Improving Predictive Maintenance Through Wear Debris Analysis

Treat your oil analysis laboratory as a real partner in the prevention of unexpected equipment shutdowns. Work closely with it to harness the power of the various predictive maintenance techniques that will keep your operations up and running profitably.

The goal of every predictive maintenance tool is to provide an early-warning system before equipment failure—and the earlier a warning comes, the better (see Fig. 1). In the last issue of this magazine, we discussed the use of atomic emission spectroscopy (AES) as one especially powerful early-warning technique (pgs. 30-35, LUBRICATION & FLUID POWER, September-October 2006). However, because of particle size limitations of 10μ need to be utilized. This article discusses a number of these techniques.

Analytical ferrography
The most powerful diagnostic tool in oil analysis is analytical ferrography—the only test that can justify shutting down a piece of equipment. It consists of detailed microscopic examination of a slide to determine:

• Particle size and relative concentration
• Metallurgy both ferrous and nonferrous
• Wear mechanism indicating root cause
• Component source of wear
• Identification of contaminants

An analytical ferrography slide is created by passing an oil sample along a glass slide over a strong magnetic field as illustrated in Fig. 2.

The slide captures both ferrous and nonferrous particles along with contaminants and upon heating can reveal additional information on the particle type such type of alloy.

Table I categorizes some of the different particle types along with their sources.

Analytical ferrography also can be used for root cause analysis in evaluating equipment failures. Root causes that can be determined through this technique include:

• Overloading
• Misalignment
• Plain bearing cavitation
• Contamination

When to use analytical ferrography
The following checklist can help you determine when this technique is appropriate:

1. Equipment is exhibiting unusual characteristics, such as overheating, high vibration readings, unusual sounds and high ΔP readings across a filter.
2. Equipment is extremely critical, which necessitates making a moderate investment of $35-$100 per sample to prevent unexpected equipment failures.
3. High failure rates experienced on startup of new or repaired equipment due to infant mortality caused by material defects, design problems or improper assembly. (Analytical ferrography, along with routine oil analysis tests, should be run quickly after startup.)
4. A new oil analysis program is established. (Analytical ferrography should be run on critical equipment to establish baseline data.)
5. Routine oil analysis indicates an unusual condition that necessitates further evaluation. The following tests will be examined as trigger points for further analytical ferrography evaluation:

   • 

     Direct Read Ferrography

   • Particle Quantifier
   • Particle Counts
   • Micropatch
   • Laser Net Fines
6. Critical equipment should have an analytical ferrography done once a year as a minimum.

Direct read ferrography
The direct read ferrography technique measures ferrous particles both large DRL >5μ and small DRS <5μ (see Fig. 2).

In this technique, a sample is flowed through an inclined glass tube across a strong magnetic field with two small openings where magnetic particles are accumulated and measured by light blockage. The first opening captures large particles >5μ and the second opening captures particles <5μ. There is some mixture of particles, so that the first opening captures some of the smaller particles. Contaminants and nonferrous metals also can be captured in the openings. The measurement of particles is expressed as Wear Particle Concentration (WPC), which is a unitless number.

Because direct read ferrography is relatively inexpensive and quick, it is a good screening tool for analytical ferrography. It also is excellent for trending ferrous wear particles.When the ratio of DRL to DRS increases, it is an indication of a greater generation of large particles, which is a precursor to potential catastrophic wear.

Direct read ferrography is commonly used in routine oil analysis for systems with large amounts of ferrous metals, such as gear boxes and screw/reciprocating compressors. Before analytical ferrography is performed in these systems, certain trigger points must be met. Table II expresses normal total WPC values by equipment type before going to analytical ferrography. Each oil analysis laboratory has its own trigger limits for WPC.

It is important to trend the DRL and DRS ratio. Typically, as the ratio increases, wear severity increases. A ratio in excess of 5:1 calls for further investigation with analytical ferrography. Some laboratories, though, are only interested in the overall DRL value.

One major petrochemical plant, based on the knowledge history of its equipment, has established the following guidelines for DRL WPC trigger points for analytical ferrography in Table III.

You will want to work with your oil analysis laboratory to understand and to set guidelines for further analytical ferrography analysis based on your DR readings.

Particle quantifier (PQ)
PQ measures ferrous density by exposing samples to a magnetic field. Presence of any ferrous metals causes distortion of the field which is then assigned a dimensionless value called the PQ index.

This test is very quick and inexpensive, but does not differentiate between large and small particles. Consequently, the same PQ value can indicate a large number of small particles, a small number of large particles or a combination of the two. It is a very trendable and can be used to justify analytical ferrography.

One oil analysis laboratory has used the PQ technique along with emission spectroscopy to trend the growth of large particles. For example, if the PQ index is increasing without a corresponding large increase in emission spectroscopy values, which measure only small particles, it is an indication that large wear particles are being generated.

Particle count
There are two types of particle counters: optical and pore blockage.

Optical counters are most commonly used for fluids where laser light can pass through. Dark fluids, such as gear and engine oils, have to use pore blockage to obtain particle counts.

Optical counters can either measure particles by light blockage or light scatter. They can not be used with fluids that contain >300 ppm water. Light blockage or scatter is correlated with fine dust calibration through microscopic particle counting to arrive at size and number of particles. This is a quantitative measure and does not identify particle types, whether they be wear or contaminants. Oil analysis laboratories report actual number of particles per milliliters or per 100 ml by size range. Typical size range counts are shown in Table IV.

In order to simplify the cleanliness of a fluid, the ISO 4406 cleanliness code is used. It is a three-number designation such as 16/14/11, which classifies particles >4μ, >6μ and >14μ with use of the information contained in Table V.

The ISO 16 means that the measured particles >4μ fall within the range of 320 to 640 particles per milliter. The ISO 14 represents particles >6μ that fall in the range of 80 to 160. The ISO 11 are particles >14μ, which fall within the range of 10 to 20.

The lower the ISO number, the cleaner the fluid. Every one-number increase in the ISO number signifies a doubling of the particles.A trigger point for analytical ferrography is when the ISO number increases by two. It is important to note particles can be wear, contaminants or a combination of both. Larger particles are more likely caused by wear; therefore trending particles >14μ may indicate the wear severity resulting in further investigation with analytical ferrography.

Pore blockage is used for dark fluids and those contaminated with water. It is a good trending tool and is expressed as a three-number code >4μ, >6μ and>14μ. The test is run by passing a fluid through a 10μ- or 15μ-micron filter and measuring flow decay and extrapolating to express the two ISO numbers of 5 and 15.A two-number increase from one period to another may require analytical ferrography. In most cases, dark fluids, such as gear oils, don’t normally have particle counts run on them. Instead, direct read ferrography typically is utilized to capture significant increases in wear of dark fluids.

Micropatch
In the micropatch technique, a fluid sample is mixed with solvent and then vacuumed through a 0.8μ absolute filter patch. Microscopic analysis is performed on the patch where particles are sized and identified. A good test to identify nonferrous metals and contaminants, it is used as an indication of significant wear that may require further analytical ferrography analysis.

Micropatch testing can be run quickly. It is most appropriate for turbines, compressors, water glycols and natural gas engines.

Laser net fines
Lockheed Martin, in cooperation with the Naval Research Laboratory, developed the laser net fines test in 1998, but the test only recently has begun to be used by commercial oil

analysis laboratories. It measures particle counts for particles >4μ, but also classifies shapes for particles >20μ up to 100μ in the following forms: cutting, severe sliding, fatigue, nonmetallic, fibers and water droplets.

The laser net fines test is run by having a pulsed laser diode pass through the sample with a camera recording the images where the particles are classified by shape and size. Its main advantage over optical particle counters is the ability to run both dark and water-contaminated samples. This tool provides very good predictive maintenance information with its ability to trend the morphology of particles in both size and shape, and can be used as a precursor to analytical ferrography, which provides more in-depth evaluation.

A case history
Analytical ferrography is not part of routine oil analysis, but when applied properly, it can prevent catastrophic failures—and lead to significant savings.

A large petrochemical facility has a centrifugal compressor driven by a 700 hp motor producing a large volume gas chemical. Recently, particle count numbers began to increase, eventually reaching the trigger point for analytical ferrography.When this test was performed, it indicated Babbitt bearing wear in the electric motor. The sample was rerun to confirm the findings and again indicated significant wear. Based on the data, operations elected to shut down and inspect the motor. Inspection revealed that the bearing was starting to become wiped and was not replaced. Instead, it was scraped and quickly put back in service, saving $300,000—not including production loss costs, which can be significant.At the next scheduled outage, the bearing was replaced as part of normal maintenance, at minimal cost and with no production losses. In short, thousands of dollars were saved because the motor was not damaged and the shutdown was planned and, therefore, completed quickly.

Many other actual cases can be cited where the use of analytical ferrography has helped prevent catastrophic failure, thus saving millions of dollars.

Conclusion
World-class companies, in order to remain competitive in a global economy, need to practice Reliability Centered Maintenance (RCM). Predictive maintenance is central to any successful RCM program, while oil analysis is central to any successful predictive maintenance program.

Keep in mind that although atomic emission spectroscopy is an important predictive tool, it is limited by size of the particles detected. Other techniques are required to measure large particle wear debris. Central to wear debris analysis is analytical ferrography, the only common oil analysis tool that can justify equipment shutdown. Because it is time-consuming and relatively expensive ($35-$100), analytical ferrography is not commonly recommended for all samples. Other tools, such as direct read ferrography and particle counts, need to be utilized as screening before running analytical ferrography.

Typically, about 10-15% of samples require analytical ferrography—and 10% of these samples reveal a critical condition. Only about 1- 1.5% of total samples evaluated reveal a critical condition, but the savings can be large if unexpected equipment failure is prevented. The type of equipment dictates the numbers of samples that will reveal a critical condition. Normally, gear boxes, many of which are overloaded, have a larger percentage of critical samples.

Many companies that have oil analysis programs are not utilizing analytical ferrography. It is important to work with your oil analysis laboratory to apply this tool when needed. Some laboratories will have a normal up-charge, so if analytical ferrography is required, it will be run at no extra cost. This is like having an insurance policy. It definitely should be considered as part of your oil analysis program. LMT

Contributing editor Ray Thibault is based in Cypress (Houston), TX. An STLE-Certified Lubrication Specialist and Oil Monitoring Analyst, he conducts extensive training in a number of industries. Phone: (281) 257-1526; e-mail: rlthibault@msn.com