Original Date: 08/26/1996
Revision Date: 01/18/2007
Best Practice : Predictive Maintenance Group
Since downtime in a mill is costly, Weirton Steel Corporation (WSC) established a predictive maintenance group in 1991. This concept was first introduced in the Hot Strip Mill on the #5 Pickler where WSC employees collected and analyzed data. In 1993, an outside contractor was brought in to expand the program to the rest of the mill. The contractor team, made up of a site manager, three analysis engineers, and nine technicians, collected data with the goal of centralizing operations for each area of the mill on a scheduled basis. Vibration and thermography data was collected. By 1994, vibration readings were being collected from 20,000 routine points each month. Thermographic readings covered all areas, once every six months. Reports were distributed to maintenance foremen, and a follow- up report, listing completed jobs and cost savings, was generated every week for the planners.
Two WSC analysts in the Hot Strip Mill’s #5 Pickler are Level-I certified. A new contractor was retained to cover the rest of the mill with a site manager, two engineers, five analysts, and four technicians. The analysts are located with the maintenance planners in each area of responsibility. Vibration data collection increased to 35,000 routine points per month. Maintenance personnel determine which machines are critical and require monitoring. A full-time engineer is dedicated to special vibration monitoring projects. Thermography now monitors one area at a time for periods of two to eight weeks. Special projects in such areas as batch anneal, clogged pipes, and misaligned couplings are conducted as required. Reports are distributed to the maintenance planners who meet and discuss the findings with the maintenance foremen; follow-up is extensive. Monthly update reports are provided to operation/maintenance management and planners who meet to discuss jobs in each area. All returned, predictive maintenance work orders are rechecked for improvement in vibration levels.
Predictive and condition-based maintenance at WSC applies several technologies. Vibration readings are taken to measure the impact force on a machine and to detect bearing defects, misalignment, imbalance, coupling problems, steam traps, and mechanical looseness. When vibration readings are taken, the technicians also perform a visual inspection to detect loose bolts or other problems. Thermography measures temperature and can detect problems in such areas as electrical switchgear, gearboxes, torpedo cars, substations, circuit breaker panels, steam lines, and process parameters. New technologies added included motor current analysis to detect problems with rotor bar current harmonics, high voltage, discharge testing, and resistance measurements; wear particle analysis to identify wear in applications where metal meets metal such as gearboxes, bearings, and hydraulic systems; and ultrasonic monitoring to detect leaks in valves, steam traps, piping, and other process systems.
Training is a key aspect of the predictive maintenance program. Classes are conducted once a month to teach maintenance personnel how to use and benefit from predictive maintenance techniques. The purpose is to enable WSC maintenance personnel to gain an appreciation of these maintenance tools and have a better understanding of the information that the reports provide, so they can make better maintenance decisions.
Predictive maintenance techniques have significantly improved operations by reducing downtime, product loss, and machine replacements due to unanticipated failures. In 1993, the savings-to-investment ratio broke even. By 1995, the savings ratio was 9:1.
Future plans call for integrating the predictive maintenance program into other WSC programs including oil analysis, preventive maintenance, and spares management. New and more useful metrics are being developed to help evaluate effectiveness and return on investment. Acceptance testing of new installations is being adopted to monitor baseline levels and determine if installation was acceptable. The knowledge and experience gained is then applied to the development of acceptance specifications for new equipment acquisitions. Not only has this new program been gaining approval within the maintenance ranks, but it is facilitated by a quarterly newsletter and training. Maintenance personnel now appreciate the value of predictive maintenance methods.
For more information see the
Point of Contact for this survey.