Inlet Gas Contamination Causes Degradation in Turbo Machinery Performance

Problem Area

Liquid and solid contamination in inlet gas can corrode, erode, and foul turbine and compressor blades, thereby reducing their efficiency and service life. The following are examples of the types of contamination found in turbine blades.

Typical Turbomachine Blades


Top

Corrosion

Corrosion is the loss of material caused by chemical reaction between machine components and
contaminants which can enter the gas turbine through the gas stream, fuel system or water/steam
injection system. Salts, mineral acids, elements such as sodium, vanadium, and gas, including
chlorine and sulphur oxides in combination with water, can cause corrosion, especially in the

Top

Erosion

Erosion is the abrasive removal of material by hard particles suspended in the gas stream. Particles causing erosion are normally 10 microns or larger in diameter. Particles with diameters between 5 and 10 microns fall in a transition zone between fouling and erosion.

Erosion of turbine and compressor blades has been well studied and documented.(1) It is generally agreed that erosion damage increases with increasing particle diameter and density, flow turning and gas velocity, and with decreasing blade size. Turbine and compressor manufacturers minimize erosion by increasing trailing edge thickness, installing field replaceable shields and using improved alloys. Nevertheless, they all recommend fine inlet filtration to prevent hard particles from entering the turbines.

Top

Fouling

Fouling is the adherence of particles and droplets to the surface of the turbomachine blading. This degrades flow capacity and reduces efficiency in a short period of time. Fouling can normally be reversed by cleaning, but it often requires downtime. Fouling is a serious problem, particularly in the oil and gas industry where sticky hydrocarbon aerosols are universally present. Traditionally, no accommodation has been made in designing turbines to tolerate deposition tendencies of particulate-laden gas streams. Although the deposition trajectories can be predicted for some turbine blades, the actual fouling is very much dependent on inlet gas cleanliness which varies unless it is controlled.

Top

Mechanism of Blade Fouling

The primary mechanisms of particle delivery in turbines are inertial impaction, turbulent eddy diffusion and Brownian motion.

Figure 1: Contaminant Deposition of Blades


Stator / Rotor

For inertial impaction, particle inertia cause its trajectory to deviate from flow streamlines in the area where flow changes direction. This causes particles to impact on blade surfaces (Figure 1). Mass flux (arrival date) due to this mechanism decreases with decreasing particle diameters and inertia becomes very small for particles smaller than 1 micron.

For turbulent eddy diffusion, particles become entrained in eddies of turbulent boundary layers and are swept toward the blade and vane surfaces. Small particles (0.1-1.0µm ) are slowed and trapped by the viscous drag forces of boundary layers.

In Brownian motion, very small particles (0.1µm and finer) are randomly transported to the turbine surfaces. Decreasing particle (aerosol) size and density increase the deposition rate due to Brownian motion.

There is another type of very small particles (0.01-0.1µm) deposition on the turbine blades caused by temperature gradient (Thermophoresis) which is usually negligible in low temperature applications. This is typical in oil and gas industry operations.

Because of the broad distribution of particles and droplets in gas streams, (0.1-10µm) almost all of the above mechanisms govern the fouling of turbine blades.(2)

The formation of low melting point eutectics is an important fouling mechanism in power recovery hot gas expanders used in fluid catalytic cracking units.(3)

Figure 2: Compressor Fouling And Its Effect On Gas Turbine Performance (Ref. 3)


Top

Turbine Power Degradation

The primary degradation effect of deposition is flow blockage at the throats of vane and blade passages.(4) This reduction in gas flow through the turbine nozzle results in a loss of gas turbine output. Figure 2 illustrates the effects of fouling on typical gas turbine efficiency and outputs. In attempting to compensate for the loss in output, turbines consume additional fuel which increase firing temperature and generate undesirable increases in heat rate, also shown in Figure 2.

It is customary to shutdown the turbine at the point of 10% power loss for cleaning. This downtime can be frequent and costly. Figure 3, which is based on field studies by a major operator and two major turbine manufacturers, gives some interesting information on turbine fouling.(5) The combustion turbine was a single shaft machine with about 50 MW power output, a pressure ratio of about 10:1 and inlet temperature of 871°C (1600°F).

Figure 3: Power Drop Due to Deposition for 10:1 Pressure Ratio Turbine (Ref. 2)

This graph (Figure 3) clearly shows that the output loss is directly proportional to inlet gas contamination level. It would only take 20 operating hours to drop the output to 90% with 549 ppm contamination in gas. For middle range contamination (58 ppm), a 10% loss occurred after 250 hours. The clean up was necessary to maintain the output within the 90-100% range. When the gas was further cleaned up by a granular filter (19.5 ppm), the 10% power loss was delayed to 540 hours.

Top

Inlet Gas Filtration

With the current advances in filtering solid particles and coalescing aerosols from gases, the turbine inlet gas can economically be filtered to contain no more than 0.01 ppm with a particle size cut-off at 0.3µm. That absolutely eliminates erosion. Fouling should virtually be controlled to limit 10% power loss in no less than 20,000 operating hours which is beyond normal plant shutdown. That can be extrapolated from the upper-right line on Figure 3.
Regarding corrosion control, particle filters and liquid gas coalescers are mechanical septa and do not separate corrosive vapor and gases from fuel gas. However, most of corrosive salts are dissolved and carried by liquid aerosols. Most aerosols are between 0.1 - 0.6µm and are quite removable by fine liquid/gas coalescers. That should reduce, not eliminate, corrosion as a bonus of fine filtration.

Top

Solution

A typical blade design lifetime for gas turbines operating with clean fuels is greater than 45,000 hours and a minimum of 25,000 hours between major overhauls.(4) That can be achieved by a series of efficient hydrocyclones and granular bed filters, provided that bed filters do not pass any media downstream. In practice, the media filters may help reduce fouling, but sand particle releasing downstream can increase erosion due to media migration downstream of filters.

Pall offers a single stage liquid/gas coalescer to protect compressors and turbines from loss in efficiency. This coalescer rated at 0.3µm (99.98% efficient by aerosols count method)(6) will consistently give an effluent of less than 0.01 ppm (liquids and solids). That limits the turbine power loss to less than 10% in 25,000 hours or longer operating time. As a result, the overhaul of blades due to erosion and fouling will not happen during the manufacturers' predicted list (e.g. 25,000 hours).

For hot gas expanders as used in FCC unit power recovery trains, Pall recommends a porous stainless steel blowback filter. The filter elements rated at 1µm absolute provide even finer filtration due to the presence of cake formation. Removal efficiencies are in excess of 99.9% by weight.

Top

References

  1. M. Menguturk, D. Gunes, M. Erten, and E.F. Sverdrup, “Multistage Turbine Erosion,” ASME Paper 86-GT-238, presented at the ASME International Gas Turbine Conference and Exhibit, Dusseldorf, West Germany, June 8-12, 1986.
  2. R.A. Wenglarz, “An Assessment of Deposition in PFBC Power Plant Turbines,” Trans. ASME, J. of Eng., Power, Vol. 103, July 1981, pp 552-560.
  3. D.H. Linden, “Catalyst Deposits in FCCU Power Recovery Systems Can Be Controlled,” Oil and Gas Journal, December 15, 1986.
  4. M.K. Pulimood, “Field Experience with Gas Turbine Inlet Air Filtration,” ASME Paper 81-GT-193, presented at ASME Gas Turbine Conference Products Show, March 9-12, 1981 Houston, Texas.
  5. R.A. Wenglarz, “Rugged Turbines for PFBC Power Plants,” presented at 1982 AIAA/ASME Joint Fluids, Plasma, Thermophysics and Heat Transfer Conference, St. Louis, Missouri, June 7-11, 1982.
  6. K. Williamson, S. Tousi, and R. Hashemi, “Recent Development in Performance of Gas/Liquid Coalescers,” presented at The First Annual Meeting, American Filtration Society, Ocean City, Maryland, March 21-25, 1988.

Top