Clarifying the link between data, diagnosis and Asset Health Indices

Abstract

Within an asset management framework, such as described by ISO 55000, the role of asset condition and performance analysis is critical to success. An Asset Health Index (AHI) is an asset score which is designed, in some way, to reflect or characterize asset condition and thus likely asset performance in terms of the asset's role. A number of different approaches to generating an AHI have been proposed, and we review several here, including systems which may be: logarithmic, dynamic, weighted, or binary. Each approach has pro's and con's which are demonstrated. For a successful AHI we need to link the available raw data- whether condition monitoring or asset history or maintenance and operational data - through to likely failure modes, or issues which will affect asset performance. We discuss the relevance of particular data sources through to failure modes, and how those can be grouped into logical assessments of an asset or an asset subsystem. Such groupings make sense in terms of users of an AHI system as they deal with, for example, the electrical performance of the dielectric insulation or the thermal performance of the cooling system. Each failure mode can give rise to a likely deterioration rate, which will help direct the actions arising: to gain true value from an AHI, it is necessary to calibrate the scoring system so that similar scores have similar consequent actions and timescales. As examples of the theory and practice of AHI development we look at some individual systems, to demonstrate their practicality, including how they deal with missing data and erroneous or invalid data. Examples are given for power transformer health, focusing on the failure modes related to winding insulation, to bushings and to thermal performance, but always within an overall AHI. In each case we look at the need for an AHI to indicate a condition, and how the owner/operator planned actions and scheduled them. We further discuss the need for calibration of AHI, including both health scoring of the components of a given asset, and between different asset classes. The AHI must provide useful information for intervention: what, and how soon, whether maintenance, refurbishment or planned replacement. This paper has a practical focus, based on real data from users of AHI's in different locations, with different foci on asset performance. Data from analyses, with actions arising and interventions planned are discussed. The integration of AHI into an overall asset management approach, to make sure that the value of an AHI is realized, is presented.