|
Original research DEVELOPING A CONDITION-BASED PROGNOSTIC MAINTENANCE FRAMEWORK FOR A FLEET OF HAUL TRUCKS IN A SURFACE MINE ENVIRONMENTPages 325-336
Hamalwa Nestor,
Abstract
This study presents a Condition-Based Prognostic Maintenance (CBPM) framework designed to enhance the reliability and maintainability of haul trucks within surface mining operations. To overcome the inherent limitations of reactive and time-based maintenance strategies, the proposed framework integrates historical maintenance data, Failure Modes, Effects, and Criticality Analysis (FMECA), and data-driven prognostic techniques. The framework is validated through a comprehensive case study conducted at Mining Site R, utilizing three years of operational data. Analysis revealed significant performance deficiencies, with Mean Time Between Failures ranging from 22 to 59 hours (against a target of ≥50 hours) and Mean Time To Repair spanning 3 to 7 hours (against a target of ≤2 hours). The FMECA identified engine blow-by and tyre punctures as the most critical failure modes. A polynomial regression model, applied to crankcase-pressure-based health indicators, achieved a predictive accuracy of 93.8%, offering a 19-week lead time for proactive maintenance intervention. The proposed CBPM framework provides a scalable, data-driven solution for predictive maintenance in mining, with the potential to substantially reduce downtime, extend equipment service life, and improve overall operational profitability.
Keywords:
Condition-Based Prognostic Maintenance, Data-Driven Prognostic, FMECA, Reliability, Regression Analysis,
Surface Mining.
|
Michael Sony, 
PDF 