Improving Asset Effectiveness Through Performance Monitoring

Dec. 19, 2006

For process enterprises, the value-added in manufacturing happens to be the primary determinant of business performance. As such, optimizing asset effectiveness is at the forefront of today’s performance-driven enterprise.

However, alterations in equipment configuration, degradation of equipment, and changes in business priorities and market demand all conspire to degrade performance. To improve business performance, operational decisions must be based on real-time performance monitoring that aligns with optimizing asset effectiveness.

Real-time Performance Management (RPM) is an integral part of an overall methodology for driving a business toward optimal asset and business performance while adapting to changing market conditions. RPM uses data collected from many different sources found in plant and enterprise systems, which is then organized in proper context for displaying to the appropriate individual. The important RPM measurements that employees see relate to performance targets for the operation. The closer these measurements are to true measures of business performance, the more likely workers will make the correct adjustments.

The old saying is true: “You cannot control what you cannot measure.” Companies spend considerable time and effort to control and optimize processes, but often do not consider other factors such as equipment, loop and controller performance. Process optimization efforts have generated significant benefits, but without including these other factors, assets will always under-perform.

Methods and applications

Selecting the appropriate performance targets, key performance indicators (KPIs), and method of calculating them from real-time process data is critical to success. KPIs are based on process knowledge gained from engineering, design and operations, along with business strategies. Methodologies for calculating performance range from simple transformation of raw data, to rigorous and empirical models, neural nets, and univariate and multivariate statistical methods.

Each method has its advantages and disadvantages, and varies in complexity and effort to implement. The method chosen should be commensurate with the expected benefits for the specific task. For instance, rigorous modeling employing data validation, reconciliation and parameter estimation techniques requires significant effort and engineering expertise to implement, and is therefore appropriate only for key processes and important pieces of equipment. Driving information from engineering into operations forms the basis for improved understanding of asset capabilities and operational improvements. Benefits of leveraging engineering data and models include better asset utilization, reduction of maintenance costs through improved surveillance programs, and lower operating costs.

Using existing plant technology, it is possible to monitor everything from individual sensors to controllers and loops, equipment, processing units and entire plants. At a fundamental level, acquiring accurate data from sensors is important to operate the plant safely and efficiently. Acting on faulty or inaccurate data can have dire consequences. Sensor monitoring is intended to validate data and detect failure, to reduce the risk of damage to equipment, and improve product quality and plant availability.Incorporating equipment performance monitoring and maintenance requirements into the planning, scheduling and other manufacturing execution systems functions is an important strategy for manufacturers to enhance asset performance, reduce costs, improve production efficiency, improve quality and achieve higher return on assets.

Tom Fiske, [email protected], is a Senior Analyst with ARC Advisory Group Inc., Dedham, Mass.