On-site oil analysis laboratories are becoming more popular as reliability professionals realize the profits a well-designed and well-managed lab can reap. The strengths of an on-site lab include rapid sample turnaround and the rich, environment-relevant diagnostics enabled by in-house staff who are familiar with the site's machines.
Ultimately, the goal of any project is to boost the bottom line of the organization. Setting up an on-site lab is no exception, and careful planning will help ensure the success of this project. Designing and commissioning a lab is an exciting task and can result in a competitive addition to the organization.
This article considers important design factors of an on-site laboratory, including sample reception, analysis, and storage and personnel requirements. Many options catering to various budgets are presented, leaving room for future expansion without duplicating existing capabilities.
Viability of On-site Laboratory
The main benefits provided by an on-site laboratory include:
Rapid sample turnaround time
In-house knowledge of machinery and environmental conditions guides, test selection and interpretation
Rapid quality control of lube oil purchases
As with any project, a thorough viability study must take place. The author developed a rough model that resulted in some sample numbers. The following assumptions were made:
Current program with a commercial lab:
Sample rate is 100 samples per month at a commercial oil analysis (OA) lab.
Average sample cost is $25.
Annual benefits from the current program are estimated at $200,000.
Proposed program with an on-site lab:
Sample rate is 200 per month. (Rate would be higher than using a commercial lab and the analysis facility is more accessible).
20 percent of the 200 samples are still being analyzed by the commercial lab.
Average price of a commercial lab sample stays the same at $25, but fewer of the basic tests and more of the advanced tests are being performed.
Annual benefits are estimated at $500,000 due to the increased sample analysis rate and increased problem-detection rate as a result of introducing new technology on-site.
Sample extraction and lab technician costs were factored into the model, which highlighted the following:
Cost per sample was nearly halved. The cost of analyzing 240 samples per month was $37,000 as opposed to 100 samples for $32,000.
The percentage increase of return on investment was 276 percent.
The model is presented in Table 1.
An on-site laboratory needs an operable environment. Some factors to consider include:
Physical Location. The point of operation would ideally be in a geographically central location, likely near the new-lubricants-receiving depot, so that new oil samples can be run while the lubricant distributor is on-site.
Electrical Supply. Check local, state and federal regulations for electrical supply requirements. Flame- or explosion-proof electrical fittings may be required.
Solvent Storage. Verify regulations for solvent storage and secondary containment requirements. Additional plant requirements may also be in effect.
Laboratory Size. If the laboratory runs successfully, expansion will likely occur. Ensure sufficient room is available for this.
Ventilation. Ensure adequate ventilation is in place. Fume hoods are necessary for the majority of tests that should be run.
Climate Control. The environment should be clean, and temperature and relative humidity should be maintained at near constant levels. Having a consistent climate environment is important for the reproducibility of some tests.
Storage. Ensure adequate internal storage in the form of drawers and closets for laboratory consumables.
Ideally, the plant will contain a room that is inexpensive to convert to a laboratory. Alternatively, an existing quality control (QC) lab located on-site might have extra space that could be used. If this is the case, ensure the operations of the QC and OA labs do not interrupt each other.
A third alternative would be to purchase a prefabricated laboratory and install it at a convenient location on-site.
The laboratory should be designed to facilitate a logical flow of work around the lab (Figure 1).
Mechanical agitation is the first process performed. Immediately thereafter, routine tests, which are sensitive to particle suspension, are carried out. Routine tests that are insensitive to particle suspension follow. The final tests are typically performed by exception only.
Sufficient storage space should be available to keep previously analyzed samples for at least three months.
The selection of suitable personnel, and their training, is essential to the successful operation of an on-site laboratory. In fact, inadequate personnel resources are the largest contributing factor to the failure of what would otherwise be a successful project.
Typically, the laboratory staff includes a mechanic who has either ended up in the position by default, or for some medical reason can no longer carry out the physical responsibilities of a mechanic. There is nothing wrong with this; in fact, it is a good thing because the mechanic will bring machinery experience and knowledge of the plant's environment into the equation. However, there are two reasons why this does not always work out:
Lack of laboratory training. This is not part of the typical mechanic's training, and without it, it is hard to develop and maintain rigorous and repeatable laboratory procedures.
Lack of authority in decision making. This is probably the biggest contributing factor to the failure of well-designed programs. If the diagnostician does not have the authority (perceived or real) to make the interpretation necessitated by the results, interpretations will likely be tailored to an idea of what might be carried out, not what must take place. These substandard diagnoses lead to a loss of confidence in the technique and often the abandonment of the project. This is management, not an operational constraint, and it's an issue that must be addressed and resolved early in the project-planning phase.
If the lab tech does not have the position in the management structure to believe his or her work will be taken seriously, he or she will need the support of someone with that authority. There is nothing wrong with questioning an interpretation; in fact the system should encourage it. But if there is an organizational bias to ignore the interpretations, the on-site lab project has little chance of succeeding.
Another factor to consider in making personnel decisions is certification. Certification brings a benchmarked level of authority to the program and encourages confidence among the people who use the lab services. The individual performing the analysis should have ICML Level I Laboratory Lubricant Analyst (LLA) and Level I Machine Lubricant Analyst (MLA) certifications, or equivalent. Ideally, the program manager would have Level II MLA or equivalent.
Some possible test options that should be reviewed for use in an on-site laboratory will now be discussed.
No on-site laboratory would be complete without particle-counting capabilities. Three available options to consider include patch microscopy, pore blockage particle counting and optical particle counting.
The simplest and most cost-effective option is a patch maker and microscope combination. This option should be considered either in startup situations or where the oils being analyzed have high contamination levels of dirt and/or water. The method for preparing the patch will be discussed in a follow-up article in an upcoming issue of POA. While not a true particle counting method, decent approximations of particle counts can be estimated using charts. Available on CD-ROM, an example of this chart is presented in Figure 2, which may be used to estimate the particle count from the patch.
Even if a particle counting program must be started with this method, the patch-making apparatus will not be wasted upon an upgrade to another option.
Pore Blockage Particle Counting
The second option to consider for particle counting is the pore-blockage particle counter. This particle counter determines particle contamination by establishing pressure increase or flow decay across a sensor as the grid fills up with particles. Pore-blockage particle counters are suited for moderately dirty fluids that typically have air and/or water contamination and dark-colored fluids such as engine oils and emulsions. Figure 3 illustrates the pore blockage particle counter's principle of operation.
Optical Particle Counting
The third option for particle counting, and in most cases preferred by the serious oil analysis practitioner, is the automatic optical particle counter. This technology typically uses laser or light blockage to count and size particles. The laser-scattering device is illustrated in Figure 4.
Unfortunately, automatic optical particle counters have an inclination to count water droplets and air bubbles as particles. Therefore, for this technique to be used effectively, water-masking and degassing techniques must be employed. Although this is extra overhead for the on-site laboratory, the benefits of accurate particle counting are high enough that the additional preparation is justified.
In most cases, optical particle counting is the most desirable option. However, proper adherence to sample preparation is essential - if this is not done, the importance of this test will be lost. Sample preparation includes the following:
Agitation. Manual agitation is not sufficient and a paint shaker must be used.
Water Masking. Because an optical particle counter counts water droplets as particles, the water needs to be dissolved into the oil to remove the interference.
Degassing. An optical particle counter also counts gas bubbles as particles. These bubbles need to be removed either by vacuum, ultrasonic bath or a combination of both.
Ferrous density is the quantification of the amount of ferrous debris in an oil sample. The technique is suited to large-particle detection like abnormal wear detection. Various means of determining ferrous density are available using magnetism to either trap magnetic particles or directly measure magnetic strength of the particles using the magnetic Hall effect.1
Ferrous density information is useful because its presence in an on-site laboratory is desirable. Due to the moderate expense of ferrous density testers, they are likely located in only the more sophisticated on-site labs. However, if gearboxes make up a high percentage of the samples being analyzed, ferrous density testing should be considered essential.
There are three types of commonly used ferrous density testers. The first is direct-reading (DR) ferrography. With this method, particles are trapped by magnetism on a glass slide and photocell detectors are used to measure the concentrations of large and small particles. Although this method is time-consuming, it provides a ratio of large to small particles, enabling the assessment of the severity of a wear situation. DR ferrography is suited to exception testing on most samples (generated by an increased particle count), but should also be used on expensive and critical pieces of equipment.
The second type of ferrous density testing uses the Hall effect. As a ferrous-containing sample is passed over a coil that is passing a current, a voltage is induced. The voltage is proportional to the ferrous content of the oil. This test is quick and is ideally suited to routine testing. Unlike DR ferrography, it does not provide the ratio of large to small particles, merely a single index. Therefore the ferrous density readings should employ an exception test, either patch testing or analytical ferrography, to further investigate nonconforming readings. Such testers are available in bench-top or portable versions.
The third type of ferrous density testing is a variation on the particle count, called a ferrous particle count. A magnet is used to hold back ferrous particles while a particle count is performed. This particle count represents the concentration and distribution of nonferrous particles in the oil. The magnet is then removed and a second particle count is performed. This represents all particles in the oil. Through subtraction, the ferrous-to-nonferrous ratio of particles, as well as the size distribution of the ferrous particles, can be determined.
An on-site oil analysis laboratory is a useful addition to the condition monitoring arsenal. However, users must be aware of the failure potential and the root causes that may be attributed to this, including incorrect instrumentation selection, incorrect test selection, poor personnel and information-flow management, and incorrect or no assimilation with a commercial laboratory. Understanding the pitfalls of embarking on such a project is vital to ensuring its success.
1. "Wear Analysis." Practicing Oil Analysis magazine. September 2002. Noria Corporation. Tulsa, OK.
2. Fitch, J. The Lubrication Field Test and Inspection Guide. Noria Corporation. Tulsa, OK. 2000.
3. Info Product Sites Inc. 2003..
4. Troyer, D. "Looking Forward to Lubricant Oxidation?" Practicing Oil Analysis magazine. March 2004. Noria Corporation. Tulsa, OK.