Continued From Part I

Dependent on the type of fuel gas, the geographical location and the forwarding means there is the potential for the “raw” gas supply to contain one or more of the following contaminants:

1. Tar, lamp black, coke
2. Water, salt water
3. Sand, clay
4. Rust
5. Iron sulfide
6. Scrubber oil or liquid
7. Compressor Lube oil
8. Naphthalene
9. Gas Hydrates

It is critical that the fuel gas is properly conditioned prior to being utilized as gas turbine fuel. This conditioning can be performed by a variety of methods. These include, but are not limited to: media filtration, inertial separation, coalescing and fuel heating. Table 2b identifies the trace metal, particulate and liquid contamination limits. It is critical that fuel gas conditioning equipment be designed and sized so that these limits are not exceeded. For further information on gas fuel conditioning, see publication GER 3942(8).

A. Particulates
Contamination limits for particulates are established to prevent fouling and excessive erosion of hot gas path parts, erosion and plugging of combustion fuel nozzles and erosion of the gas fuel system control valves. The utilization of gas filtration or inertial separation is instrumental in ensuring that the particulate requirements as defined in Table 2b are met. GE recommends the use of stainless steel piping down-stream of this last level of filtration to prevent the generation of corrosion-derived particulates.

B. Liquids
As identified in Table 2b, zero liquids are allowed in the gas turbine fuel gas supply. Refer to note 16 of Table 2b for current input requirements and turbine controller actions. The introduction of liquids with gas fuel can result in nuisance and/or hardware damaging conditions. These include rapid excursions in firing temperature and gas turbine load, primary zone re-ignition and flashback of premixed flames. In severe conditions, liquid carryover to the first stage turbine nozzle may result in damage to downstream hot gas path components. When liquids are identified in the gas fuel supply, phase separation and heating must be employed to achieve the required superheat level.

C. Sulfur
There are several concerns relative to the levels of sulfur contained in the fuel gas supply. Many of these are not directly related to the gas turbine but to associated equipment and emissions requirements. These concerns include but not limited to:

  1. Hot Gas Path Corrosion
    Typically , use of sulfur bearing fuels will not be limited by concerns for corrosion in the turbine hot gas path. Experience has shown that fuel sulfur levels up to about 1% sulfur do not significantly affect oxidation/corrosion rates. Specifying the fuel alkali le vels to values shown in Table 2b controls hot corrosion of hot gas path parts resulting from alkali sulfate formation. Unless sulfur levels are extremely low, alkali levels are usually limiting in determining hot corrosion of hot gas path materials. For low Btu gases, the fuel mass flow rate at the turbine inlet is increased over that for natural gas, and the alkali limit in the fuel is therefore decreased.
  2. HRSG Corrosion
    If heat recovery equipment is used, the gas fuel sulfur concentration must be known so that the appropriate design for the equipment can be specified. Severe corrosion from condensed sulfurous acid results if a heat recovery steam generator (HRSG) has metal temperatures below the acid dew point.
  3. Selective Catalytic Reduction (SCR) Deposition
    Units utilizing ammonia injection downstream of the gas turbine for NOx control can experience the formation of deposits containing ammonium sulfate and bisulfate on low temperature evaporator and economizer tubes. Such deposits are quite acidic and therefore corrosive. These deposits and the corrosion that they cause may also decrease HRSG performance and increase backpressure on the gas turbine. Deposition rates of ammonium sulfate and bisulfate are determined by the sul-phur content of the fuel, ammonia content in the exhaust gas, tube temperature and boiler design. Fuels having sulfur levels above those used as odorants for natural gas should be reported to GE. In addition, the presence of minute quantities of chlorides in the compressor inlet air may result in cracking of ANSI 300 series stainless steels in the hot gas path.
  4. Exhaust Emissions
    Sulfur burns mostly to sulfur dioxide, but 5%-10% oxidizes to sulfur trioxide. The latter can result in sulfate formation, and may be counted as particulate matter in some jurisdictions. The remainder will be discharged as sulfur dioxide. To limit the discharge of acid gas, some localities may restrict the allowable concentration of sulfur in the fuel.
  5. System Material Requirements
    When considering fuel gases containing H2S (sour gas fuels) material selection for system piping and components shall comply with NACE Standard, MR0175(9).
  6. Elemental Sulfur Deposition
    Solid elemental sulfur deposits can occur in gas fuel systems downstream of pressure reducing stations or gas control valves under certain conditions. These conditions may be present if the gas fuel contains elemental sulfur vapor, even when the concentration of the vapor is a few parts per billion by weight. Concentrations of this magnitude cannot be measured by commercially available instrumentation and deposition cannot therefore be anticipated based on a standard gas analysis. Should deposition take place, fuel heating will be required to maintain the sulfur in vapor phase and avoid deposition. A gas temperature of 130°F or higher may be required at the inlet to the gas control valves to avoid deposition, depending on the sulfur vapor concentration. The sulfur vapor concentration can be measured by specialized filtering equipment. Contact GE for further information on this subject.
  7. Iron Sulfide
    The presence of sulfur in the gas may promote the formation of iron sulfides. Under certain conditions, iron sulfide is a pyrophoric material that can auto ignite at atmospheric pressure and temperature when exposed to air. Extreme care must be taken when servicing gas fuel cleanup equipment to avoid accidental exposure to oxygen and subsequent combustion of filter material.
  8. Trace Metals
    Sodium is the only trace metal contaminant normally found in natural gas. The source of sodium in natural gas is salt water. Limits on trace metals are established to prevent the formation of corrosive deposits on hot gas path components. These deposits can be detrimental to gas turbine parts life. In addition to sodium, additionally harmful trace metal contaminants can be found in gasification and process gases. GE will determine limits on these contaminants on a case-by-case basis.

A. Particulates
The major source of particulates is from corrosion products in the pipeline. These products are
continuously formed over the life of the pipeline at rates that are determined by the corrosive
components and water content of the gas. For this reason most gas suppliers control moisture content to avoid the possibility of water condensing and forming acid with CO2 or H2S.

B. Liquids
Gas at the wellhead is typically produced in a wet saturated condition and is treated to remove water and heavy hydrocarbon liquids. Depending on the degree of treatment and the supply pressure, the gas may be delivered in a dry condition. After pressure reduction it is possible for condensates to form as the gas cools during expansion. The condensates may be either water or hydrocarbons or both. Carry over of lubricating oil from compressor stations is another source of liquids.

C. Sulfur
Sulfur is usually combined with either hydrogen as H2S or carbon as COS. It is produced with the natural gas. Typically the gas supplier will limit H2S to a concentration of less than approximately 20 ppmv by removing sulfur in a treatment system. Sulfur may also be present in very low concentrations (< 100 ppbv) in the form of elemental sulfur vapor.

D. Trace Metals
The details of various sources of alkali contaminants in the following text apply to MS 3000, MS 5000, B, E and F class machines. The sources of contaminants for 6C, FB and H-class machines are discussed in GEK 107230(1). Contributions to the alkali content of the combustion gases can come from any of the material streams supplied to the combustor; fuel, air, water or steam. The basic parameter which can be used to define the allowable alkali metal content admitted to the turbine is Xt, the combined sodium and potassium con-tent of the combustion gas at the entry to the first stage nozzle. This concentration must not exceed the values stated in Table 2b. Since there is no simple test method for measuring Xt in an operating turbine, it must be calculated from the alkali metal contents of the fuel, air, water and steam flows.
T(Xt) = A(Xa ) + F(Xf ) + S(Xs ) + W(Xw )

Where T = total flow to turbine (= A + F + S + W):
= alkali contaminant concentration in total flow as Na.
A = Air flow,
Xa = contaminant concentration in air.
F = Fuel flow,
S = Steam flow,
Xs = contaminant concentration in steam.
W = water flow
Xw = contaminant concentration in water.
The allowable levels of alkali contamination in the different flows entering the gas turbine are
discussed below:

E. Air
There are four sources of alkali metal contained in the compressor discharge air, (a) Inlet filter carry over of sodium chloride in ambient air (b) carry over of sodium dissolved in water used for evaporative cooling (c) carry over of sodium dissolved in water used for inlet fogging and (d) carry over of sodium from water used for on-line water washing. When concentrations of trace metals in fuel, water or steam are not precisely known, a value of 0.005 ppmw, GER 3419(10), can be used for systems with or without evaporative coolers. This value, based on experience, would cause an insignificant contribution to the overall contamination level and have a minor impact on parts lives.

For systems with inlet foggers, the water carry over is 100% compared with approximately 0.003% for evaporative coolers and the potential sodium carry over is therefore proportionally higher. The maximum inlet fogging water flow rate is approximately 1/3 of the natural gas flow rate and, depending on the sodium concentration, could use up a significant portion of the total allowed at the turbine inlet. Refer to GEK 101944(11) for information on water purity requirements. If it is anticipated that the specification could be exceeded, General Electric should be consulted for recommendations on the selection of the water source and use of proper air filtration equipment.

F. Steam
Steam for gas turbine injection is typically taken from a suitable extraction point on a steam turbine or HRSG. The limiting purity requirements for this steam are those for the steam turbine. These limits are defined in GEK 72281(12).

G. Water
The maximum alkali meta l (sodium plus potassium) content of water to be used for injection is
discussed in GEK 101944(11). The maximum water injection rate is approximately equal to the fuel injection rate. If the alkali content approaches the maximum allowable value stated in GEK
101944(11), it may use all of the allowable margin at the turbine inlet, leaving none available for the fuel. If it is anticipated that the specification could be exceeded, GE should be consulted for recommendations on the selection of the water source.

The concentration of sodium in steam and water, at the levels specified, can be
measured directly using an on-line sodium analyzer or, in the laboratory, with an
ion or pH meter fitted with a “sodium specific” electrode or by an atomic
absorption spectrometer fitted with a graphite furnace.

H. Fuel
The final source of contamination to be considered is the fuel. Most cases of alkali metal
contamination and corrosion of hot gas path components are related to liquid fuel contamination. It is rare that natural gas will contain trace metals but gasification fuels may contain alkalis carried over from the gas clean up system. The issues and requirements with gas fuels are discussed separately in the prior sections.


The effective particle size for erosion considerations is determined by a particle’s terminal settling velocity. The size and density distribution of the solid particles must be such that not more than 1.0 percent by weight of the particles shall have a terminal settling velocity in air (70°F) and 30 inches Hg, absolute) greater than 14 inches per minute, and not more than 0.1 percent shall have a terminal settling velocity in excess of 23 inches per minute. For a solid spherical particle, Stoke’s Law of settling permits calculating the terminal settling velocity if the particle size and shape and particle specific gravity are known. The following tabulation gives the spherical particle diameters equivalent to the limiting terminal settling velocities for particles of specific gravity 2.0 and 4.

The following relationships can be used to determine turbine inlet contaminant and fuel equivalent contaminant concentrations to compare to limits given in Table 2b. In general, for mass balances on the flows and on the contaminants from fuel, air and steam/water at the turbine inlet.
E = F + A = S                                                                                                           (5)
(XE)E = (XF)F + (XA)A + (XS)S                                                                  (6)
A, F, S and E are the mass flows of air, fuel, injected steam/water and combustion gases at the turbine inlet, respectively XA, XF, XS, and XE are the contaminant concentrations (ppm by weig ht) in the inlet air, in the fuel, in the injected steam/water, and in the combustion gases at the turbine inlet, respectively.

The fuel equivalent concentration of contaminants (XFe) is (dividing (5) and (6) by F)

XFe = XE(1 = A/F + S/F) = XF + XA(A/F) + XS(S/F)                                                     (7)

For example for a natural gas contaminated with salt at a sodium concentration of 0.1 ppm, an air sodium contamination of 5 ppb, and a steam contamination of 5 ppb, the equivalent sodium in the fuel for A/F =50 and S/F=1 is from equation (7)

XFe = 0.1 + 0.005(50) + 0.005(1) = 0.355ppm                                                             (8)

This is the amount of sodium considered to come from a fuel source only, which gives the same
sodium concentration at the turbine inlet as from the combined three sources. The turbine inlet concentration, XE, is found by rearranging (7)

XE = XFe /(1 + A/F + S/F) = 0.355/(1 + 50 + 1) = 0.006827ppm = 6.827 ppb        (9)

These values are well within the sodium specification values of 1 ppm on a fuel basis (XFe), and 20 ppb for the turbine inlet concentration (XE). XE and XFe values for all contaminants are given in Table 2b. For equivalent contaminant relationships in 6C, FB, and H-Class machines, refer GEK 107230(1).

The gas fuel pressure at the purchaser’s connection FG1 should first be determined. The hydrocarbon and moisture superheat can then be calculated from the equations shown on Figure 1. For example, at a gas fuel pressure of 490 psia, the moisture and hydrocarbon superheat requirements are 18°F and 42°F respectively.

For a typical pipeline gas the maximum allowable moisture content is 7 lbs/mmscft. At 490 psia the moisture dew point can be determined from Figure 2 and is equal to 23ºF. The minimum gas temperature to avoid moisture condensation is equal to the moisture dew point plus the required superheat, or 23º+ 18º = 41ºF.

For hydrocarbons, the maximum expected hydrocarbon dew point should be determined by
consultation with the gas supplier and a review of historical data if available. The hydrocarbon dew point will be dependent on the source of the gas, the degree of gas processing and may vary seasonally with overall gas demand and the economics associated with liquids removal and recovery.

For a fuel with a maximum hydrocarbon dew point of, for example, 35ºF at 490 psia, the theoretical minimum gas fuel temperature is 42º+ 35º= 77ºF. The temperature, in this example, is higher than the minimum required to avoid moisture condensation, therefore it establishes the minimum gas temperature to avoid both hydrocarbon and moisture condensation. However, the actual minimum gas fuel temperature will be based on the hydrocarbon dew point breakpoint. (see section III, C) In the example above, the assumed hydrocarbon dew point was 35°F. (at 490 ps ia) The hydrocarbon dew point for this case is higher than the breakpoint. (-25°F) Therefore, GE will assume that the gas entering the superheating system is saturated, and will require the full hydrocarbon superheat be applied. For example, if the gas entering the superheating system is at a temperature of 55°F, then the
minimum gas temperature would be 97°F. (55°F + 42°F) The 97°F temperature is what the turbine controller will begin to take protective action on. The use of the 97°F temperature, instead of the 77°F temperature is based on GE’s experience of gas fuel dew point variation on sites.

For another example, if the hydrocarbon dew point is calculated as -30°F, (at 490 psia) then the hydrocarbon dew point would be less than the breakpoint. (-25°F) Therefore, the minimum gas fuel temperature would be based on the moisture dew point and superheat. Assuming a 7 lbs/mmscft, the dew point is 23°F and superheat required is 18°F, resulting in a theoretical and actual minimum gas temperature of 41°F. (23°F + 18°F)

Disclaimer: This Paper is Property of GE, it has been printed here for your reference and ease of our readers.

One thought on “Part 2: Specification for Fuel Gases for Combustion in Heavy-Duty Gas Turbines”

Leave a Reply