The growing use of electrostatic oil cleaners in power-generating stations and paper mills to control varnish and contaminants has prompted interest among lubricant and additive manufacturers, as well as end users about the performance of lubricants in conjunction with these systems.

Figure 1. IGV filter from GE 7FA gas turbine after three weeks in intermittent service before (top) and after (bottom) solvent wash.

Introduction
Turbine oil contamination is more predominant in hydroelectric and steam turbines than in gas turbines because the oil in these units is rarely drained, inadequately filtered (until recently) and often exposed to moisture, bacterial growth, particles, and sludge and varnish buildup within the system. Water represents a real risk to the oil and the equipment and should be aggressively controlled.2 Fine filtration and water separation units, in the form of kidney-loop systems, have offered improvement.

Gas turbines present unique challenges. The turbine oil is subjected to much higher operating temperatures than in steam or hydroelectric turbines due to the transfer of heat of combustion to the oil through the bearings. It is common for the turbine oil to be exposed to intermittent bearing temperatures as high as 500°F (260°C), with the bulk oil temperature in the reservoir ranging from 150°F to 200°F (66°C to 93°C) during operation.3

In turbine designs where a single common oil reservoir is used for both bearing and control oil, critical control actuators are susceptible to varnish formation and valve sticking due to a low and infrequent oil flow rate combined with a drop in oil temperature in the control lines. Varnish deposits inevitably build up over time on critical surfaces, such as an inlet guide vane (IGV) valve (Figure 1), often resulting in costly turbine trips.

Lubricant degradation occurs at high temperature from oxidative and chemical reactions. Oil decomposition by-products form polar compounds which adsorb on metallic surfaces to form varnish. The mechanism of deposit formation on servo valve filters in gas turbines is not clearly understood. Recent papers published on this topic4 explain the general mechanism of varnish formation as follows: "Lubricant varnish is defined as a thin insoluble film that develops throughout the intervals of a machine's lubrication system over time. It is considered a contaminant, composed predominantly of lubricant degradation by-products and depleted additive molecules. These by-products are submicron in size, they are insoluble in the oil, and they are highly reactive and easily polymerized.

A major power generation company conducted an internal inquiry on servo valve varnishing and concluded that varnish formation depends on the following variables:

  • Low or stagnant oil flow

  • Static spark discharge creating excessively high localized hot spots in the oil

  • Contaminants in tight tolerances

  • Cyclic temperatures

  • Inadequate oil cleanliness

  • Submicron polymeric insolubles formed from oxidative degradation of the turbine oil

Installing an on-site electrostatic oil cleaner in conjunction with a 15-micron oil filter at Beta 200 eliminated varnish formation and power trips.

ConocoPhillips conducted an internal study on the performance of Group II turbine oils in gas turbines in continuous operation, where the oil temperature is controlled and held constant per manufacturer recommendation, the oil is adequately filtered, and the oil flow rate is high. Under these conditions, servo valve varnishing did not occur after more than 25,000 hours of operation.

Thus, deposits, such as fine polar insolubles which lead to varnishing, can be removed by the use of advanced separation technologies such as electrostatic technology, balanced charge agglomeration and ion-exchange resins. But the question remains: Are lubricant additives removed or damaged by these technologies?

Test Protocol with Electrostatic Oil Cleaner
Four Group II-based industrial lubricants were evaluated in a commercial electrostatic oil cleaner:

  • New ISO 32 turbine oil with 22,500 hours TOST life as measured by ASTM D943

  • New ISO 32 ashless hydraulic oil with 15,000 hours TOST life by ASTM D943

  • Used ISO 32 turbine oil with >25,000 fired hours in a GE 7FA gas turbine

  • Used ISO 46 compressor oil with 3,900 hours in service

Physical and chemical properties of the various oils were evaluated before and after electrostatic cleaning, and some of them were evaluated by the quantitative spectrophotometric analysis (QSA) test method4, to determine if additives were removed during the electrostatic cleaning and to document their performance.

The small electrostatic oil cleaner used in this test protocol is a modified system that incorporates a built-in reservoir for the oil being tested, as shown in Figure 2. Based on the design flow rate, the estimated oil residence time for the total oil volume in the system was approximately 10 minutes. In a 24-hour period, the oil passed through the cleaner 144 times, which made the filtration conditions more severe for the oil additives than those observed in actual use in the field.

Tech_Conoco_Fig2.jpg

Figure 2. Electrostatic Oil Cleaner Test Assembly

Test Unit Setup
Prior to the start of a test, the electrostatic oil cleaner unit was completely cleaned with light base oil. The reservoir was then filled with test oil, and the unit was turned on to begin the test. The voltage applied to the cleaner's collector filter was between 11.8 and 11.9 kilovolts. Each test oil was run in the electrostatic oil cleaner for a predetermined number of hours. Oil samples were taken after each run and analyzed.

Change in Kinematic Viscosity at 40°C, cSt
Samples taken after one hour and 166 hours indicated a slight reduction in the oil viscosity for each of the test oils. This reduction was most likely attributed to dilution with the residual light neutral base oil used to flush the unit. All samples remained within their ranges of viscosity grade.

Change in Acid Number and Foaming Characteristics
The change in acid number (AN) of the four test oils ranged from no change (for the new turbine oil and used compressor oil) to a significant decrease (for the new ashless hydraulic oil and the used turbine oil). The decrease in the AN of the new ashless hydraulic oil suggests a partial removal of additives, while the AN decrease of the used turbine oil could be due to a combination of a reduction in the level of oxidation by-products and partial reduction of rust inhibitor additive by the electrostatic oil cleaner. Nevertheless, the tested oils still provided adequate rust and corrosion protection based on ASTM D665 and D130 results, as shown in Figure 3.

The effectiveness of the defoamant additive formulated in these oils was measured after ASTM D892 Foam Test and found less than new oil values. This property was not checked prior to running the filter test; therefore, definite conclusions cannot be made that the reduction was due to processing through the electrostatic oil cleaner. Additional research on the effect of this technology on defoamant additives is needed to confirm that this additive type is not adversely affected by electrostatic oil cleaners.

Figure 3. TAN Values During Testing

Change in Pour Point
No change was observed in the pour point of the test oils after passing through the electrostatic oil cleaner, indicating the pour point depressant was not removed or affected in this process.

Change in Oxidative Stability and Oil Cleanliness
A slight reduction in oxidative stability by the RPVOT method (ASTM D2272) was observed in some oils after passing through the electrostatic oil cleaner. The new turbine oil experienced a reduction of 8.66 percent after 27 hours of exposure to the electrostatic cleaning (equivalent to 162 turns on the oil volume) and the new ashless hydraulic oil experienced a reduction of 10.86 percent after 166 hours (equivalent to 996 turns), as shown in Figure 4.

This difference is within test precision and repeatability (12 percent) and is therefore not conclusive. While the trend in the data for the New ISO 32 Group II Turbine Oil and New ISO 32 Group II Ashless Hydraulic Oil suggests a slight (less than 10 percent) reduction in oxidative stability, the absence of a decline in the Used ISO 32 Group II Turbine Oil indicates further testing is needed to resolve the contrasting results.

While the used turbine oil showed no change in oxidative stability after 166 hours, the used compressor oil showed about 35 percent reduction in oxidative stability, indicating that the electrostatic oil cleaner removed a significant portion of oxidized material with antioxidants attached to them. This compressor oil was obtained from a Gardner-Denver 100-horsepower rotary screw compressor after 3,900 hours in service. The compressor oil was near the end of its service life and was expected to contain significant varnish deposits.

The significant improvement in the oil cleaning effectiveness was confirmed by the reduction in ISO particle count results, as shown in Table 1. Note the different number of hours that the four oils were subjected to the electrostatic cleaning process.

The new turbine oil was on the unit for only 27 hours because the test ended on a Friday before a weekend. That is why the particle count did not drop as much. The authors were looking for a trend.

Figure 4. Oxidative Stability Data (RPVOT)

Summary and Recommendations
In today's demanding environment for power generation, gas turbines put unique stresses on turbine oils. As a result, there can be a tendency to build up impurities in the form of varnish, sludge and oil degradation by-products. If these contaminants are not removed from the oil, they can plug filters and plate out onto system components more so under cyclic temperature conditions and during low-to-stagnant oil flow rate, especially in tight tolerance areas, such as servo valve filters. This results in erratic control system operation.

Table 1. Particle Count Test, ISO 4406:1999

If the hydraulic system is maintained clean by using an electrostatic oil cleaner that has no negative impact on foaming and additive removal from the oil, which impacts the oil's oxidative stability, and adhering to the OEM recommended filter, the hydraulic system will remain clean, filter plugging is eliminated and the turbine control system will operate trouble-free.

Table 2. Oil Property Limits

Adhering to the following preventive maintenance program and following OEM recommendations can eliminate servo valve varnishing problems in gas and steam turbines. Further details on in-service turbine oil monitoring can be found in the ASTM D4378-03 standard:

  1. Conduct monthly oil analysis on the turbine oil to determine its condition. In addition to the routine tests, conduct colorimetric filter patch test or ultra centrifuge test to determine potential for varnish formation. Conduct particle count test by ISO 4406:1999 method to determine oil cleanliness. Use the RPVOT test as a guide only to determine the turbine oil's predicted remaining useful life. The limits in Table 2 should be adhered to.

  2. Use filters in the 5- to 10-micron range with high Beta ratios.

  3. Do not use filters <10 microns in conjunction with electrostatic oil cleaners.

  4. Follow OEM recommendations.

  5. Periodically inspect the reservoir and filters for static spark discharge.

Editor's Note:
The data presented in this article does not present a sufficiently consistent pattern to establish definitively that oil additives are negatively impacted by the application of this technology, but it does suggest the need for further testing.
Varnish is the result of heat and contaminants within the oil system. Electrostatic cleaning helps to minimize the effects that varnish has on the system, but a root cause analysis should be performed to potentially eliminate the fundamental cause(s). Root causes include additive dropout, bulk oil oxidation, microdieseling and electrostatic discharge.

Acknowledgements

Buddy Atherton, Technical Sales Manager, UAS / Kleentek
Andrew Jeng, Chief Scientist, ConocoPhillips
Ernest Pettit, Associate Research Technician, ConocoPhillips

References

1. ASTM D4378-03. Standard Practice for In-service Monitoring of Mineral Turbine Oils for Steam and Gas Turbines, 2003.
2. Fitch. Oil Analysis for Maintenance Professionals. Tulsa, Okla, Noria Corp., 1998.
3. GE Power Systems. GEK 107395a, May 2001.
4. Thompson and Livingstone. "Improve Your Predictive Maintenance Program with a Varnish Potential Test." Noria Lubrication Excellence and Reliability World 2005.
5. Sasaki and Uchiyama. "A New Technology for Oil Management: Electrostatic Oil Cleaner." UAS / Kleentek Industrial Co. Ltd., Tokyo, Japan.
6. Winslow, Naman and Jenneman. "Bacterial Contamination of Turbine Lube Oil Systems." Noria Lubrication Excellence 2004.