New Technology Changes the Landscape for Varnish Removal

For the past 30 years, we’ve tried to maintain cleaner lubricating environments by reducing the pore size in the system filters. For most of us, the prescription was simple. If the particle count and the ISO Code (ISO 4406-1999) are too high, this was interpreted as the filter pore size was too large. This, however, may not be the correct response with many modern lubricants.

New Filtration Technology for a New Contamination Control Challenge
The main full-flow filters must be viewed in terms of their original purpose of controlling particles from damaging frictional surfaces and plugging oil ways and clearances.

Kidney-loop filtration is increasingly common in controlling the fine material buildup and providing particle-count control.

In the past, we thought only of particle count and followed by finding the ingression source. Today, there is a second route for ingress: varnish and oxidation product buildup in the oil.

New Technology
A new electro-technology has been developed to overcome varnish buildup by creating a filter with fibers treated to actively adsorb varnish and pre-varnish material. This principle is enhanced by applying a low static voltage to half the filter material in a positively charged fashion and half the material in a negatively charged fashion. As a result, all charged species and all polar species are brought into contact with the fibers in the filter and removed.

Extensive laboratory and beta site testing have been conducted at the factory and with a modern combined-cycle generating facility. The results on a macro scale are noteworthy. The large-frame gas turbine ran for five years with no technology other than what had been provided by the manufacturer. The lubricant was a modern Group II formulation from a major supplier, and the varnish potential had been monitored for some time.

Varnish potential testing was developed some years ago to rank and attempt to rate the probability of clearing restrictive oxidative deposits within the machine oil system from an oil analysis test. This is a daunting challenge. Trying to correlate the probability of a deposit within a system, based on the level of less than 0.01 percent contamination, has thus far been met by the lab that originated this test, Analysts Inc.

Beta Site Case Study
With the onset of the beta site trial for this new technology, the varnish potential was reduced from 60 to 5 during the first days of operation (Figure 1). The entire system was then cleaned during the next 25 days.

Figure 1. Varnish and Gravimetric Solids

During this period, the pentane insolubles (ASTM D4055 0.045 micron) decreased from 48 to less than 15 mg/l. This reduction in gravimetric solids is confirmed by the laboratory analysis performed with the particle-counting technology which can count and size particles smaller than five nanometers. Figure 2 shows the submicron contamination in the oil before the start and at the end of testing. The two graphs show the complete eradication of material above 0.1 micron during the period. As previously stated, the authors believe the remaining material is all color bodies.

Figure 2. Particle Count Below One Micron Diameter

 

Figure 3. Particle Size Distribution After Oil Purification

New particle-sizing and counting technologies make this small particle removal performance easier to track and trend. This testing in conjunction with the varnish potential test makes a powerful affirmative indicator pair of the presence of varnish material. Figure 2 from the laboratory testing clearly demonstrates why the 0.45 micron or 0.8 micron patch used for the varnish potential test was completely white at the end of the study. The authors believe the field test results will correlate well to this data.

Testing the fluid of previously varnished oil revealed a striking change.

They believe that the remaining material is primarily color bodies and could possibly be removed in the coming weeks.

Filtration System Design
The separation concept is that oil flows through layered media top to bottom, passing through capillary-sized channels. One filter medium is charged positively, and the other is charged negatively. Oil is fed into a controlled path through capillary-sized paths within treated wound depth media. Each controlled flow element contains several thousand square feet of surface area.

Positive and negative electrical charges are placed on the bulk media. A double layer forms at the boundary between the flowing fluid and the surface of the media. This double layer is responsible for separating the contaminants from the fluid stream. The nature of the media surface creates the holding function for the particle retention function.

A double layer forms at the boundary of each capillary channel with a charge opposite to the charge applied to the bulk of the medium. This double layer strips material with the same polarity charge as the polarity of the bulk medium. As the conductivity of the fluid flowing through the channel increases, the width of the double layer expands, making the effect more pronounced.

The electro effect forces the contaminants into contact with the medium, and the adsorptive nature of the surface retains it. The adsorptive nature of the treated medium causes a preferential retention of amorphous oxidative varnish material.

Conclusion
It appears that this new technology demonstrates a capability to remove and retain large amounts of polar and solid material in an aggressive fashion. The reservoir previously displayed a wide bathtub ring at the fluid level, but this has been almost entirely removed in 30 days. The foaming condition has also been altered to a level between normal and ideal.

The authors anticipate this new technology will be able to control varnish, gravimetric solids and particle count to an improved level. If the theory about color body removal is correct, the fluid will return to original transparency within a short period of time. Further testing and reports are anticipated. POA

Acknowledgments

The authors wish to thank their colleagues for their contributions to this article: David Wooton of Wooton Consulting, Brian Thompson of Analysts Inc., Akira Sazaki and Dan McCormick.

About the Authors

For further information, contact Gerald Munson at gmunson@filmaxinc.com or Dan McCormick, dmccormick@filmaxinc.com.