The high voltage switchgear at a smelter was enclosed in a large metal clad building to protect the equipment from airborne materials and the weather. Many years later, it was discovered that the insulation built into the metal wall panels was flammable.
Fire modelling was carried out by specialist consultants that clarified how a fire might start and develop and what equipment would be damaged if it did. The largest cost of a fire would have been lost production and damage to the smelter if it lost power and cooled down in an uncontrolled manner.
The plant’s insurers requested that the walls be replaced to reduce the risk of fire damaging the plant. It was unclear whether the cost of replacing the building would be justified.
Using the results of the fire study with estimates of the time required to rebuild and of the cost of doing so, as well as assessments of the likelihood of a fire from similar facilities elsewhere, a model of catastrophic loss of power was developed and the NPV of foreseeable losses was assessed over twenty five years.
Several mitigation plans had been proposed over the years with a wide range of forecast implementation costs. It became clear that none of these was backed up with sound engineering studies. However, on the basis of the estimates that were available, the only option that would eliminate the risk was far more expensive than the value of the risk to the business and the others were marginally cost-effective.
The business took the results into an internal exercise to refine maintenance and other protective measures that would reduce the likelihood of a fire and used the information about the excessive cost of complete protection to negotiate an agreement with its insurers. It also began internal studies to refine estimates of the cost of the mitigation options so that they could be re-evaluated with improved information.