Some of the most critical maintenance activities require considerable commitments of time and effort. Examples include naming all stock items and all equipment, setting up preventive maintenance (PM) programs and developing spare parts lists. Many of these activities can be greatly simplified while still achieving most of the benefits if you know where the greatest opportunities lie. The best place to start can be identified by performing a "Pareto analysis."
Typically, improvements based on Pareto analysis are expected to achieve 80 percent of the benefits when 20 percent of the work is done. Good information records will identify the 20 percent that will provide the greatest benefits. In maintenance, the "Pareto effect" is often much stronger than 80/20. Here are a few examples.
The most important spare parts lists are for those equipment items on which corrective maintenance or routine parts replacement is carried out most frequently. In one plant where a project to establish new spare parts lists had begun, an analysis of work-order records showed that 60 percent of all corrective maintenance was carried out on less than 3 percent of all equipment. Starting with this 3 percent obviously would provide a much faster return on the cost of developing this critical information than working through the entire plant. In fact, once the 3 percent of spare parts lists was completed by maintenance experts, it would then be reasonable to use a process, such as the "automatic" spare parts list function available in many maintenance computer systems, to gradually develop the lists for the other 97 percent. Please keep in mind that automatic spare parts list generation usually results in many errors unless it is carefully managed by area maintenance experts.
In a large pulp mill, a study of three years of equipment-level downtime records showed that, of the more than 12,000 items in the plant, 87 (less than 1 percent) caused 80 percent of all unscheduled maintenance downtime. The establishment of a PM program for these 87 items reduced the unscheduled downtime by more than 50 percent within 18 months.
This is an excellent example of the value of good maintenance/operating information. In this case, all downtime that resulted in less-than-target production on each shift was recorded, and the equipment that caused each loss was identified. This arguably is the best measure of maintenance because it not only provides a measure of maintenance's "product" (reliability), but it also provides this information in a form that lends itself to a Pareto analysis, which can be used directly to improve performance.
When equipment maintenance costs are important, a Pareto analysis of long- and short-term costs for all equipment items can reveal the highest-cost items, which can provide more useful information than a simple monthly report on the cost of all equipment.
When there is a lack of data, a substitute for a Pareto analysis is to interview experienced operators, tradespeople and their supervisors to find out, for example, which equipment is repaired most frequently.