February 4, 2016

The Anatomy of Oil Analysis Response

Data Interpretation, Decision and Response Workflows

The basic definition of reliability could be that something (or someone) does what is expected, when it is expected. A car that starts when you turn the key, takes you where you want to go, and returns you home without failure would be described as reliable. A compressor that provides a suitable flow rate of gas at a suitable pressure, on demand and without fail would also be describes as reliable. Reliability itself isn’t difficult to understand, but improving reliability is far more complex. There are multitudes of tools and techniques, philosophies and approaches that can be applied, and within these effective paths to get to the goal of improved reliability. If one of your tools is oil analysis, this article will help you find that path.

Data Interpretation

“It is a capital mistake to theorize before one has data.” – Sir Arthur Conan Doyle

An oil analysis report, even a basic one, is full of data. Even basic oil analysis will have up to 40 data points for the current sample, as well as results from a number of previous samples. All of the data provided is important, but there are effective and efficient ways to work through it.

Understand what it is you are sampling. Understanding the equipment that is being sampled is critically important. Different components operating under different conditions will generate different results, all of which might be viewed as normal given when and how the sample was collected. Maintenance history like filter changes or oil top-ups will have an effect on data interpretation. Knowing if the equipment has been run particularly hard or has sat idle for long periods will inform the results. Context is necessary to understanding.

At Fluid Life, we have a database of close to 7 million oil samples on a wide range of equipment. It is likely that we not only have an extensive history of results from your exact make and model of equipment, but for that equipment in your industry. Using statistically-based flagging limits, you can know what is normal and what is not.

Read your results immediately. Your oil analysis program exists to provide you information about the internal condition of your equipment. Too often the story is told of equipment that catastrophically failed, and when an investigation is done it is found that the oil analysis (or vibration analysis, or thermography, or other condition-monitoring tool) report correctly identified and even tracked the deterioration of the equipment. The report was simply not read in a timely manner, thus the proper steps to prevent the failure were not taken.

Fluid Life’s CARE (Condition Assessment and Reliability Evaluation) program ensures that your oil analysis results are reviewed as soon as they are available by our team of dedicated Reliability Specialists, even when you cannot. When critical issues emerge, you are contacted immediately with not only an expert assessment of the analysis results, but informed advice on what next steps should be taken.

Identify critical results. Regardless of what lab you use, obvious clues will be provided to you to help you determine which sample results are severe, unacceptable, reportable, or normal. Some labs will use a color-coded system; some will use letters; some will identify problems in the comments surrounding results and recommendations. Regardless, it is important to prioritize your work to ensure that your biggest problems are addressed first, and your time isn’t wasted reviewing routine results.

At Fluid Life, we utilize a system called Sample RankTM, a simple scale running from 0 (nothing to worry about) to 10 (very serious), based on individual results and observed trends. Simply going through your reports and identifying those with Sample RankTM scores between 8 and 10 will allow you to prioritize your review. A study has recently been completed that demonstrates that by focusing only on equipment with Sample RankTM ratings of 8 or higher (typically about 20% of samples received), you can realize up to 90% of the savings achievable through your oil analysis program.

Review the Results and Recommendations comments first. Commercial labs run hundreds or thousands of samples daily. Lab analysts are generally familiar with what looks normal, what doesn’t, and what to do about it. By reading these comments before diving into the results, you will get a good picture of what to look for.

Review the data. All quality labs will identify unusual or unexpected results in some way. Looking for these “flagged” results will reveal what is or is not going on in the equipment. Make sure you understand what each result is trying to tell you. If you don’t know what a particular test is for, don’t be afraid to ask. Remember, you need to read the data in the proper context, with knowledge of the equipment, its operating and maintenance history, and other background information to properly understand what the results are saying.

Fluid Life offers both public and private training on oil analysis interpretation, lubrication fundamentals, oil sampling techniques, and contamination control.

Decision

“A good decision is based on knowledge and not on numbers.” – Plato

Your review of your oil analysis results, informed by your knowledge of the equipment and its history, have revealed that an issue exists, one that needs to be addressed. What you decide to do must be based on your understanding of the issue in question. Not every problem requires the immediate shutdown of the equipment, nor is it possible to immediately shut down equipment based on a bad oil analysis report. On the other hand, some results may indicate that the best path of action is to shut down and inspect or repair critical equipment as soon as a problem is identified.

Take, for example, a report that has identified through an ISO particle count that the oil is dirty, well above what is recommended for the equipment in question. Elemental iron, as reported by ICP spectrometry, has increased as well. Other results are within normal parameters; no water is present, the viscosity is where it should be, the additives appear to be a match for the new oil sample (meaning that the right oil is in the equipment), and other wear elements are not unusually high. Your options might be to:

  1. Immediately shut the equipment down; drain the oil; conduct a full inspection; repair any damaged components; flush the system; refill with the proper lubricant; bring the equipment back on line as quickly as possible.
  2. Plan to shut the equipment down at the next available opportunity, and then proceed as above.
  3. Keep the equipment running; inspect for possible points of ingress of contaminants and repair (if found); filter the oil using a kidney-loop filter cart; re-sample and monitor.
  4. Keep the equipment running; re-sample using Optical Particle Classification (OPC) or Ferrography to better understand the mode of failure and monitor.
  5. Keep the equipment running; re-sample and monitor.

Each decision listed above is correct in the right circumstances, and other correct courses of action may exist. Understanding the criticality of the equipment to your overall process, whether or not the equipment is twinned (installed spare), what effect the actions taken will have on the overall process, the size of the lubricant reservoir, the frequency of your oil analysis, the rate of change in your results, other condition monitoring results, and even the recent maintenance history will all influence your decision. In an ideal world, you would have conducted a Failure Modes, Effects and Criticality Analysis (FMECA) before any indication of impending failure so you would have a good idea as to how to proceed. Even if you haven’t, though, you can base your decision on all of these parameters and more. For the example above, your logic might be as follows:

  • The equipment is critical to the operation of the plant. If it shuts down, the entire process is halted, and start-up will take several hours, if not days. For this reason, the equipment is twinned, and the installed spare can be started up at any time, keeping the plant on line. Go with option 1.
  • The equipment is critical to the operation of the plant, however, the plant has a minor shutdown scheduled in the next two weeks. There are no other indications (vibration, thermography, etc.) that show any impending doom. Go with option 2.
  • The equipment is critical to the operation of the plant, but has not been twinned. If it shuts down, the entire process is halted, and start-up will take several hours, if not days. The reservoir is quite large, measuring hundreds or thousands of litres. Go with option 3.
  • The equipment is important to the operation of the plant, but is not critical. The oil reservoir is quite large, measuring hundreds of litres. Go with options 3, 4, or 5.
  • The equipment is important to the operation of the plant, but it is not critical. The sample was taken by a new employee, and may have been compromised. Go with options 4 or 5.

Response

“Praemonitus, praemunitus” – Latin proverb, loosely translated as ‘Forewarned is forearmed.’

You’ve discovered an anomaly in your oil analysis that points to a problem, and you’ve decided how to approach the problem based on your knowledge of the equipment, the plant, and the operating plan going forward. Regardless of whether the decision was to immediately shut the unit down, schedule a shutdown (or take advantage of a previously-scheduled shutdown), filter or just monitor, ensure that your plan is carried out. Generate an appropriate work order (if work is to be done) and log it once it is complete. If filtering, ensure a sample is taken to confirm that it has worked. If the plan is to re-sample, either with the same analysis or additional testing, ensure that it happens in a timely fashion and that you follow up by reviewing the analysis as soon as possible. With the knowledge gained from viewing your data in the context of your familiarity with the equipment and its place in your process, you will be able to take action based on solid decisions.

Fluid Life’s CARE program is a sustainable solution that puts you back in control of your oil analysis program. An established and disciplined process, cutting edge software tools and expertise of our Reliability Services team deliver excellent financial returns. With the CARE program, you can expect:

  • Oil analysis interpretation delivered by Fluid Life experts
  • Collaborative decision-support for corrective actions
  • Tracking work order creation and completion
  • Closing the loop on problem conditions
  • Monthly reporting of program and financial benefits

Take control of your oil analysis program and use it to its full potential with Fluid Life and CARE.