Gas-turbine valve sticking . . . the plot thickens
By Greg Livingstone and Jon Prescott, EPT Inc, and Dave Wooton, Wooton Consulting
Lubricant varnish is the primary cause of valve sticking/seizing in the hydraulic circuits of gas turbines (GTs). Because valve sticking often leads to starting problems and turbine trips, lubricant varnish has been one of the hottest topics at recent user group meetings. It also has been a frequent subject of technical letters by OEM engineers, conference presentations by lubricant professionals, webinars by oil suppliers, and so on.
One of the most important causes of lubricant varnish—particularly impacting cycling and peaking GT plants—is auto-degradation. Unfortunately, the traditional varnishpotential tests will not tell you if auto-degradation is occurring, though such tests remain essential to your overall oil-condition monitoring program. Similarly, electrostatic separators, seen by many as a onestep “cure-all,” will not eliminate the problem. Note that the term “electrostatic separators” as used here includes electrostatic oil cleaning, balanced-charge agglomeration, electrostatic filtration, etc.
To fully understand if your lubricant is undergoing auto-degradation, you need to assess the antioxidant health of the fluid and examine it for specific types of degradation byproducts. If auto-degradation is identified, you’ll need a holistic approach that includes monitoring of initial oil quality, analysis and additive replenishment for in-service oil, and the installation of appropriate oil-cleaning technology that removes existing varnish and slows the auto-degradation process.
Plants achieving a complete solution to lubricant varnish will achieve higher unit reliabilities. Those attacking only part of the problem will continue to be frustrated by the brown, sticky goo.
A common tale
Here’s a story that’s becoming alltoo- common in the combined-cycle/ cogen sector of the US electric-power industry: A peaking plant installed an electrostat ic separator and assumed that its varnish problems would be eliminated. To monitor the problem, operators tracked the performance of their new oil-cleaning equipment by testing the varnishpotential rating using Quantitative Spectrophotometric Analysis (QSASM). The initial results were excellent: Varnish-potential ratings decreased, and the valve sticking problems went away.
A few months later, however, the varnish-potential ratings started to increase. Soon, the test results showed a consistent trend upwards— with the QSA rising close to the same level as before the installation of the electrostatic separators. And then, one cold morning when the unit was fired, the inlet-guide-vane (IGV) valve stuck, resulting in a fail-tostart event for the plant. Frustration ensued. Fingers were pointed. This plant had just experienced the effects of auto-degradation (Fig 1).
Auto-degradation is defined here as the creation of soft contaminants in a static body of oil. Note: A shutdown lube-oil system for a cycling or peaking GT is one such “static body.” Soft contaminants are the lubricant degradation byproducts most responsible for varnish formation, and often are more troublesome to remove than the hard particulates on which lubricant authorities traditionally focused.
As was discovered by the example plant above, electrostatic separators do not solve all varnish problems. By themselves, neither will the alternative filtration technologies. For example, one alternative that has been spotlighted of late is cellulose depth filters. Because they have a polarity, soft contaminants do indeed have an attraction to cellulose, which has its own polarity. Hence these filters can absorb soft contaminants—just as the charcoal water filters used in the home successfully adsorb chemical impurities from tap water.
The problem for GT users is that soft contaminants are extremely small—about 0.08 micron in diameter. Before they develop a polarity so that they can be absorbed by a cellulose depth filter, they must agglomerate by a factor of approximately 10—to particles roughly 0.8 micron. Unfortunately, the turbulent conditions in an operating GT lube-oil system do not allow such agglomeration to occur. Furthermore, agglomeration is a reversible process, so any agglomeration that occurs while the gas turbine is shut down is reversed soon as the turbine’s lube-oil pump is restarted.
Another alternative technology— mechanical filtration—cannot remove the tiny soft contaminants until they agglomerate by a factor of approximately 20—to particles about 2 microns in diameter (Fig 2).
Net result is that the majority of soft contaminants remain in turbine oil when the lubrication system is equipped with only cellulose depth filters or mechanical filtration.
Causes of autodegradation
Several phenomena are known to cause auto-degradation in lubricants. Focus here is on the four most important to GT users: precipitation, auto-oxidation, condensation, and agglomeration.
Precipitation occurs through a temperature-related solubility effect. If oil is saturated with soluble soft contaminants at a relatively high operating temperature, insolubles will precipitate out of solution as that oil cools during shutdown. Imagine scooping many spoonfuls of sugar into a hot cup of coffee, until the coffee won’t hold anymore. At this point, the solution is “saturated” with sugar. When that cup of coffee cools, it becomes supersaturated with sugar, forcing the now insoluble amounts to precipitate out, leaving a sweet syrup in the bottom of your cup.
The oil in the pressure line to the IGVs is like that cup of coffee, with insoluble soft contaminants acting like the sugar as the oil cools down. When the turbine is firing and the IGV valve is constantly adjusted, the temperature of the oil is relatively high. However, when the unit is off-line and on turning gear, the oil cools down, particularly in the IGV line where flow is virtually eliminated. In many plants, the cooling effect is exacerbated because the IGV line is exposed to outside ambient conditions.
Auto-oxidation is a second cause of auto-degradation. The most important aspect of GT oil formulation is the level and type of antioxidants— the sacrificial additives that disrupt the auto-oxidation process, and assist in protecting the oil. As antioxidants are depleted during service life, the oil becomes more susceptible to auto-degradation—hence to the formation of varnish.
In GT hydraulic systems, auto-oxidation typically occurs when there is an over-abundance of free radicals in the oil. Source of the free radicals? Lubricant authorities believe one of the major sources is the spark discharge from mechanical filters. Evidence of spark discharges have been indentified the main lube oil, hydraulic, and last-chance filters of virtually all powerplant designs.
When sparks occur, a high-energy plasma around the spark generates severe oil degradation and free radicals to continue the oil’s destruction over time, while the spark itself impinges on the metal surface causing melting and pitting. The craters that ensue also kick out metal particles that enter the oil as hard contaminants (Fig 3).
Condensation (technically referred to as Aldol or Claisen condensation) occurs when several low-molecularweight soft contaminants combine to form a higher-molecular-weight, long-chain molecule. This can result in the transition of soft contaminants from a soluble (harmless) state to an insoluble (harmful) state.
Agglomeration is a very similar process to condensation, except that it occurs with soft contaminants that are already insoluble. The typical size of insoluble soft contaminants is well under 1 micron; however, through agglomeration they can grow to 1 micron or larger. As mentioned earlier, the rate of agglomeration is inversely related to the level of turbulence in the oil.
Auto-degradation in my oil?
Brian Thompson, laboratory manager for Analysts Inc’s mid-continent operation in Louisville, has investigated auto-degradation at numerous sites. In a recent study conducted on dozens of turbine oil systems, Thompson found that over 80% of the GTs he studied showed evidence of autodegradation in their lube-oil systems. Some of these oils had been in service less than two years. He also found differentiation in the severity of auto-degradation based on the unit manufacturer, model, duty cycle, and brand of oil. Interestingly, not a single steam-turbine-oil system or control-oil system showed evidence of auto-degradation, even though some of these systems had an elevated varnish potential.
GTs with integrated lube oil and hydraulic circuits—such as the GE F7A—are most prone to auto-degradation. Turbines designed with separate hydraulic oil reservoirs—such as the MHI 501G—are not immune, but the problems tend to be less severe, or simply happen later in the life cycle of the oil. Duty cycle of the plant also plays a significant role in auto-degradation, with cycling and peaking gas turbines being more vulnerable than base-load plants.
Required tests. When a sample of oil is pulled from an oil reservoir and allowed to sit for 7 2 to 96 hours, the measurable amount of soft contaminants increases. The observation is a change in varnish potential rating (VPR) by the QSA test. In a lube system that is experiencing severe autodegradation with electrostatic separators installed, it is possible to see a VPR of 7 onsite, and 95 in the lab several days later. (The VPR scale goes from 1 to 100.) As often happens in the real world, onsite testing is not always convenient, and in some cases may even be impossible. Specialized laboratory tests to indicate the potential for auto-degradation include:
- RULER, which measures the amount of phenolic and amine antioxidants. Note that this test can be performed onsite.
- FTIR, which can detect lubricant degradation byproducts and some additive concentrations.
- QSA, which measures the lubricant’s varnish potential.
- RPVOT (Rotation Pressure Vessel Oxidation Test, ASTM D2272), which measures the oil’s resistance to oxidation but does not correlate to auto-degradation.
Clean-up/mitigation
What options does a gas-turbine user have to eliminate or at least minimize auto-degradation? The most important is a comprehensive, carefully structured oil-condition program that: (1) removes the soft contaminants (cleans up the oil); (2) minimizes oil degradation (finds the root cause of the problem and mitigates it); and (3) strengthens the quality of the oil (replenishes the oil additives)
Cleaning up the oil begins with the installation of electrostatic separators. As discussed above, this technology by itself does not solve the problem, but it has been shown to extend the time before auto-degradation becomes a serious problem. Because degradation products are pro-oxidants and will catalyze new degradation, electrostatic separators also will lengthen the life of the antioxidant package in the oil. Electrostatic separators also will slow down the rate of auto-degradation.
Minimizing degradation. The ideal situation is to stop the fluid from degrading in the first place. For this, a root-cause analysis is needed. At many plants, the analysis has pointed a finger squarely at spark discharge from the mechanical filters. Though the velocity of the oil through the filters may be well within manufacturer’s specifications with no characteristic “snapping” sounds, high levels of very small spark discharges still may be occurring in your mechanical filters.
For most lube- and control-oil systems, removing all mechanical filtration is not a viable option. However, some simple changes can significantly reduce the spark phenomenon. If electrostatic filter technology is inplace to handle small particulates, the mechanical-filter ratings can be raised to levels that still provide adequate protection against catastrophic failure, while lowering the levels of spark. Filter OEMs are hard at work creating new designs that reduce sparking, but since the charges come from fluid shearing in the filters regardless of media, this may only be part of the solution.
Lower fluid velocities through the filters—and therefore lower levels of oil shearing— seem to be the key. With this in mind, users should consider increasing the micron ratings and maximizing the physical size of their filter housings and elements. At some sites, users are questioning the value of last-chance filters at servo valves a couple of years after unit commissioning, and are experimenting with simultaneously using both sides of the duplex lube-oil filters.
Another important factor in autodegradation is eliminating slow flows of cold oil. Users of “sandwich” type metered bypasses at the servo valves—referred to as a “CRV Plate” in Europe where it has been used more extensively—report excellent results in addressing varnish. Some sites are adding heat tracing to long, slowflowing lines like IGV oil, to reduce temperature fluctuations. As has been mentioned, this can have a dramatic effect on soluble degradation products that may precipitate and form varnish at lower temperatures.
Additives. New turbine oil is formulated with additives designed to help control the degradation process. These additives act as the first line of defense. In the early stages of service life, they typically are very effective, but as they perform their mission the additives are slowly depleted. At approximately 50% of their initial concentration, additives will begin to lose their ability to control degradation, and problems will be detected. This is a good time to consider additive replenishment. Note that replenishing additives in the field poses risks if not performed by experts.
When the additives have been depleted to approximately 25% of initial concentration, oxidation rates in the oil significantly increase, and serious problems become inevitable. At this point, the oil deteriorates to a dangerous level, condemnation levels are reached or exceeded, and users must consider flushing the in-service oil and replacing it with new oil (Fig 4). ccj oh
