Procuring The Highest-Quality Oil Sample

Oil samples can reveal a lot about the condition of your equipment. Make sure they’re accurate.

As in all aspects of life, the end result of any endeavor is only as good as the effort put into the exercise and the quality of elements used to create the result. Such is the case in lubricant and wear-particle analysis. Here, accuracy of results is highly dependent upon the care and method used to collect and then deliver a quality used-oil sample into the hands of a laboratory for analysis.

Procuring and delivering an analysis-ready, superior-quality used-oil sample requires discipline and consistency, as summed up in the “7 Best-Practice Principles of Oil Sampling” in Table I. Choosing the best procedure, method, hardware and sample location is key. These choices will likely differ based on whether they are taken from a pressurized or non-pressurized system, and whether the machine or gearbox is designed or set up for best-practice sampling techniques. They may also differ due to the consistency of the sampling methods, the training of the person taking the sample and the sampling cleanliness protocol used.   

The best sample choices for a piece of equipment are driven by three main objectives: 1) to maximize sample data density; 2) to minimize sample-data disturbance; and 3) to maximize sampling consistency.

Data density

Each oil sample carries a unique time-stamped composition signature of base oil chemistry, additive-package level and chemistry, and wear-particle type, size and count. These factors are then compared against a virgin oil sample to determine the oil’s chemical condition and the machine’s moving-parts condition at that moment in time. In turn, the more representative the sample is, the more accurate the diagnosis. Every sample, consequently, must contain the maximum amount of data density (representative data) it can—which is best achieved by extracting the sample in the most appropriate place.

For pressurized systems, e.g. hydraulic, and recirculating oil systems, oil is pumped from a reservoir under pressure, through a series of filters in a piping distribution system to the bearing surface areas, from where it is returned to the reservoir to be once again filtered and cooled for recirculation. Maximum data density is always found downstream of the lubricated bearings and upstream of the return filter, where it is laden with contaminants that have just been washed from the bearing surfaces. To assure the most representative sample, take it:

• When the machine is running at temperature and under regular working condition load.
• From a live fluid zone, meaning no dead pipe legs (static areas) or line ends.
• From a sample port connected to an elbow used to create a turbulent zone and ensure a colloidal (well-mixed) sample.

Samples can be extracted in a low-pressure (LP) system using a simple ball valve drain tap screwed into an elbow. For high-pressure (HP) systems, a ball valve can still be employed, but with the addition of a helical coil attachment used to reduce the pressure of the fluid stream once the valve is opened. A more sophisticated way to take HP samples is to use a vacuum pump connected to a push-style sample port (similar to the way a grease nipple works): The probe attached to the pump is inserted into the spring-loaded sample port to allow pressurized oil to flow into the sample bottle that’s screw-attached to the vacuum pump unit.

For non-pressurized systems such as a self-contained splash- or bath-lubricated gearbox, a sample can be extracted three ways. The first (and least desirable) method uses a simple ball valve screwed into the reservoir drain port. Although easy to set up, a large flush volume is needed prior to taking the actual sample—and the user still runs a high risk of picking up sludge contamination from the bottom of the reservoir. (To lessen this risk, a pilot sample tube can be inserted to the one-third level mark of the reservoir.)

The second method employs a drop tube attached to a rod to ensure the tube opening is approximately positioned at the one-third reservoir level mark when the tube is lowered into the reservoir through a fill opening. This is done to help ensure no non-representative sludge contamination is allowed in the sample. The tube is then connected to a suction or vampire pump to extract the sample. Again, sample disturbance can be high if the sampling procedure is not performed carefully.

The third and ideal sample method employs a combination pilot-tube/level-gauge device affixed at the correct reservoir sample level. As most reservoirs don’t come with such devices, this approach will require an after-market equipment purchase and installation

Data disturbance

It’s important that your oil-sample data be neither disturbed nor contaminated by the actual sampling and sample-handling processes. For example, if they’re not minimized, reservoir sludge, dirty sample/drop tubes and dirty sample bottles can all distort data readings. Simple, but effective, tactics for managing data disturbance, sometimes referred to as “interference,” include:

• Cleaning hands, cleaning the sampling port/area, cleaning sampling equipment.
• Using only virgin sample bottles designed for oil-analysis sampling (glass is preOnly filling sample bottles 60% to 70%, providing headspace that lets the lab agitate and successfully re-suspend the solids for testing purposes.
• Performing the 10x flush rule for every sample, e.g., flushing the sample valve and tube (when used) with approximately 10x the required sample-volume space of the oil that’s to be sampled into a non-sample container before the real sample is taken.
• Using a ziplock sandwich bag as a glove to handle clean sample containers, that when filled, can be stored untouched, ready for shipping (thus minimizing the time sample bottles are open to the elements).

Sampling consistency

To ensure high-quality sample results that can be trusted, the sampling protocol must assure consistency. This is achieved by:

Developing an engineered oil-sampling program in which every sampling port and method is documented and regular sampling frequencies are set up in a work order system. (Commencing such a program, bearings are usually start-sampled on a 500-hr. frequency; industrial hydraulic systems on a 700-hr. hour frequency, light-duty gearboxes on a 1000-hr. frequency; and heavy-duty gearboxes on a 300-hr. frequency.)

• Using an oil-sampling program to develop, as well as train on, standard operating procedures;
• Always sampling from the same location.
• Regularly sampling virgin-oil when new lubricant stock arrives on site.
• Using the same laboratory for sampling, and ask for dedicated lab technician(s) to perform your plant’s sampling.
• Always filling in the sample-data form accurately, including sampling date and time stamp.
• Sending a sample to the lab within 24 hours of its collection (if longer than 24 hours, the sample must be retaken).