Follow the Chain of Success for Oil Analysis Optimization

Bennett Fitch, Noria Corporation
Tags: oil analysis

How often should I take an oil sample? From which machines should I sample? These are typical questions I’m asked when visiting plants during the development of a lubrication program. While they are good questions, I always follow up my answers with an explanation of why effective oil analysis requires consideration of several important factors. If one of these factors fails, then the entire oil analysis program will likely fail as well.

This article will describe each of these key factors, which form a chain of success for oil analysis optimization. The three main elements in this chain involve obtaining a representative oil sample, ensuring reliable testing and determining the optimum course of action. As the illustration below suggests, each link in the chain relies on those above it. If a link fails (becoming the weakest link), then all the links below it are compromised.

Obtaining a Representative Oil Sample

Select the Right Machines for Oil Analysis

Each sample obtained for oil analysis can be costly, so sampling every machine in a facility with thousands of lubricated machines is not feasible. The best method is to determine the Overall Machine Criticality (OMC) and the Overall Lubricant Criticality (OLC), which are based on machine and lubricant failure causes in terms of probability and severity.

Takeaway: Sometimes just one finding, such as the discovery of elevated wear debris levels on a critical machine, can justify the cost of the oil analysis program. If money is spent for oil analysis on every machine or machines are poorly selected for oil analysis, these returns may not be realized.

Clean and Correct Sampling Containers and Extraction Tools

One of the main objectives of oil sampling is minimizing data disturbance. Using the right sampling tools and ensuring their cleanliness will be vital. Oil sampling bottles should be certified to one of the three cleanliness levels: clean, superclean or ultraclean. As for extraction tools, nothing in the fluid’s pathway from the machine to the bottle should further contaminate the sample and disturb the data.

Takeaway: In oil analysis, it only takes a small amount of contamination to cause concern. If the sample becomes further contaminated during the sampling process, the results can trigger premature cautionary or critical alarms.

Correctly Located Sampling Ports and Sampling Frequency

The precise location where an oil sample is extracted must be carefully chosen so the analysis results will be representative of the oil in the machine’s wear protection zones. Two samples taken from the same machine but in different locations can potentially have different results for tests such as particle counts, elemental analysis and Fourier transform infrared (FTIR) spectroscopy. Similarly, the sampling frequency should be often enough to detect dangerous spikes in unfavorable results but not too often so that time and money are wasted.

Takeaway: Both the sample location and frequency have the ability to significantly affect your oil sampling objectives but in different ways. Without the proper location, the sample result may not be representative. Without the proper frequency, a crucial oil analysis result may be missed before it is too late.

Proper Sampling Procedures

During the sampling process, an oil sample can be altered in a variety of manners, many of which are not apparent. Understanding the best method for preserving a representative sample from the machine to the bottle is imperative. Also, knowing how to obtain a proper sample is not instinctual. If samples are not drawn by the same person or the sampling procedure isn’t followed consistently, it is likely that different practices will be employed.

Takeaway: Following a proven sampling procedure is vital for one of the most important oil analysis objectives - obtaining a representative sample.

Forward Samples Immediately to the Laboratory

Just like the blood in your body, oil can offer hidden clues regarding a machine’s past, present and future condition. For example, one day you may go in for routine blood work and then receive a call soon afterward with results indicating a life-threatening condition that must be addressed immediately, even though you may not have had any known symptoms. The same is true with oil analysis. When a serious mechanical failure occurs, don’t be the one with the recent sample still sitting on your desk awaiting to be sent out for analysis.

Takeaway: Machines can be extremely expensive, and the downtime associated with their failure could be even more costly. Don’t wait any longer than necessary to assess the state of your machines. Tens of thousands of dollars could be at risk.

Related Article: Oil Sampling Do's and Don'ts

Ensuring Reliable Testing

Certified Training of Laboratory Technicians

Many people assume that their laboratory will have a staff of properly trained and certified technicians who know how to operate all the lab’s instruments, but this may not always be true. Perhaps the reason a particular laboratory is less expensive is because it is using cheap labor. A quality-control program should be in place with written procedures for consistently providing the best instruments for your oil analysis.

Takeaway: The tests performed on your oil samples deserve uniformity. Without it, accurately comparing one sample result to another will be impossible.

Optimized Selection of Tests

A typical laboratory will have an assortment of oil analysis tests from which to choose. Don’t allow the lab to select just the standard test slate for all your samples. Unless the appropriate tests are chosen for each sample point, early warning signals may be missed. More often than not, opportunities for maintenance cost savings are overlooked.

Takeaway: Test slate optimization requires a two-pronged approach with routine and exception testing. This methodology should be used with all test packages for each machine type and criticality.

Correct Sample Preparation at the Laboratory

The importance of not disturbing an oil sample during the sampling process has already been discussed. However, this would be a waste of time if the laboratory does not properly prepare the sample for analysis. Each testing procedure has written preparative instructions that must be followed for admissible results.

Takeaway: A lot of time and effort can be spent on sample preparation by the end user, but don’t stop there. Ask your laboratory the right questions to ensure it is using the appropriate sample preparation methods.

54%

of lubrication professionals perform oil analysis on equipment at their plant on an annual basis, according to a recent poll at MachineryLubrication.com

Use of Calibration Standards, Low Interference Instruments and Test Methods

You also should not make assumptions about other practices at the laboratory. Be sure the lab you select performs appropriate calibrations and utilizes methods and instruments known for low interference.

Takeaway: Due diligence is required when choosing a laboratory. Ask the necessary questions to ensure the lab will take your oil analysis seriously.

Related Article: How to Choose an Oil Analysis Laboratory

Onsite Oil Analysis and Inspections

An excellent way to enhance your oil analysis program, regardless of whether a commercial (offsite) laboratory has been selected, is to incorporate onsite oil analysis and inspection methods. Certain types of tests are quite standard and should be conducted on every sample, such as particle counts and moisture content. The instruments for these tests have dropped significantly in price, making them even more feasible for onsite use.

Takeaway: Onsite oil analysis tests and simple inspections can further optimize your ability to obtain quality information and become more cost effective over time.

Determining the Optimum Course of Action

Effective Organization of Analysis with Proper Trending

Although laboratories generally attempt to display oil analysis results in a user-friendly manner, they may not always be easily understood by end users. Comprehending how results are organized will be necessary for proper interpretation. For example, with certain types of tests, trending with graphs is required to detect a fault. Therefore, you should be aware of how the data is organized and investigate whether it can be improved for easier interpretation.

Takeaway: The organization of oil analysis results is the laboratory’s responsibility. The more effective the organization is for interpretation, the less likely a preventable machine failure will be missed.

Accurate Data Interpretation by the Laboratory

Oil analysis results typically come with a summary paragraph that suggests actions the end user should take. The accuracy of this lab-based interpretation depends on the integrity of each preceding link in the chain as well as on the information provided about the machine’s operating and environmental conditions.

Takeaway: End users often do not have enough time to interpret oil analysis results and must rely on the laboratory for interpretation. Without an accurate interpretation, effective oil analysis cannot be achieved.

Enhanced Data Interpretation by the End User

Following the laboratory’s interpretation, the end user must make the final decision. No one knows the machine’s history and application better than those who see the machine on a daily basis. If these individuals were trained to interpret oil analysis reports, the program’s success could be improved.

Takeaway: The end user is primarily responsible for the data interpretation and understands the machine’s operating conditions better than anyone else. There is not a more qualified individual to include in the data interpretation.

Take Corrective Action and Determine the Root Cause

Obtaining a representative oil sample and using a capable laboratory are pointless if corrective actions are not taken to prevent a failure. You should not expect a return on the time and money spent unless action is performed as a result of the oil analysis. Be sure to always follow up on the results and form a consensus on what the corrective action should be. Remember, the problem will only persist if the root cause isn’t discovered and resolved.

3 Objectives of Good Oil Sampling

1. Maximize data density

2. Minimize data disturbance

3. Select the proper frequency

Takeaway: Oil analysis cannot be justified without findings that lead to the prevention of a mechanical failure. Optimizing oil analysis only makes sense when it is used as a tool for greater reliability.

Continuous Improvement and Sampling Adjustments

Oil analysis never seems to get enough credit. Whenever it is effective in providing proactive maintenance, machine failures cease. Over time, people may forget that the lack of failures is attributed to oil analysis. Maintaining this awareness through continuous improvement will be key for success. Making adjustments to alarms and limits or sampling frequencies may be necessary for true optimization.

Takeaway: Just as important as determining the sampling frequency for each sample port, periodic adjustments to these frequencies should be addressed as further evidence of how change can be achieved through oil analysis.

Utilizing Oil Analysis as a Key Performance Indicator (KPI)

Oil analysis performance indicators are essential for realizing the overall health of a plant over time. In comparison, analytics are conducted on blood work results throughout a hospital or even throughout a certain geographic region to identify potential trends of a growing health concern or a dangerous outbreak. Likewise, while oil analysis KPIs may be beneficial for your machines, they also can prove to be valuable indicators of the progress made in improving your plant’s overall health as well as the advantages of oil analysis.

Takeaway: KPIs help you to recognize when there is a weak link in the chain. A performance indicator like a regression in oil analysis returns should be a call to action to re-evaluate which of the previous chain links has been compromised. Justifying a plant-wide oil analysis program will require these types of metrics.