A chemical plant is due for a major overhaul and the manager has to make an assessment of a number of uncertainties associated with the project. These include the time the overhaul will take to complete (after 35 days the losses of production caused by the overhaul could have serious consequences for the company) and the risks that there will be leakage of dangerous chemicals into local watercourses during the cleaning process. The following extracts have been taken from the manager’s draft report which details plans for the overhaul:
(i) ‘I assessed the most likely duration of the overhaul to be 30 days. I then tried to take an optimistic view and assumed that, if all goes well, we could finish the work 5 days earlier than this (i.e. in 25 days). I then took a pessimistic perspective and estimated that the longest the project will take is 34 days. I am therefore certain that we should complete the overhaul within 35 days.’
(ii) ‘Essentially the overhaul will be split into eight independent phases. I think the chances of us completing each phase without a pollution problem are high, say 90%. Overall, I therefore estimate that we have almost a 90% chance of avoiding a pollution problem during the project.’
(iii) ‘There must be a high chance that there will be no serious corrosion in the main pump. The last five pumps we’ve inspected at other plants were all corroded and the chances of getting six corroded pumps in a row must be very low indeed.’
(iv) ‘I’m mindful of the theft of equipment we had last week at our Briston plant. If we don’t take appropriate security precautions I am virtually certain that we will lose important equipment in this way during the overhaul, with possible disastrous effects on our ability to complete the project within 35 days.’
(v) ‘I estimated the probability of the West boiler requiring repair to be about 10%.’ (On a later page:) ‘Given the likelihood of seepage into the pipe feeding the West boiler, there must be a high chance of this boiler being corroded. I therefore reckon that there is a 50:50 chance that we will have to repair this boiler as a result of the seepage and corrosion.’ Comment on these extracts from the report in the light of Tversky and Kahneman’s work on heuristics and biases.