Root Cause Analysis Forensic Investigation

Forensic Services Newsletter

If you Googled the phrase 'root cause analysis' 12 months ago you obtained 226,000 hits. Today it is 1,380,000. It has become very trendy to use this term, with its described use so diverse, that if I am asked whether we use a root cause analysis approach, I am not sure how to answer.

Failure analysts have to keep asking why? However taken to its extreme, every accident can be traced back to the big bang at the beginning of the universe. It is understood that this is ridiculous, so there has to be a limit to asking this question. That fact that you got out of bed is a step on the way to the traffic accident you were involved in on that day, but just about everyone would accept that it is not relevant.

Melting furnace, with a leak at an oil pipe. Instead of fixing the leak, the leaking oil is channeled to a 200 litre drum. If a fire starts, what is the root cause?

Where we encounter most problems with the phrase 'root cause', is the implication that the root cause is the cause, the one and only cause that is at the core or beginning of the event. Supposedly if someone determines the root cause, then of course it stands to reason that this cause takes precedence over all the others that might be suggested. However this neglects the fact that most accidents are a combination of a number of causes and the so-called root cause may well depend upon who is asking. For a single accident, the adjuster, failure analyst, manufacturer and the insured may all justifiably decide on a different root case.

I imagine many are now thinking proximate cause and the definition of this that carries sway in courts of law. It may well be that an adjuster, lawyer and insurance consultant can reach agreement on the proximate cause, but if the same consultant was working for the insured or the manufacturer, would the answer be the same?

When there is a failure of a major item of equipment, such as a turbine, the investigation report by the manufacturer may dwarf all others, not just in size but in the detail and range of sophisticated test equipment used. What surprises many is the fact that many of these investigations do not come up with an answer at all, or at least one that satisfies the insurers. This is sometimes looked upon with suspicion, but it must be considered that the manufacturer begins these investigations with a different viewpoint than others, with the prime question put by the management team and their lawyers to the in-house laboratory being, Is it our fault? Given this framework, the analyst launches an exhaustive range of tests designed to answer this question. There might not be the slightest interest in whether the answer will, for example, assist in determining which insurance policy applies. At the steelworks where I worked years ago, I was taught that chemical analysis was the first test to perform when a customer returned a failed piece of steel with a complaint. This was because this test often showed the steel was not made by us or even in Australia.

Failure analysts working for the insurance industry also encounter difficulty when an accident occurs as a result of a combination of factors, particularly if one is payable and the other not. We might be asked which factor was the most significant, but sometimes this is like sitting on a three-legged chair and being asked which leg is the most important.

A person operating a cutting torch, with no gloves, safety glasses, other suitable clothing, or skill. If such a person starts a fire in a factory, what is the root cause?

Turning to the insured and putting aside the insurance claim, there is a need to prevent a repeat accident. The traditional argument is that to prevent further accidents you first need to know the cause. In fact sometimes the reverse is true, the root cause will depend upon the need of the person asking. In a recent investigation of an explosion it was clear that the conflagration was ignited when a contractor carried out welding on the outside of a tank with flammable material inside. Technically this is an adequate description of the cause, a bit like welding on the outside of a petrol tank. But does this tell the factory how to prevent further accidents? This factory had experienced engineers and they knew that what happened was fundamentally wrong. It turned out that the cause of the accident, from their perspective, was that the hot work permit was filled out by workers who were not the best qualified to do so. To prevent further accidents, the qualified engineers needed to take back control of the permit system. Was then the root cause of this accident the faulty management system? That depends on who is asking.

It is apparent that the phrase root cause analysis is entrenched and as failure analysts working generally for insurers, we will continue to be provided with investigation reports by insured's and manufacturers that purport to have determined the root cause. Experience has shown that some reports will be excellent, others severely faulted, neither of which has anything to do with the use of the phrase root cause. There is a temptation to jump on the bandwagon, but it seems best to continue to carry out failure analysis as best one can.

If you have a query you may contact us in any of the following ways.

Telephone Fax
+60 3 2145 2022 +60 3 2145 1022

Our web site contains a great deal of information and you can use the site search facility at