Monday Accident and Lessons Learned and a Root Cause Analysis Tip: Cracks in Wind Farm Turbines – Is There a Generic Cause?
An Article in the The Pratt Tribune says that cracks on wind farm turbine blades in Flat Ridge, Kansas, are due to a “manufacturing defect” and a “quality control defect in the manufacturing process.”
The problem was discovered during the root cause analysis of cracks at another wind farm. Since the same manufacturing process was used for the blades at the Flat Ridge site, the manufacturer has decided to replace the blades under warranty rather than waiting to see which ones start to crack.
First, if I am a manufacturer, a manufacturing defect is not a root cause. It is a causal factor that needs to be investigated. Since they have found out how the cracks occur (“During the manufacturing process numerous layers of laminate are used to make the fiberglass blades. During that process little folds develop that build up in a very defined location on the blades that will eventually lead to small cracks in the blades and cause pealing.”).
The “little folds developing” needs to be examined. It is probably the causal factor that needs to be examined to find it’s root causes.
Next, should these “little folds” have been caught in the design/manufacturing testing process?
Finally, how long have reports of these “little cracks” been coming in?
Because these cracks seem to occur across the manufacturing process (all the blades seem to have them), the root causes will be Generic Causes that apply to all blades manufactured. But are the problems with the design/manufacturing/testing process also present in other parts of the manufacturing process (not just the “little folds” on these blades)? That is a much bigger generic cause question.
What do you do to find the real fixable causes of manufacturing problems?
Do you look beyond the immediate causes to find root causes and then probe further to see if there are Generic Causes?
If you would like to learn a process that is used by industry leaders around the world, see: