Development of an integrated model for assessment of operational risks in rail track

摘要

In recent years there has been continuous increase of axle loads, tonnage, train speed, and train length which has increased both the productivity in the rail sector and the risk of rail breaks and derailments. Rail operating risks have been increasing due to the increased number of axle passes, sharper curves, wear-out of rails and wheels, inadequate rail-wheel grinding and poor lubrication and maintenance. Rolling contact fatigue (RCF) and wear are significant problems for railway companies. In 2000, the Hatfield accident in the UK killed 4 people, injured 34 people and led to the cost of £ 733 million (AUD$ 1.73 billion) for repairs and compensation. In 1977, the Granville train disaster in Australia killed 83 people and injured 213 people. These accidents were related to rolling contact fatigue, wear and poor maintenance.Studies on rail wear and lubrication, rolling contact fatigue and inspection and rail grinding analyse and assess the asset condition to take corrective and preventive measures for maintaining reliability and safety of rail track. Such measures can reduce the operational risks and the costs by early detection and prevention of rail failures, rail breaks and derailments. Studies have so far been carried out in isolation and have failed to provide a practical solution to a complex problem such as rail-wheel wearfatigue-lubrication-grinding-inspection for cost effective maintenance decisions. Therefore, there is a need to develop integrated economic models to predict expected total cost and operational risks and to make informed decisions on rail track maintenance.The major challenges to rail infrastructure and rolling stock operators are to:1. keep rolling contact fatigue and rail-wheel wear under controllable limits,2. strike a balance between rail grinding and rail lubrication, and3. take commercial decisions on grinding intervals, inspection intervals, lubrication placements, preventive maintenance and rail replacements.This research addresses the development and analysis of an integrated model for assessment of operational risks in rail track. Most significantly, it deals with problems associated with higher axle loads; wear; rolling contact fatigue; rail defects leading to early rail replacements; and rail breaks and derailments. The contribution of this research includes the development of:failure models with non-homogenous Poisson process and estimation of parameters.economic models and analysis of costs due to grinding, risks, downtime, inspection and replacement of rails for 23, 12, 18 and 9 Million Gross Tonnes (MGT) of traffic through curve radius 0-300, 300-450, 450-600 and 600-800 m; and application of results from this investigation to maintenance and replacement decisions of rails. Cost savings per meter per year are:* 4.58% with 12 MGT intervals compared to 23 MGT intervals for 0-300 m* 9.63% with 12 MGT intervals compared to 23 MGT intervals for 300-450 m* 15.80% with 12 MGT intervals compared to 23 MGT intervals for 450-600 m* 12.29% with 12 MGT intervals compared to 23 MGT intervals for 600-800 m. a lubrication model for optimal lubrication strategies. It includes modelling and economic analysis of rail wear, rail-wheel lubrication for various types of lubricators. Cost effectiveness of the lubricator is modelled, considering the number of curves and the total length of curves it lubricates. Cost saving per lubricator per year for the same curve length and under the same curve radius is:* 17% for solar wayside lubricators compared to standard wayside lubricators. simulation model for analysis of lubrication effectiveness. Cost savings per meter per year for:* 12 MGT grinding interval is 3 times for 0-450 m and 2 times for 450-600 m curve radius with lubrication compared to without lubrication.* 23 MGT grinding interval is 7 times for 0-450 m and 4 times for 450-600 m curve radius with lubrication compared to without lubrication. a relative performance model, total curve and segment model. an inspection model for cost effective rail inspection intervals. Cost savings per year for same track length, curves and MGT of traffic:* 27% of total maintenance costs with two inspections, compared to one inspection considering risk due to rail breaks and derailments. a risk priority number by combining probability of occurrence, probability of detection and consequences due to rail defects, rail breaks and derailments. integrated model combining decisions on grinding interval, lubrication strategies, inspection intervals, rectification strategies and replacement of rails.Cost saving per meter per year for 12 MGT is:* 5.41% of total maintenance costs with two inspections, compared to one inspection considering risk due to rail breaks and derailments.* 45.06% of total maintenance costs with lubrication for two inspections, compared to without lubrication.Cost saving per meter per year for 23 MGT is:* 5.61% of total maintenance costs with two inspections, compared to one inspection considering risk due to rail breaks and derailments.* 68.68% of total maintenance costs with lubrication for two inspections, per year compared to no lubrication.The thesis concludes with a brief summary of the contributions that it makes to this field and the scope for future research in wear-fatigue-lubrication-grinding-inspection for maintenance of rail infrastructure.