So, how do you determine the optimum repair time? Simple, it is the minima of the sum of the costs of repair and the costs of not repairing as a function of time in service. We know all about the costs of repair so that need not be elaborated. How do you get the costs of not repairing? Simple, you note the area under the sawtooth cost of waste, delay and/or customer complaint as a function of time in service. In other words, defect frequency gets larger and larger as the process roller is run longer and longer and longer. When the roller is changed out, the defect level drops. It may not go to zero because the new roller is not perfect and there are other contributing factors, but that is beside the point.
If you replace a roller too early, you cost your company money. If you replace a roller too late, the far more likely situation, you also cost your company money. Go to the very minor trouble of putting numbers on things instead of guessing when to repair/replace. This minor study may pay for itself on the very first repair cycle. It almost certainly will on the long run. This subject will be reported on again next week at AIMCAL’s Web Coating and Handling Conference. See you there. If you want a complete copy of the notes, email me with ‘Roller Economics’ in the subject line.