Performing oil analysis tests to determine the presence of varnish in a system is challenging for a few reasons. The sample of oil that is obtained for analysis may not be indicative of the condition of the lubricating system. If the lubricant in a heavily varnished system is changed without performing a flush to remove the deposits, the new oil may initially indicate a low varnish potential even though there are deposits throughout the system.
Often in peaking or cycling gas turbines, the varnish potential will increase during times of the year of high demand and decrease when the unit is offline for several weeks. This is due to the degradation products coming out of solution in the oil and forming deposits through-out the lubricant system.
The lubricant degradation products (soft contaminants) responsible for causing varnish deposits can be challenging to measure. They are very small, often estimated at less than 0.1mm in size, although they can easily agglomerate together. Soft contaminants are soluble and easily transition in and out of solution depending upon temperature and pressure. Current testing methodologies only measure soft contaminants when they are in suspension, so sample handing prior to testing impacts results. While the soft contaminants are in solution, they don’t typically impact the bulk chemistry of the fluid making it very difficult to measure.
Most routine analytical tests are not successful in measuring the onset of varnish or the varnish potential of a lubricant. Fortunately, there have been several testing advancements in recent years that provide much more effective identification of underlying issues.
The first step in determining turbine oil's potential to produce varnish is to understand the overall health and remaining vitality. This is influenced by a variety of things, but the fluids antioxidant system health and concentration is paramount. The best way of determining the antioxidants remaining health, and consequently the fluids remaining useful life, is by directing measuring them, using a technique called Linear Sweep Voltammetry, or RULER. This measurement identifies the antioxidant type and measures the quantity of material present. The results are typically reported as a percentage of new oil and separated by antioxidant family.
There are other physical and chemical property measurements that may be useful in assessing the overall health of the fluid, including the fluid’s ability to handle contamination such as air and water. When additive health is lost, and/or the fluid has been cross-contaminated with incompatible chemistries (from other lubricants, water treatment chemicals, or process chemicals) one of the first performance criteria that fails is demulsibility (ASTM D892). Foam (ASTM D892) and air release (ASTM D3427) characteristics also change for similar reasons. When compared to performance of the new fluid, each of these can indicate that some type of damaging change has occurred in the fluid. Since chemical incompatibility can be a cause of degradation products and varnish, these are important characteristics to know about the lubricant.
Profiling the amount of contamination in the fluid by particle counting, water and spectrographic analysis is also of value.
Once the overall health of the fluid is determined, measuring the amount of soft contaminants is the next step in determining a fluid’s propensity to produce deposits. The Membrane Patch Colorimetry test is the most widely used test for this purpose. The MPC method is quite simple. Fifty milliliters of oil are mixed with fifty milliliters of solvent (usually petroleum ether) and filtered through a 0.45mm nitrocellulose patch. The color of the patch is determined by spectrophotometer and the results are reported in CIE LAB DE Value, representing the total amount of color on the patch. In general the darker the color stain on the patch, the higher the fluid’s varnish potential.
This technique requires that the sample remain idle for up to a week at room temperature prior to analyzing to ensure the soft contaminants have come out of solution. There is generally good correlation between MPC values and actual field varnish issues, but interpretation is not always straight forward. For example, the controls on a GE 7FA are much more sensitive than on a GE 7EA. The same MPC reading for both systems may indicate a more severe condition in the 7FA system due to these sensitivies. Duty cycle also plays a large role in MPC interpretation.
Since lube varnish can play a large role in a plant's reliability, the most important take away in this post is to make sure that you're testing for this. In some cases, gravimetric analysis and ultracentrifuge tests can augment the MPC patches, providing a more complete contamination profile of the fluid.