Turbines

      PRODUCTS: LMG-20E™; LMG-30EŽ

 

     Gas turbine operators face many operational problems relative to the overall operation of their plants. Many of these problems are mechanical in nature and will not be covered in this discussion. The problems discussed will be chemical in nature and are problems that LMGI products are designed to minimize.

High Temperature Corrosion

       High temperature corrosion in gas turbines is caused primarily by vanadium. When the vanadium fuel contaminant is combusted, it combines with oxygen to form primarily:

    Vanadium Pentoxide  -   V2O5    

 

This compound melts about 675 °C, well above the temperatures found along the combustion path in the turbine. Vanadium pentoxide is a "solvent" for the passivating metal oxide that forms on the cobalt-chromium-nickel alloy surface. It dissolves the oxide coating causing another layer of the metal to form the oxide, which in turn is dissolved (each time the oxide dissolves, it is removed from the metal surface), which in turn causes another layer of the metal to form the oxide, etc. This process continues until eventually the metal disappears! When sodium is present the corrosion rate is accelerated. The higher the sodium to vanadium ratio, the greater the corrosion rate. The corrosion rate is also accelerated the higher the temperature of the gas path.

Filters

       Filters are very important on gas turbines. Gs turbines are multi-million dollars pieces of equipment with several components that have very tight tolerances (clearances). The filters are designed to remove as much "solid" material from the fuel as possible to protect flow dividers, fuel nozzles, fuel pumps, etc. Also the hot gas stream is "blowing" through the gas turbine at a fairly high velocity so any particulates in the fuel that pass through the combustions system can cause erosion of the rotating blades. This is not acceptable.

A.    Low pressure filters are normally inside of large filter housings. There should be two separate housings so each is capable of being used while the filters in the other one are being replaced. Depending on the fuel and the fuel flow, there will be from three to nine filter elements in each housing unit. An element is typically a cylinder about 35 cm tall by 20 cm in diameter. If the element is pleated paper, the paper is folded like an accordion and the ends joined to form a circle. If the filter is a metal screen it will probably be shaped the same way. The only difference between the two is essentially the media of the filter. Each end of the filter will be sealed in the end cap to keep fuel from passing around the ends of the filter. A rubber gasket on each end is compressed when more than one element is in a stack to keep fuel from leaking between filter elements. The direction of flow will almost always be from the outside to the inside. So "dirty" fuel enters the outside of the filters and is cleaned as it moves to the inside. On the inside of the filters there is a metal pipe with holes drilled around its outside to allow clean fuel to pass into the pipe. From there the clean fuel is normally forwarded to the gas turbine. Pressure from the fuel forwarding pumps drive the entire process (these pumps are normally located in a building near the tank farm, remote from the filters, heat transfer equipment is located in the same location).

Paper filters have provided very good service through the years, but they suffer from several drawbacks. Since the fuel is almost always heated, the paper filters can suffer fatigue. The resinous material that is used to seal the end of the filters into the caps has actually dissolved in some instances. As temperatures of fuels have increased, this has occurred more frequently. Filter manufacturers ave done an excellent job keeping their filters ahead of the increasing severity of their environments. When the filter undergoes fatigue, the filter pleats are bent over one another. As would be expected when this happens the, effective filter area is reduced. This will cause the pressure to increase more rapidly than expected, the filter elements will be considered to be plugged, and they will be changed. When fatigued elements are removed from the filter unit, it will be obvious this was the problem by even a casual inspection of the filter elements. Paper filters will have pore sizes of 5, 10, or 25 microns depending on the fuel and the turbine manufacturer's requirements. Less viscous fuels tend to have the smaller sizes.

Metal screens tend to be more robust than paper filters. Metal screens will be very expensive to purchase initially, but they can be cleaned (usually in an ultrasonic bath) to remove trapped debris. As their useful life continues, the holes in their surface will become more and more blocked with particles that could not be removed until the filter screens need to be removed from service permanently. This is expensive as many locations will have roughly nine elements in a filter unit, two units on a gas turbine, and the site will have from two to as many as 24 gas turbines. To keep all the filters serviceable is a major undertaking.

B.    High pressure filters are located after the main fuel pump. High-pressure filters normally comprise a single sheet of metal screen supported on a stiff "spring" encircling the inside of the filter element. The size of these filter elements are about the same as the paper elements described above. In fact in at least one case the low-pressure filter element was substituted for a high-pressure filter. High-pressure filters do very little filtering, their main job is to act as the last place to stop any particles large enough to damage the flow dividers or fuel nozzles. Still there have been times when these filters have been coated with a waxy material. Increases in fuel temperature are normally sufficient to control this problem. High-pressure filters typically have 75 or 100 micron pore sizes.

Storage Tanks

       The major problem with storage tanks is water accumulation in their bottoms. This is a normal situation. As fuel (in some cases well over a million liters) is stored in tankage, the water that represents only a tiny fraction of the total fuel settles to the bottom. As an example a relatively dry crude oil may contain only 0.1% water, but in a million liters that represents 1000 liters. With sufficient time, this will settle in the tank. A very normal operation at gas turbine sites is the draining of the tank bottoms. There should be a regular schedule set up to do this and it should be adhered to. The contaminants sodium and potassium are contained in this tank bottom water (as well as any other water around fuels). If a slug of this water should be pulled into the suction of the fuel pump and get all the way to the gas turbine, corrosion damage could result in the hot section. Frequent and regular tank draining is as important as adding the magnesium additive.

Solutions to Chemical Problems

       Problem solutions are divided into two general areas: fuel washing to remove the water-soluble contaminants sodium and potassium and the subsequent addition of a magnesium additive to inhibit the effects of vanadium and lead.

Fuel Washing

       The principle of fuel washing is to mix into the fuel "clean water" and then remove the water. The water-soluble sodium and potassium are removed with the water. Sodium and potassium arise from salts (typically sodium chloride - the salt in salt water - and potassium chloride) that are contained in the crude oil as it is pumped from the ground. Thus the only way sodium and potassium can exist in the fuel is to be present in water droplets. It is important to realize the sodium and potassium are very concentrated in the droplets of water. It is not uncommon to have the concentration be above 2000 ppm of both elements in the water droplets. Thus for a resid that contains 1% water and sodium plus potassium of 40 ppm, the water droplets would contain 40/0.01 = 4000 ppm in the water droplets. These are important points to remember, fuel washing relies on these principles.

       To remove the sodium and potassium it then becomes a "simple" matter or adding a quantity of sufficiently pure water, mix it thoroughly into the fuel to contact the water droplets in the fuel to dilute the concentration of sodium and potassium in the water droplets, and then to remove as much of the water as possible. These steps are accomplished in the following manner.

A.    Addition of a demulsifier is normally done first. Water and oil/fuel can form an emulsion when mixed vigorously. Petroleum emulsions are stabilized by naturally occurring materials found in nearly all petroleum products (asphaltenes, waxes, sediment, dirt, metal soaps, and other materials). When an emulsion of oil and water forms the water resists removal from the fuel. If this occurs during fuel washing, all that is accomplished is to add additional water (normally containing more sodium and potassium) to the fuel. So in order to eliminate existing emulsions and prevent the formation of new emulsions a demulsifier is added first. Emulsifiers mimic the chemical structure of many of the materials that stabilize emulsions with an oil-loving tail and a water-loving head. This allows them to penetrate the emulsion stabilizing film that forms on water droplets. However, specialized molecular branching of the demulsifier disrupts the film, which allows for coalescence of the droplets. As the droplets grow, they become large enough that gravity will cause them to "drop" from suspension in the fuel.

B.    Addition of water follows the addition of the demulsifier. Normally between 5 and 10 percent of the fuel volume of water is added. The quality of the water should keep the contained sodium and potassium as low as practical. It is not necessary to have water with zero sodium and potassium for it to be useable, their concentration should just be low (typically in the range of 50 ppm together).

C.    The water is then intimately contacted with the fuel to insure that the droplets of water are diluted with the fresh water. For electrostatic precipitators this is normally accomplished with a mix valve. In the case of a centrifuge, this is done with a mix tank. A mix valve is a globe type valve that is closed sufficiently to obtain a pressure on the fuel/water/demulsifier stream as it passes the valve. The net result is the passage of the materials through the valve causes the water to break into very fine droplets. These droplets will mix with the droplets that are already present in the fuel. This achieves the dilution of the "brine" desired. In a mix tank, slowly rotating paddles are used to mix the fuel and water mixture. Since centrifuges have more difficulty in breaking emulsions, the mix tank causes fewer emulsions for the centrifuge to resolve. Electrostatic units are more efficient in breaking emulsions so their mixing method can cause emulsions, but this allows much more intimate mixing.

D.    The fuel/water mixture is fed into the electrostatic unit or the centrifuge for resolution of the mixture into clean fuel and water. Both methods follow Stoke's Law. Stoke's Law is a relationship of how materials settle. The application of Stoke's Law can be viewed watching raindrops falling down a windowpane. The law is stated as follows:

Rate = 

(Sp.Gr. W - Sp.Gr. O)gd2


18vO

 

• Sp.Gr. is the specific gravity of either water (W) or oil (O);
• g is the acceleration of gravity;
• d is the diameter of the water droplet;
• v is the viscosity of the oil;
• Rate is the rate at which the water droplet falls through the oil.


Both electrostatic and centrifuge methods minimize viscosity by heating the oil (because an electrostatic unit is closed, higher temperatures can be achieved). Heating also reduces the specific gravity of the oil and of the water although to a lesser degree. Beyond this the two methods differ. Centrifuges increase the gravity term due to centrifugal force while electrostatic units rely on increasing the diameter of the water droplet. Since the diameter is a squared term, many times the electrostatic method can effect water separation more quickly. However, normally the selection of a fuel washing method is dependent on the size of the installation: electrostatic treaters being more efficient for large installations. Also, centrifuges have enjoyed widespread use throughout most of Saudi Arabia due to a process called waterless centrifuging. Centrifuges have (during the 1990's) been widely distributed in residual oil facilities too. However at the close of the 1990's the difficulties and inefficiencies of using centrifuges for residual plants is beginning to move the industry back to electrostatic units for these extremely heavy oil facilities.

  1. Electrostatic method: the method is quite simple.
    A grid is placed in the vessel. A high voltage, low amperage alternating current is passed between the grids the water layer acting as the grounding grid. Droplets of water alternate (due to the charges on their surfaces) between being elongated and spherical. Also the water surfaces are attracted toward each other. The elongation of the droplets causes the droplets to rupture the stabilizing film on their surfaces. When this film is ruptured the droplets coalesce with an adjacent droplet. This forms a larger droplet. As the diameter of the droplet increases (it is a squared term in Stoke's Law) they fall through the liquid faster. The coalesced water collects at the bottom of the electrostatic vessel where it is removed from the system. The clean oil leaves through the top of the vessel. Electrostatic treaters can be sized to treat very large flows of fuel. The electrostatic method is used to "desalt" nearly every drop of the world's crude oil as it enters the refining process. There are no moving parts in the electrostatic method except for fuel pumps to move the oil/water mixtures. By using multiple stages, the feed to each successive stage is the outlet from the previous stage with water flowing in a counter-current manner (cleanest water washing the cleanest oil, dirtiest water treating the dirtiest oil) fuel with very low sodium and potassium levels can be achieved (<0.5 ppm Na + K). This occurs because the concentration of sodium and potassium in the water droplets in the fuel is reduced. Using the example from above (1% water in the resid only now we want 0.5 ppm Na + K) the new water concentration is 0.5/0.1 = 50 ppm in the droplets. Obviously the lower the water content of the resulting fuel, the higher the concentration of the sodium + potassium in that water can be.>
  2. Centrifuge method: the method uses a centrifuge to effect the separation of water from fuel. The mixture of water and fuel is passed into the centrifuge. There centrifugal force cause the fuel and water to be separated. A centrifuge contains many closely spaced "plates" to enhance the capability of the centrifuge. The heavier water pushes down in the centrifuge and is removed out one line. The lighter fuel moves upward in the centrifuge and is removed through a separate line. One major disadvantage of the centrifuge relates to its ability to remove all manner of solids that may be in the fuel. The centrifuge can also remove asphaltenes and other fuel debris. These collect in the bowl of the centrifuge. It is not uncommon at a centrifuge site to have four centrifuges where only three would actually be required to treat the fuel volume. The fourth centrifuge is opened for cleaning out the bowl to remove the black gooey mess that collects there from fuel debris. Due to the nature of centrifuges, there are also many moving parts. In addition to each of the pumps needed to forward the fuel/water mixture (there are normally many centrifuges needed to treat a comparable volume of fuel that one electrostatic vessel can treat), there are many centrifuges in operation, each with its own drive to cause the centrifuge plate stack to rotate at very high speeds.


E.    Forward clean fuel to storage. No matter how the fuel is washed, after it has been washed it will be sent to clean fuel storage. Before the fuel is sent to storage there is one critical step that must be performed: the quality of the fuel must be determined. Remember the whole purpose of fuel washing is to remove sodium and potassium to a level that meets the turbine manufacturers requirements for these elements (0.5 to 1.0 ppm depending on the manufacturer and the fuel). Samples of the "clean" fuel will be checked for these elements using an emission spectrometer. If the fuel meets the specification it will then be sent to storage. If it does not, it will be returned to raw fuel storage where it will then be returned to the fuel washing process along with other fuel from raw storage. With both electrostatic and centrifuge systems, it is normal to have to return fuel at the start of the operation until the entire system has reached equilibrium. Also the level of demulsifier being used may not be proper to sufficiently lower the amount of water in the fuel to meet the trace metal requirements. Once everything is correct, all subsequent fuel should meet the required specifications. Only spot samples will be taken once the system is at equilibrium. Treated fuel storage has several advantages. The fuel will continue to shed water which will actually improve the quality of the fuel as the gradual drop out of water will also carry more sodium and potassium with it. And more importantly the treated fuel will be readily available when the plant requires it. Most fuel washing plants are sized smaller than the hourly requirements for fuel if all the gas turbines would be operating. This is done as a cost savings measure and also recognizes the fact that most plants do not run around the clock, nor are all gas turbines always operating. The fuel washing plant can easily keep up with fuel needs by running during gas turbine down times.

Magnesium Additives

       Early in the development (1950's) of gas turbines the corrosive effects of vanadium were noted. Many gas turbine manufacturers embarked on research programs to discover a solution to the corrosiveness of vanadium. As a result of all this work, one metal stood out as the most economical and effective of those tested: magnesium.

       The minimum treating ratio of 3 parts of magnesium to 1 part of vanadium was determined to be correct in the late 1960's to early 1970's. Initially the treat rate was set at 3.5 to 1 to insure adequate magnesium would be added. The more appropriate 3:1 was agreed upon as an industry standard since the early 1980's. The actual stoichiometric amount of magnesium required to just react with vanadium to make safe compounds is only about 0.7:1. However, additional magnesium is added because not only is the desired magnesium orthovanadate formed, but other less desirable magnesium vanadium compounds are also made. To force the reaction to the desired product, more magnesium is required. Other magnesium products are also formed (magnesium oxide and magnesium sulfate). More magnesium needs to be added to offset these less desirable compounds. And finally, since the time allowed for the reaction is very short (high gas velocity in the region of the flame), the greater the amount of magnesium added, the greater are the opportunity for a vanadium atom to find a magnesium atom.

       Most if not all gas turbine locations that use additives have selected either type sulfonates or carboxylates. Oil soluble additives, although sometimes more expensive to use than water solutions are much more convenient to use so that most users have decided the cost savings are not important. Another advantage of oil soluble products is they are delivered to the user ready to be used. With water solutions it is necessary to batch dilute the crystals and to either take a chance on the concentration, do an analysis, or treat with higher levels than needed to be on the safe side. Using more additive than required reduces any cost advantage the water-soluble products may have.

       No matter the source of magnesium, the mechanism to solve vanadium corrosion is the same: raise the melting point of vanadium pentoxide to one above the gas turbine temperatures. By adding magnesium, vanadium orthovanadate is formed instead of vanadium pentoxide. This reaction is reproduced below:

V2O5 + 3MgO ģ 3MgO • V2O5 (or rewritten as Mg3V2O8)

       Magnesium orthovanadate melts above 1200 (C. This temperature is well above typical gas turbine temperatures, especially blade temperatures due to blade cooling. When the system temperature is lower than the melting point of a compound, the compound (magnesium orthovanadate in this case) is not melted, it is a solid. Thus magnesium orthovanadate is solid in the gas turbine. Vanadium pentoxide is only corrosive while it is molten. When converted to the orthovanadate (in the flame) it will pass harmlessly through the system. Thus by adding the appropriate quantity of magnesium (3:1) the system will be protected from corrosion. This has been the case for well over 25 years of magnesium use in gas turbine applications using heavy fuels.

       There is one disadvantage to using magnesium additives in gas turbines. The increased amount of metal that needs to pass through the gas turbine will lead to more rapid deposits on turbine blades. This becomes a problem as the deposits add to the weight of the rotating section of the gas turbine. This extra weight causes a loss of power output from the gas turbine. Eventually the gas turbine will need to be stopped to clean the deposits from the blades. With heavy residual use the need to stop operation may be every 200 hours of operation when the metal content is high. This operation is completely normal and expected. These cleaning cycles were considered when the fuel for the gas turbine was being selected.

       To clean the turbine it must be stopped completely. All heavy fired gas turbines have a washing system as part of their equipment. The turbine inside, after cooling, is sprayed with water. In some installations it may be standard procedure to completely fill the turbine section. The deposits are allowed to soak for a set period of time and then the water is drained out. This normally will return the turbine to full power. The washing operation is made easier by another compound resulting from magnesium: magnesium sulfate. Magnesium sulfate results from the combustion of sulfur found in all fuels. In the oxygen rich gas turbine environment, much of the sulfur will be converted to sulfur trioxide. This combines with magnesium oxide and forms magnesium sulfate.

       Magnesium sulfate is very water-soluble. So during washing operations the magnesium sulfate in the deposits dissolves thereby loosening all other deposit materials. This makes the washing operation fairly easy to perform. Two cautions need to be made concerning washing.

  1. Each turbine manufacturer has their own procedure. This can also vary among different types of gas turbines from the same manufacturer.
  2. Another compound formed from magnesium can interfere with good washing. This compound is magnesium oxide. When turbine temperatures are too high, magnesium sulfate can be converted back to magnesium oxide. Magnesium oxide is not water-soluble. Thus if there is too little magnesium sulfate, incomplete washing may result. This can be compensated for by lengthening the soak period and by repeating the washing cycle as required. This decision will be up to the turbine operator.

 


 HOMEPAGE