criteria is required. This, despite the efforts of many users to categorize the broad couples of qualitative likelihood-qualitative consequence with a semaphore type of coloring, i.e. red, yellow & green to characterize the acuity of risk.

Instead, QRAs provide a rational platform for decision making based on the comparison of the assessed risks with Quantitative Risk Tolerability Curves (QRTC), provided due regard is given to uncertainties. This statement remains true for risk acceptability, i.e. when a QRA is used to define risk to human beings in terms of lives lost as it will be demonstrated later on in this chapter.

In recent years "standardized levels of risk reduction" have been formulated, and at least three of these definitions are now in common use among analysts. These three levels of risk mitigation also represent a convenient way to elude explicit tackling of risk tolerability, especially when the delicate theme of human life has to be dealt with.

ALARA: ALARA is an acronym for the phrase As Low as Reasonably Achievable. It is most often used in reference to chemical or radiation exposure levels.

ALARP: ALARP stands for As Low as Reasonably Practicable, and is a term often used in the milieu of safety-critical and high-integrity systems. The ALARP principle is that the residual risk shall be as low as reasonably practicable. For a risk to be ALARP it must be possible to demonstrate that the cost involved in reducing the risk further would be grossly disproportionate to the benefit gained. The ALARP principle arises from the fact that it would be possible to spend infinite time, effort and money attempting to reduce a risk to zero. It should not be understood as simply a quantitative measure of benefit against detriment. It is more a best common practice of judgment of the balance of risk and societal benefit.

BACT (Best Available Control Technology) For example: an emission limitation based on the maximum degree of emission reduction (considering energy, environmental, and economic impacts) achievable through application of production processes and available methods, systems, and techniques. BACT does not permit emissions in excess of those allowed under any applicable Clean Air Act provisions. Use of the BACT concept is allowable on a case by case basis for major new or modified emissions sources in attainment areas and applies to each regulated pollutant.

The determination of the acceptable risk level generally differs depending on whether a phenomenon is natural or man made, a private or a public (societal) issue.

Essentially Figure 2 demonstrates the decrease of risk against increasing mitigative costs. When risks are very high, a relatively small investment generally allows reducing risks quickly whereas investments increase asymptotically when risks are reduced beyond a certain level. The graph shows the point at which an acceptable threshold of risk mitigation might be settled – stating explicitly that the mitigative costs will realistically be too high to achieve a theoretical total abatement of risk. The blue vertical line is set in the ALARP/ALARP/BACT zone, i.e. a zone where risks are As Low as Reasonably Achievable (ALARA), As Low as Reasonably Practical (ALARP), or obey to the Best Available Control Technology (BACT) concept. Needless to say that the definition of these risk abatement levels is not common in many industries. In summary, the blue line is depicted "in the commonly accepted reasonable risk abatement zone", and its intersection with the risk abatement and mitigatory investment functions defines the residual risk and the investment necessary to attain it.

Figure 2 Risk and Mitigative Costs

Tolerability for Physical Damages

Once the risks incurred by an operation are estimated, rational and sustainable decisions on risk mitigation are generally requested by clients wishing to adopt risk management methods and maximize the investment they have made by performing a risk assessment. These can only be taken after an explicit risk tolerability function in defined. The tolerability function can be: derived formally (mathematically) from client’s financial data, defined empirically, derived from public opinion tests, “negotiated”. Generally the final tolerability curves selected by clients are the result of a mix of these various approaches. Tolerable risk curves are always project and owner specific and indicate the level of risk, which has been deemed acceptable by an owner for a specific project or operation (possibly taking into account public opinion). This means, as an example that within large companies corporate risk tolerability may differ quite substantially from a branch operation’s one. Risks, which plot to the left and below the tolerability curve, are deemed bearable. Risks, which plot to the right and above the curve, are deemed unbearable and some measures of mitigation are considered necessary to reduce their likelihood. Reducing the likelihood of an impact may be, for example, as simple as imposing “no stop” zones on a road. When working empirically, two curves should be developed, one representing the optimistic, the other the pessimistic view of tolerability. The band between the curves represents a range of uncertainty on tolerability defined by an organization. When data are available theoretical curves can be developed and then discussed with key personnel.		Figure 3 Tolerability envelope
		Figure 4 Intuitive curves vs factual curves
		Figure 5 Annual probabilities vs. Costs of Accident is M$

The development of empirical-estimated tolerability curves requires caution and continuous calibration as the extent of correlation between an individual's estimate or ranking of probabilities and the true value/ranking is usually quite weak, sometimes even in the order of zero. However it has been demonstrated that pooled judgments correspond better with the truth as the number of individuals increases. For instance, the average correlation between individual judgments and the correct rank order may increase tow folds when pooled across seven individuals, and twenty individuals may have an excellent ranking. No wonder, or God save us, juries are made out of twelve people. Jokes left aside, this is one of the reasons why risk assessments, and in particular risk tolerability curves should always be defined by a group, and not by an individual.

Risk tolerability is obviously a function of an organization’s wealth. In the case of some industries, in particular extractive industries, this translates into a function of time. In the mining industry, for example, as ore reserves are depleted, tolerability decreases because the company has less future wealth to buffer a hit (the operational safety margin decreases with time). In Figure 6 two curves are depicted: one with maximum theoretical tolerability evaluated at inception of the operations, the other with the theoretical tolerability evaluated at the end of operations (figure 5.7). These curves appear as a purple and a blue function at the right end of the main graph on Figure 6. In the same figure the envelope of empirically selected risk tolerability curves for various industries world-wide depicted in figure 5.3 is super-imposed. Figure 8 depicts another case from a large to very large mining operation. This case displays tolerability curves selected by a very large and long remaining mining life operation in South America. Prior to the risk assessment the client had already formulated written empirical guidelines for risk tolerability. These are graphically represented by the four pointed stars on Figure 8 The client selected a sophisticated empirical set of curves that denotes: a) Constant risk tolerability (read further for an explanation related to constant risk tolerability), b) Compliance with prior existing written criteria, c) Risk exposure similar to the one of the prior example.		Figure 6 Tolerability envelope
		Figure 7 Tolerability over time
		Figure 8 An envelope of tolerability from a mining company

Acceptability for Harm to Human Beings

All the examples described in the prior section are geared towards the tolerability for physical losses and do not encompass the possible casualties and accidents with human cost, bearing of course a heavier emotional component. Some clients adopt however a very actuarial approach and do not hesitate to allocate a cost to human life (like commonly done by insurance companies, governments when they perform large scale risk assessments etc.). In those cases a question generally raised is which cost per casualty and how “homogeneous” that cost should be when a client operates various facilities in various countries.

Examples of probability of events vs. expected casualty count per accident can be found in the literature.

Various societal acceptability curves are depicted in Figure 9 including the ones proposed by the Australian National Committee on Large Dams (ANCOLD) which can be seen as an industrial standard.

Where societal risk is concerned, the key issue is whether the totality of the risks associated with an activity that has the potential to lead to a disaster is acceptable, taking into account the social costs and benefits of the high risk activity in question (indeed in a risk study, great attention has to be exerted in ensuring that the acceptability curves are derived for the risks considered: curves derived for hazardous industrial activities cannot be used for natural hazards like typhoons, quakes or flooding). That’s why, as we will see below, large casualties events seen through a societal approach may feature surprisingly high acceptability levels, in apparent contradiction with location based risks (i.e. risks defined by a risk assessment for a plant, a facility, one road etc.)

Figure 9 Various Acceptable Curves

Tolerability to Media

A somewhat different approach can be adopted for tolerability to media risks, as depicted in the Figure 10.

Figure 10 tolerability to media risk

Here the tolerability is defined qualitatively, for example by crossing the reach of expected public scrutiny (i.e. who is going to be interested in the mishap), with its level, and finally deciding which ones of these cross classes are tolerable, difficult, or unacceptable, perhaps with a simple “traffic light: green, yellow, red” approach.

Often, people unfamiliar with tolerability curves, like this type of qualitative approach to tolerability better then the quantitative one, despite the fact it makes comparisons less obvious and introduces quite a fuzzy factor when compared to numeric representations. Others will argue that the numeric evaluations give a wrong sense of security because of their apparent precision, argument to which we would like to respond by reminding that any tolerability curve depicted or discussed should always present a generous width, depicting precisely the inevitable uncertainties.

Visit the book section of our website.

Visit the consulting section of our website.