Jun-2010

Condition Monitoring - the analysis of used oil

The elemental analysis of used lubricating oil has become an essential part of “condition monitoring”, the use of physical and chemical techniques to measure the condition of plant and equipment with the objective of preventing equipment failure and optimising maintenance programs.

Olaf Schulz
SPECTRO Analytical Instruments

Article Summary

Specialist service laboratories and major plant operators analyse hundreds of oil samples per day for a wide range of elements to detect component wear and the presence of foreign matter that may accelerate wear. There are many techniques for elemental analysis, but only Inductively Coupled Plasma — Optical Emission Spectrometry (ICP-OES) has the speed and sensitivity to succeed in these high throughput applications. In some situations, wear is so severe that quite large metallic particles can be found in the lubricating oil. The presence of such particles can indicate imminent component failure, with potentially catastrophic results, especially in critical locations such as aircraft engines. Tracing the origin of such particles can be vital if such failures are to be prevented, or investigated. Energy Dispersive X-ray Fluorescence (EDXRF) spectrometry has proved to be an ideal technique for this type of investigation. The systematic analysis of lubricating oils in service can result in lower operating costs, reduced downtimes, extended plant and equipment lifetimes and more effective maintenance programs

Condition analysis: introduction
Lubricating oil analysis has been used to monitor the condition of engines and other machinery for over 50 years. It has been likened to the use of blood tests by a clinician in determining the condition of a patient. This analogy is quite useful, as it suggests the main object of the exercise: to assess the condition of the mechanical system that the oil is lubricating, rather than that of the oil itself. It can be applied to most lubricated mechanical systems, such as engines, gear transmissions, hydraulics and the like, and has wide application in areas such as construction machinery, power generation and transport, including aviation, fleet operations and public transport.

A major objective of oil analysis is the detection of wear. There are several causes of wear, such as friction between moving surfaces, abrasion by contaminants such as grit, corrosion processes and so on, but most give rise to the presence of microscopic metallic particles in the lubricant as components wear away. Quantitative measurement of metallic elements in the oil can therefore be a useful indicator of wear. Furthermore, as different metals are used to manufacture different components, elemental analysis can often provide a clue as to which components are subject to wear. Analysis can also detect the presence and possibly the origin of foreign matter in the oil, such as dust that may have entered an engine via a defective filter. Many other changes can occur in oils under fault conditions, such as dilution by fuels, or contamination by water or antifreeze. Processes such as oxidation can lead to changes in lubricant properties such as viscosity, leading to accelerated wear rates. Not all these processes can be detected by elemental analysis, so several different physical and chemical measurement techniques are necessary for comprehensive condition monitoring, but elemental analysis has become the essential tool for wear detection. Interpretation of the analytical results from oil analysis is itself a complex and specialised task. Many lubricating oils contain additives to improve their properties that are themselves metallic compounds, and some of these metals may also occur in the wearing components themselves. Therefore the presence of a particular element does not necessarily indicate wear. Indeed, these additives are used to improve or extend the lubricant properties of the oil, and may be consumed over time. This is known as “additive depletion”, and unless the oil is changed or the additives replenished may itself lead to increased wear, so the level of additives needs to be monitored. Mechanical systems and engines are often subjected to a running in period during which wear can be quite rapid but is actually beneficial. For these reasons, system management decisions are rarely made on the basis of single oil analysis measurements or against predetermined “limit values”, but by following trends established by regular sampling regimes. Software packages designed to assist in the interpretation of measurement data are commercially available.

Unless wear is severe, metallic particles entering the lubricant are usually very finely divided (10 microns or less) and remain largely suspended in the oil without settling out. Oil samples like this can be treated essentially as solutions and are amenable to analysis by several well established laboratory techniques. In more severe wear, larger particles can be produced, that settle out and require a different approach. Large metallic particles in any lubricating oil are a cause for concern, and one popular technique is the use of magnetic sump plugs to “harvest” such particles for subsequent analysis, and hopefully identification of their origin.

Table 1 indicates the possible significance of some elements found in used oil: it is by no means exhaustive.

Analytical considerations and techniques
In the analysis of used lubricating oils, the analytical tasks can be roughly divided into two areas, the direct analysis of the oil itself, and the analysis of particles found in the oil.

As mentioned above, fine wear metal particles remain suspended in the oil. Additive elements are usually in solution. Under these circumstances the oil sample can be regarded as homogeneous and analysed by solution techniques. Typical concentration levels for wear metals lie in the range from 1 to 500 parts per million, and some additive elements can be at few thousand ppm. For most elements, these concentrations are well within the scope of spectroscopic techniques such as ICP-OES and EDXRF. Both techniques can be used satisfactorily when sample numbers are relatively small, but when high sample throughput is required the relatively high speed of simultaneous ICP-OES has made it the technique of choice in service laboratories, particularly as the sample preparation required is usually limited to a simple dilution with a solvent such as kerosene. The Spectro Arcos simultaneous ICP-OES spectrometer can process up to 80 samples per hour and can be combined with automatic sample handling, so fits this requirement perfectly. The Spectro Genesis has a throughput of up to 20 samples per hour.

The requirement for speed and the need to measure many elements in each sample means that sequential techniques like Atomic Absorption Spectrometry (AAS) and sequential ICP-OES are usually considered too slow for high-throughput applications. ICP-OES spectrometers usually require a laboratory environment and some logistical support. Sometimes distance or the need for a rapid response — aviation or motor racing might be examples — may necessitate the use of more portable or rugged instrumentation. Even quite sophisticated EDXRF systems such as the Spectro Xepos can be used in these situations, and small portable instruments such as the Spectro xSort are useful for rapid screening measurements. Another technique that can be used in these situations is the use of small optical emission spectrometers equipped with a rotating disc electrode (RDE), or “Rotrode”. Rotrode instruments are typically operated manually and do not offer the sensitivity and stability of ICP-OES systems.