Concluding on the best possible strategy for rail maintenance, Rasmussen points out that the ideal solution often is a combination of reactive, condition-based and predictive maintenance, the latter having the greatest efficiency. The most important aspect is to have a model for efficient predictive reprofiling that prevents as many surface defects as possible from initiating while at the same time limiting reprofiling activities. The trick here is to only reprofile when and where it is needed, he points out.
“By using an efficient predictive reprofiling approach, it is possible to prevent most of the surface defects from initiating. Despite an efficient model to predict defects and identify when reprofiling is needed, it must be expected that not all defects are possible to prevent. Therefore, it is recommended to combine it with an efficient condition-based reprofiling strategy, which requires efficient monitoring of the rail surface quality to detect surface defects at as early a stage as possible.”
Rasmussen adds that even with an efficient combination of predictive and condition-based reprofiling, it must be expected that some costly repairs of the tracks will be needed. The aim is to limit repair and replacement of rails due to surface defects as much as possible and instead extend the lifetime of the rails. If the surface defects can be reduced by 75 per cent using predictive reprofiling, and an additional 20 per cent by condition-based reprofiling and the last five per cent by reactive repair or rail replacement, he would describe that as a very successful strategy.
Rasmussen underlines that data-driven models and strategies for predictive and condition-based reprofiling depend on the properties of the specific lines, conditions and traffic loads, and access to knowledge and reliable data is crucial if this strategy is to succeed.
“To exploit all relevant data from a range of different sources in an efficient way, especially for predictive and condition based reprofiling, a linear asset management (LAM) system is required. LAM systems such as IRISSYS, OPTRAM or ZEDAS are developed to gather all data in a railway network in one system, which makes it possible to coordinate data and, among other things, identify correlations. A LAM system is a fundamental tool in achieving an efficient data-driven strategy for predictive and condition-based maintenance, especially for rail reprofiling,” he says.
Rasmussen emphasises that future mobility must be sustainable. A key factor is initiating a sustainable transformation of public transportation. In Denmark and many other countries, billions have been invested in having railways play a leading role in the green transition of transportation. And the ability to extend rail lifetime is an important part of achieving this goal.
“Rail production requires a huge amount of energy, causing very high CO2 emissions. The emissions from production of one kilometre of rail are estimated to be 136 kg CO2e. By implementing an efficient strategy for rail maintenance, it is often possible to increase rail lifetime by 50-100 per cent, resulting in a corresponding reduction of CO2 emissions due to the reduced need to produce new rails. An extended rail lifetime therefore benefits progress towards a more sustainable world,” Carsten Jörn Rasmussen concludes.