Most frequently, offline condition monitoring of lubricants is performed on a calendar-based schedule. During defined intervals (e.g., monthly), oil samples are drawn and subjected to detailed chemical and physical analyses. In addition to offline lubricant assessment, other sensors like vibration, particle and temperature sensors can be utilized for online condition monitoring – for damage detection – of the rotating equipment; more recently complex electrochemical impedance sensors are available for early warning of a multitude of parameters. However, the large-scale implementation and acceptance of new sensor technologies can take time, especially when a multitude of parameters are being measured and digitalization plays an ever increasing role in this process.
The general trend of lubricant condition monitoring is to move from calendar-based to evidence-based maintenance with the goal to accelerate sustainability. Reducing the incidence of additional offline oil monitoring can prolong an oil’s service life and thus reduce costs and contribute towards sustainability. One large part in calendar-based lubricant monitoring is the underlying logistic issue of accessibility. Depending on the accessibility of the rotating equipment, it can take significant resources to obtain and transport the samples to a lab for further analysis. Thus, it is most favorable to use sensors that are simple in design to proactively trigger an alert of knowing when a sample should be obtained and transported to the lab.
Because offline lubricant analysis provides extensive insight into the lubricant’s characteristics, it is highly unlikely that current online lubricant sensors will eliminate the need for the detailed lab results any time soon. We believe that a cost-effective and straightforward sensor should be used as a trigger to initiate a detailed oil sample analysis. The challenge is to further process the obtained chemical and physical properties from that analysis in order to provide an advanced insight as to the lubricants’ status in which a normal oil report cannot produce. Such an application must be able to proactively determine the oil’s stability and beneficial properties at the molecular level so as to provide a concrete risk assessment; not every alert causes a system risk as false positives can occur which highly depends on the lubricant’s molecular composition. That is, just because an additive depletion in lubricant A can be detrimental to the rotating equipment, it may not be the case with lubricant B in the same application. Thus, post processing of a very detailed oil analyses is of highest value for proactive condition monitoring.
At any stage, it is important to combine and assess the output data form all available sensors to paint an entire picture of the rotating equipment’s performance. The tribological behaviour at the rotating interface is a cascading process and a lack of understanding of lubricant chemistry and how it relates to a boundary/mixed lubrication regime can facilitate failures in rotating equipment. Thus, understanding and digitizing the detailed chemical properties at the molecular level and its relation to trans-scale bearing failure is critical as to its effects on early stages of damage initiation. We are proposing that this is the “Missing Link” in predicting rotating equipment reliability prior to operation and more importantly, during smooth operation as the lubricant chemistry is constantly changing throughout the service life.
In order to provide proactive Reliability-as-a-Service, the lubricant chemistry must be considered in detail. Complex and large-scale computational capabilities are needed to process the molecular characters of the lubricants which has remained a significant challenge. However, we have overcome this challenge with real-time complex system analysis and applied it to the molecular behaviour at the interface accounting for cascading events. Our Reliability-as-a-Service solution, SeerWorksTM Reliability, combined with a simple cost effective and highly sensitive online sensor will trigger a real-time online alert that promotes a subsequent offline oil analysis. By doing so, proactive evidence-based condition monitoring is promoted and allows the operator to optimize a large part of existing calendar-based condition monitoring protocols.