Purpose: The application of condition-based maintenance on selected equipment can allow online monitoring using fixed, half-fixed or portable sensors. The collected data not always allow a straightforward interpretation and many false alarms can happen. The paper aims to discuss these issues. Design/methodology/approach: Statistical techniques can be used to perform early failure detection. With the application of Cumulative Sum (CUSUM) Modified Charts and the Exponentially Weighted Moving Average (EWMA) Charts, special causes of variation can be detected online and during the equipment functioning. Before applying these methods, it is important to check data for independence. When the independence condition is not verified, data should be modeled with an ARIMA (p, d, q) model. Parameters estimation is obtained using the Shewhart Traditional Charts. Findings: With data monitoring and statistical methods, it is possible to detect any system or equipment failure trend, so that we can act at the right time to avoid catastrophic failures. Originality/value: In this work, an electro pump condition is monitored. Through this process, an anomaly and four stages of aggravation are forced, and the CUSUM and EWMA modified control charts are applied to test an online equipment monitoring. When the detection occurs, the methodology will have rules to define the degree of intervention.
This paper describes the analysis, from a statistical point of view, of a maritime gas turbine, under various operating conditions, so as to determine its state. The data used concerns several functioning parameters of the turbines, such as temperatures and vibrations, environmental data, such as surrounding temperature, and past failures or quasi-failures of the equipment. The determination of the Mean Time Between Failures (MTBF) gives a rough estimate of the state of the turbine, but in this paper we show that it can be greatly improved with graphical and statistical analysis of data measured during operation. We apply the Laplace Test and calculate the gas turbine reliability using that data, to define the gas turbine failure tendency. Using these techniques, we can have a better estimate of the turbine’s state, and design a preventive observation, inspection and intervention plan.
Vibration Analysis (VA) is now routinely used for condition monitoring and failure diagnosis in Condition Based Maintenance (CBM). In the context of VA, a methodology is proposed, based on biplots, to simultaneously display both vibration frequencies and their measurement points, in support of monitoring and diagnostics tasks.In this research, real observational data obtained measuring mechanical vibrations on four generators aboard a Portuguese Navy Ship in real operating conditions is used. A portable vibration collector was employed, and the measurements were taken at 13 measurement points in each one of four generators, using the same collector settings. Spectrograms resulting from vibration measurements were transformed into biplots and used for decision support according to the proposed methodology. Data analysis showed a robust stability in the macrostructure of biplots when observations resulting from different generators of the same model and at the same assumed conditions was analyzed. This invariance allows the specification of reference conditions, rules to detect changes of operating conditions and the emergence of failures. The proposed methodology, once embedded in dedicated software, will reduce the interpretation error in diagnosis and prognosis associated to variability in personnel training and experience. Consequently, it will increase the safe use of VA in an increasing number of situations.