More and more of the traditional tests done in an external lab can be done using online or on-site technologies, stated Krethe, who is based in Brannenburg, Germany. However, while this helps to get information much earlier, the interpretive knowledge of experienced engineers from the external lab is missing and must be replicated in-house. Sensors provide a real-time picture of oil condition and can go a long way to ensure long oil life and reliable operation by allowing corrective action to be taken immediately, he said.

There are some trade-offs, however. Krethe noted that not all miniaturized methods deliver the same precision as well-established lab devices, but they are suitable for monitoring oil aging trends. Also, a single device usually cannot provide a complete picture of oil condition. Methods have to be combined to provide the most effective maintenance tool, he told the Colloquium, which was held in January at Germanys Technische Akademie Esslingen.

For the most reliable oil monitoring program, its best to combine online or on-site methods with detailed lab analysis, he recommended. With this approach, the sensors provide immediate warning of potential problems, while the lab analyses help identify and track the underlying cause of the problem.

This combines the advantages of all methods, providing safe and cost-effective condition monitoring for the oil, Krethe added. He went on to describe the types of online sensors that are available today and how they correlate with established laboratory tests or failure modes.

During the last several years, Krethe said, a number of online sensors have been developed. He classified them into two broad groups, single-value sensors and condition sensors.

Single-value sensors collect information on a single oil or operating parameter, such as viscosity, water (humidity or relative saturation), cleanliness or soot.

Condition sensors can look at the oils base number and acid number, its electrical conductivity or permittivity.

The advantages of using online sensors are that they provide continuous, real-time detection of oil condition. In addition, sensor signals can be integrated into a warning system or automated maintenance system to provide instantaneous response to changes in oil condition. The user can set the limit (such as a certain percent of water content) and when that limit is exceeded, the alarm trips to alert the operator to take action. Finally, data gathered in real-time by the sensor can be processed to gain greater insight into system operation.

Disadvantages or limitations include the fact that sensor output often does not correlate exactly with well-known lab techniques. Also, sensor calibration is sometimes complicated and often may be valid only for a specific oil or blend.

In addition, relying on a single, isolated measurement can lead to misunderstandings about actual oil condition. A limited number of single tests cannot provide a detailed picture of the complex oil aging process, emphasized Krethe. Finally, sensor location is critical to provide an accurate picture of oil condition. A sensor upstream of a filter in a hydraulic system will give a very different picture of the fluids condition than a sensor placed after the filter, for example.

Oil aging is not a simple one-step process, and a number of factors can rapidly or slowly undermine the lubricants health. Oxidation, nitration, contaminants like soot and water, and thermal changes all take their toll on its quality. After describing these oil aging factors, Krethe presented an overview of the parameters typically tested to monitor oil degradation.

Various options – online, on-site or at an outside lab – are available to measure the parameters (see table, page 24). Even more sensors are still under development, while others have reached the field-test level. Also, he said, The usability of the sensor and economic feasibility have to be evaluated for every application.

Special test kits are available for on-site measurement of acid and base number. They require the user to mix chemical agents together to produce a color indication. However, not all kits can handle all base oil types.

Another option for on-site and online testing of acid number, base number and other oil related values is Fourier transform infrared spectroscopy (FTIR). Once largely the domain of full-scale laboratories, FTIR can now be found in portable set-ups and even in-line sensors. However, while FTIR is a powerful tool, its precision depends on skill and the availability of reference spectra to calibrate the tool. The mixture of different oils can also affect precision, said Krethe. In addition, miniaturized FTIR sensors do not work reliably with dark oils.

The precision of viscosity measurements is highly dependent of on viscosity range. Some sensors do not produce reliable results with highly viscous fluids.

Finally, Krethe explained, Sensors based on measuring changes of conductivity and permittivity are sensitive to oil contamination. And particle counters used to determine the lubricants cleanliness class can detect wear particles but not differentiate between wear particles and dirt. If the contaminants are wear particles, most users would want to know immediately so they can take action before the equipment fails.

Despite these issues, the importance of online sensors will greatly increase in the future, said Krethe, because effective oil condition monitoring saves money, and the costs of online sensors are more than offset by the benefits of using them.