Impact of Fuel Composition on Gas Turbine Engine Performance Free

Industrial gas turbines have proven to be increasingly versatile over the years of being able to use fuels made up of a wide range of constituents. The thermochemistry of the fuel is consequential to the resultant performance of the engine, carbon emissions, the operability of the combustion system, and the durability of hot section components. These are examined within the paper along with a property review of gaseous, liquid, and solid hydrocarbon fuel, a discussion of related applications, and a few other musings. Some general observations can be made when considering attributes of the hydrocarbon fuel being used, in particular the hydrogen–carbon ratio (H/C).

• Shaft power and thermal efficiency increase as the hydrogen to carbon ratio (H/C) increases.

• A higher H/C produces more H₂O and less CO₂ in the products of combustion.

• For conventional petroleum and natural gas-based fuels, fuels with higher H/C have higher specific energies. Fuels with lower H/C contain a wider range of heavy hydrocarbons and particulates resulting in more stringent fuel treatment and injection processes.

• Synthetic gases with high H₂ and CO content typically contain inert gases, thus lowering their heating value compared to natural gas and other hydrocarbon-based fuels.

• Fuel composition not only affects the energy content of the fuel but also affects its physical and chemical properties that highly influence combustor operability.

One fascinating aspect of combustion is how a wide range of substances in solid, liquid, or gaseous states react with oxygen in air to result in an exothermic reaction that can be extracted as useful energy. Under human control and guidance, hydrocarbon fuels used in conventional processes start their intended journey to productively release energy resulting in a finality of heat and, ideally, fully oxidized carbon and hydrogen forming carbon dioxide and water. The fuel property variations that exist with the existing wide range of fuels result in changes in the exergy emitted from the system. The mechanical energy from the system is pure exergy, while a quantity of heat or a hot pressurized fluid contains a mix of exergy and anergy. Wettstein et al. [1] references exergy as “useful energy.” Exergy is the part of energy that can be converted into any other forms of useful energy while anergy cannot be converted.

As society strives toward alternative energy sources and potential energy carriers, one needs to consider nonhydrocarbon fuel candidates that are abundant, economically and/or environmentally feasible to produce. Even many metals “love to stick to oxygen” releasing energy in the form of heat and light [2]. There have been recent studies suspending nano- or micron-sized metal particles in fuel droplets to enhance combustion energies resulting in higher energy density, shorten ignition delay time and enhance fuel oxidation by catalytic effect. The added metallic to hydrocarbon fuels consequently results in less CO₂, NO_x, and other species [3]. Though the study of what burns in our world is quite interesting in its own right, our focus is to show systemic influence hydrocarbon fuels, from pure hydrogen to pure carbon and everything in-between, have on functioning industrial gas turbines. This range of fuels also includes common industrial gases that result from a variety of hydrocarbon gasification processes.

This paper is divided into two primary sections: (i) an engine cycle analysis using two GT engines that investigates the influence of fuel thermochemistry on engine performance, and (ii) a literature review that discusses the influence of fuel chemical and physical properties on its preparation and combustor operability.

From simulation, the effects of changing fuel type can be seen on output power and efficiency of the engine while sustaining normal operating constraints such as speed and operating temperature, which is typically turbine rotor inlet temperature (TRIT) into the turbine just after the combustor. Our initial case study shows a comparison of typical fuels used on industrial gas turbines including natural gas, diesel (DF-2), and kerosene. A standard fuel composition for each blend was selected for simulation purposes. A frequently asked question from a customer could be “Why does the engine produce less power when operating with diesel fuel compared to natural gas, despite operating to the same TRIT?.” Our case study answers this question considering a single shaft or cold end drive and a two shaft or hot end drive (HED) variants of our Solar Turbine engines that can burn these gaseous and liquid fuels, otherwise known as Solar Turbine's “dual fuel” combustion systems.

Running a Taurus^TM 60 7901 rating to full load at standard conditions (sea level, Tamb = 59 °F, relative humidity = 60%), with natural gas, delivers a power near 8000 hp, a 12.3:1 pressure ratio, and an inlet airflow of 47.5 lbm/s. The Taurus 60 7901 is a single shaft engine running to a constant speed. Since both diesel (DF-2) and kerosene have lower heating values (LHV) less than natural gas, more fuel is needed to reach operating TRIT. This additional mass flow slightly increases cycle pressure ratio as the cycle rematches. Airflow changes only slightly as it follows the compressor characteristic. The notable change in operation is reduced power and efficiency due to the reduced hydrogen carbon ratio in the fuel.

Table 1 shows pertinent engine parameters relativized to natural gas in the first column. Despite constant engine speed, flow, and TRIT, the engine output power and thermal efficiency is significantly reduced apparently only due to the properties of the gas expanding across the turbine, which is the resultant mixture from the products of combustion.

To explain how the difference occurs, we examine the enthalpy versus entropy (h-s) diagram depicting each fuel for the Taurus 60 7901 cycle at full load shown in Fig. 1. The station numbers are noted (1,2,3,7). Figure 2 provides a pictorial reference, showing station designations defined for a Solar Turbines engine. Station 0 represents the inlet to the system. Station 1 is the inlet to the engine. Station 2 is the exit of the compressor. Station 3 is the exit of the combustor and inlet to the turbine. Station 5 is an interturbine location which is the separation between the gas producer (GP) and power turbine (PT) (turbines on a HED engine, respectively). Station 7 is the exit of the turbine exhaust to the atmosphere.

Equation (1) power across the power turbine (PT).

Table 2 shows enthalpy values normalized to natural gas at station 3. Both diesel (DF-2) and kerosene depict lower relative enthalpy at the inlet and exit of the turbine. This shows a lower enthalpy differential across the turbine which results in less work and hence lower power despite higher fuel flow to compensate for a lower LHV. The effect of the lower work is compounded by the necessity of providing about the same power to drive the compressor which consumes about two-thirds of the turbine power. This leaves proportionally much less power to deliver to the output shaft.

Examining the same fuels on a Titan^TM 250 30000 two-shaft engine (HED) shows a stronger effect on output parameters than what was shown on the single shaft engine. Table 3 shows engine performance parameters normalized to the natural gas values for a Titan 250 30000 engine. Note that a Titan 250 delivers about 30,000 hp output power, 24:1 pressure ratio, and about 150 lbm/s inlet airflow. The table shows a significant drop in output power and efficiency with the fuels with lower hydrocarbon ratio but also shows that the engine match has changed resulting in less inlet airflow and lower power turbine pressure ratio.

Similar to Table 2, Table 4 shows enthalpy values normalized to natural gas at station 3 for the Titan 250 30000 cycle. Both diesel and kerosene again depict lower relative enthalpy deltas across the gas producer and power turbines. There is less work across both turbines combined with less inlet airflow for the lower hydrocarbon fuels resulting in a more significant reduction in power and thermal efficiency compared to the single shaft example with the Taurus 60. In general, two shaft engines relative to single shaft engines have stronger output performance sensitivities associated with engine faults and property changes due to a rematch effect which also varies the flow and pressure ratio operating to a fixed TRIT with fixed turbine geometry.

Figure 3 shows output power and thermal efficiency relative to methane (CH₄) for fuels ranging from pure hydrogen to hydrocarbon fuels with decreasing the mass-based H/C, or H/C_m. Figure 4 shows the same data excluding hydrogen plotted versus H/C_m. In general, the lower the H/C_m the lower the output power and thermal efficiency at full load. However, the other influence to note is the relative molecular weights of the different hydrocarbon fuels. With lower H/C_m and monotonically increasing molecular weight, the sensitivity to the effect on output performance is consistent and nearly asymptotic. The lower molecular weight and higher LHV in ethylene (C2H4) compared to the adjacent H/C_m molecules shows a different resultant in output performance where power is notably less. Higher LHV will result in lower fuel mass flow addition, and a slightly lower amount of gas expanded across the turbine. Fuel LHV and H/C_m (or H/C) are the distinctive fuel properties qualifiers that drive the resultant performance of the gas turbine engine. Figure 5, which shows performance improvement of the other fuels relative to diesel (DF-2) fuel, simplifies depicting the performance benefit between diesel and methane fuels. The Titan 250 increases by about 4% in power and 1.4% in heat rate improvement when running to natural gas relative to running with diesel fuel.

Another possibility to strongly alter output performance is to add inert gases such as CO₂ and N₂. The engine will still run to a constant TRIT and will take substantially more fuel flow due to a lower heat content to meet the control objectives, resulting in higher output power. Synthetic and by-product gases typically have high inert contents and are classified as low Wobbe index (WI) fuels shown in Table 5. Another type of operation that results in power augmentation is the injection of atomized H₂O in amounts of up to 1:1 with fuel to aid emissions control. However, the energy used to pressurize and prepare this addition of mass flow can also be viewed as a parasitic loss to the overall system depending on its required work input.

Another attribute to be mindful of associated with H/C_m is the amount of CO₂ and H₂O being created from the products of combustion. Though CO₂ is an odorless, colorless, plant friendly gas, it is the most significant long-lived greenhouse gas, increasing in concentration since the industrial revolution. More CO₂ in the atmosphere leads to more global warming¹. The focus for all who burn fossil fuels is to minimize pollutants and more frequently overall carbon emissions; not water vapor so much. In the case of water vapor being produced, this is also a greenhouse gas causing 60% of the warming effect, but whether this is contributing to further global warming is less clear. Water vapor is a condensable gas that is limited to what can be contained within the atmosphere. The water vapor forms clouds and condenses to make rain based on atmospheric temperature. If there is global warming, you can have an effect of more water vapor contained in the air as the temperature increases by other noncondensing greenhouse gases. Water vapor itself can cause a warming effect, but contradictorily creating more clouds, energy from the sun is reflected and does not reach the Earth's surface to warm it. Understanding this balance remains an active area of climate science research [5].

Back to our study of how different fuel attributes affect gas turbine performance, Fig. 6 shows the relative change in the product H₂O and CO₂ mass composition for higher H/C_m from diesel (DF-2) fuel to methane assuming complete combustion. The higher H/C_m results in more H₂O and less CO₂ in the products of combustion. The proportion of H₂O and CO₂ is directly proportional to the H/C_m. The plot shows a 48% increase in the mass composition of H₂O and a 24% reduction of CO₂ from diesel to methane.

Equation (2) is the stochiometric equation of the reaction of a hydrocarbon and air.

Figure 7 shows the stochiometric equations for each of the fuels depicted in the detailed cold end drive and HED comparison studies. Methane is shown in the reaction demonstrating a 2:1 relationship between H₂O and CO₂ in the products of combustion. Natural gas, which has a H/C (and H/C_m) similar to methane, is represented in Table 5 as C1 H3.8.

Depending on the combustor sizing and loading, the products of combustion will move toward chemical equilibrium which considers the first and second law of thermodynamics. Dissociation of the major species mentioned above will occur with trace species concentrations dependent on the chemical kinetics of the fuel-air mixture. The engine performance analysis mentioned above focuses on the total available energy content of the fuel. However, combustor operability issues such as blowout, flashback, combustor instabilities, and auto-ignition can hinder the amount of chemical energy extracted from the fuel. These issues are strongly dependent on the fuel chemistry and properties. Sufficient margin from these combustion phenomena must exist to provide a safe, reliable, and efficient operability window of the combustor. The following sections Hydrocarbon Thermochemistry, Fuel Properties and Handling, and Combustion Characteristics, further show the impact of varying the H/C and overall fuel composition.

The benefits of burning a hydrocarbon fuel with a higher H/C in an industrial gas turbine engine are realized by analyzing the fuel's thermochemistry. The fuel's calorific value, also known as the heat of combustion or fuel heating value, is dictated by the enthalpy contribution from each species. Analyzing a hydrocarbon molecule's main constituents (hydrogen and carbon) on a mass basis, a higher H/C (H/C_m) increases the fuel's heating value due to the greater contribution from hydrogen (LHVH2 = 51,574 Btu/lbm) compared to carbon (LHVC = 14,100 Btu/lbm). A higher H/C also decreases the fuel's mass due to hydrogen's lower molecular mass (MMH2 = 2.016 lbm/lbmol; MMC = 12.011 lbm/lbmol) [6]. Therefore, a heavier hydrocarbon results in a higher molecular mass, hence density, and lower calorific value due to the decrease in H/C.

Figure 8 illustrates the H/C spectrum varying from pure hydrogen (H2) to solid carbon (C(s)). Typical normal alkanes (n-alkanes) are shown in between which follow the trend of decreasing energy content per unit mass as the H/C decreases. For these alkanes, the heating value trends to a near constant value for molecules larger than n-Decane (C10H22) which has an H/C of 2.2. Natural gas typically consists of light alkanes with methane being the main constituent. Branched-chain alkanes (i-alkanes) also exist with their straight-chain/normal counterparts such as n-butane and its isomer i-butane (isobutane). Although these isomers have the same molecular mass due to the same chemical formula, branched-chain isomers have slightly lower heating values than straight-chain alkanes and are more volatile.

It should be noted that the solid carbon energy content presented in Fig. 8 may be an overestimate compared to some classes of coal. Coal not only contains large amounts of oxygen (depending on the coal classification) within carbon and other volatile components, but its macromolecular structure is based on aromatic carbon structures which contain lower heat content and are precursors to soot formation [7]. Asphaltenes are an example of large aromatic hydrocarbons present in crude oil.

The alkene hydrocarbon family (ethylene—C₂H₄, propylene—C₃H₆, butene—C₄H₈) are unsaturated hydrocarbons that do not normally exist in large quantities in crude oil but are typically produced by conversion processes in refineries. Alkenes (olefins) have slightly lower LHVs than alkanes with the same number of carbon atoms. This lower LHV results in lower shaft power as shown in Fig. 4 using ethylene (C₂H₄) on the Titan 250 compared to the other alkanes. Alcohols (methanol—CH₃OH, ethanol—C₂H₅OH, propanol—C₃H₇OH) are viewed as alternative fuels that can boost octane ratings in gasoline engines and potentially reduce combustion particulates due to the added oxygen content. The synthesis of alcohols, especially ethanol, stemmed from the need to replace methyl tertiary butyl ether in gasoline grades due to its health and environmental impacts. However, alcohols have high vapor pressures compared to diesel fuels which can promote vapor lock and are highly corrosive [8]. Naphthenes (cyclopropane—C₃H₆, cyclopentane—C₅H₁₀) are saturated hydrocarbons that are less volatile and contain lower specific energies than alkanes of similar carbon numbers. Aromatic hydrocarbons (benzene—C₆H₆) are ring compounds that are typically undesirable in gas turbine engines due to their marked tendency to soot formation, low heating value, and exhibit strong solvent action on rubber that can cause trouble in fuel systems [9].

Although optimizing the energy content of a fuel on a mass basis is beneficial for applications involving continuous fuel flow such as industrial gas turbine engines, it can also be analyzed on a volumetric basis to maximize the energy content stored in a fixed volume such as fuel tanks on aircraft. Heavier hydrocarbons have higher energy densities (energy content per unit volume), while lighter hydrocarbons have higher specific energies (energy content per unit mass). These different design constraints allow for fuel chemistry optimization such as the substitution of heavier alkanes with naphthenes (cyclo-alkanes) to maximize the energy density due to their higher densities. Fuels rich in naphthenes are sometimes called high-density fuels. The H/C variation in fuels used in industry is a result of the mixture's extraction and refinement processes based on its gas and/or petroleum sources which heavily dictates its chemistry. In addition to its impact on engine performance as mentioned above, a variation of fuel H/C will also affect its physical properties and its combustion chemistry resulting in an impact on fuel handling and combustion characteristics.

In a gas turbine combustion chamber, the fuel must be injected, vaporized, and mixed with air prior to combustion. The physical properties of the fuel play a crucial role in these processes and largely dictate combustion performance. Therefore, one needs to consider the changes in fuel properties when investigating the H/C effects on engine performance. Considering the typical straight-chain alkane hydrocarbons mentioned above, the light molecules down to n-butane (C₄H₁₀) exist as a gas at standard ambient conditions. Heavier hydrocarbons will exist as liquids, and eventually reach the solid phase as the molecular mass continues to increase. This change of phase at standard ambient conditions is due to the greater amount of kinetic energy needed to remain in gas or liquid phase compared to intermolecular forces. Therefore, the phase of the fuel needs to be considered to understand its proper handling and treatment prior to combustion.

Table 5 provides a comparison of approximate fuel properties for gaseous fuels commonly used in industrial gas turbines that are blends of hydrocarbons, diluents, and other energetics such as hydrogen (H₂) and carbon monoxide (CO). The reader should note that these approximate values can vary based on the specific blend and are shown for comparison purposes. The most common gaseous fuel for industrial gas turbines is natural gas. However, global interests in alternative energy and energy storage efforts has led to the increase in interest of gasified biofuels, synthetic gas blends, and by-product gases such as coke-oven gas (COG) and blast furnace gas which can be sourced from steel production. COG is released during coke production that feeds into a blast furnace that produces blast furnace gas and pig iron. The tabulated gasified biomass shown in Table 5 is a synthetic gas (syngas) blend that can be produced by gasification of carbon-containing wastes. Because of their lower H/C compared to conventional petroleum and natural gas, these lower-WI fuels with high levels of inerts and H₂-CO content can also be treated to increase the H/C through the application of the Fischer–Tropsch process. This process can be divided into low-temperature and high-temperature conversion processes to produce high molecular-weight and low molecular-weight hydrocarbons, respectively. These conversion processes are ideal in areas with an abundance of coal or biomass-based fuels where they can be converted to higher energy-content transportation fuels like diesel fuels and gasoline.

Although gaseous fuels are advantageous in terms of high thermal stability, they often require treatment to remove particulate matter and sulfur compounds to reduce corrosion rates and maximize turbine life. Natural gas can be too “sour” due to high hydrogen sulfide (H₂S) content that requires removal prior to combustion. Other impurities such as water and liquid hydrocarbons can also be present in gaseous fuels depending on its conditions. The wide energy density range across typical gaseous fuels can create challenges to the turbine control system, fuel handling equipment and combustion hardware due to their large variation in volumetric flow rates.

Liquid fuels demand more stringent conditioning due to their wider range in hydrocarbons and particulates, large variation in physical properties and the nature of heterogeneous combustion. The liquid fuel's chemical and physical properties strongly affect the combustion process through competing effects between the fuel's evaporation, mixing and reaction rates. Depending on the quality of atomization and mixing, the maximum rate of heat release and thus combustion efficiency is governed by any one of these three mechanisms [9]. Fuel properties such as viscosity, density, and surface tension play a major role in spray combustion. The viscosity not only affects the power required to pump the fuel through the fuel system but also affects its atomization and droplet evaporation. The higher the viscosity of a fuel, the poorer the quality of atomization which can lead to soot formation. Carbon deposits within the combustion system can damage hardware by high thermal radiation and clogging. Liquid fuel systems may also require separate air atomizing systems during ignition depending on the type of atomizer used in the fuel injector. The relative density (specific gravity) of a fuel is related to its boiling point and chemical composition and is a good indicator of the mixture's H/C, heating value and tendency to form carbon deposits.

Volatility characteristics depend on its distillation range, vapor pressure and flash point. High-volatile fuels can provide better ignition and stability characteristics but can also promote vapor lock. Low-volatile fuels such as heavy fuels usually are too viscous for injection thus requiring heating for acceptable viscosity levels as well as prevention of wax crystal formation, which is reflected in higher freezing points. The presence of iso-paraffin hydrocarbons in lieu of their straight-chain counterparts can decrease the freezing point [7]. When using heavy fuels in gas turbines, a distillate oil with a higher American Petroleum Institute (API) gravity (lower relative density and viscosity) such as a light diesel blend (No. 2 Distillate, also known as DF-2) is typically used for engine startup and shutdown sequences to avoid coke formation in fuel injectors.

Liquid fuel treatment processes are used to remove particulates such as vanadium and lead, depending on the fuel's contaminant criteria, as well as preventing emulsions. Sodium, potassium, and calcium can also be present in the form of seawater that may have resulted from compressor ingestion near ocean environments, salty wells or transportation over seawater, which can react with fuel-bound sulfur to form sodium sulfate (Na₂SO₄). These contaminants can also react with engine components resulting in loss or deterioration of material through hot corrosion leading to engine degradation [4]. Therefore, it is crucial to implement appropriate fuel-treatment processes to reduce engine degradation. Liquid fuel treatment processes include fuel washing systems or vaporized fuel oil systems configured to treat contaminated fuel [10].

Table 6 compares typical properties of common liquid fuels including methanol and ethanol. The liquid fuel properties presented are also an approximation due to the possible variation in fuel composition depending on the fuel application. This is demonstrated in Fig. 9 which illustrates an example of a gas chromatograph reading for a typical kerosene blend showing the dominating carbon lengths. For example, aviation gasoline (Avgas) contains less aromatics than automobile gasoline to minimize fuel effects on elastomers and to provide a higher heating value. Kerosene blends (Avtur) can be tailored for a variety of jet-fuel applications such as commercial aviation (Jet A/A-1), military aviation (JP-8) and Navy-marine applications (JP-5) which contain different additives to reduce freezing point temperature, reduce volatility for safety, or improve performance in harsh environments.

Heavy residual fuels are the remainder of the crude oil distillation process and contain more particulate matter than distilled blends. They are generally classified based on API gravity or viscosity and cover a wide range of heavy oils and residuals which are used to produce asphalt. The H/C and specific energy do not change significantly when comparing gasolines to light distillate oils. Figure 10 shows the specific energy range for the typical liquid hydrocarbon classifications.

Focusing on the lower range of the H/C spectrum for reference, Table 7 compares the lower heating value between pure carbon graphite form, typical wood and three coal classifications. Lignite coal is the lowest rank while anthracite coal is the highest rank of coal. Along carbon, wood and coal have other constituents depending on the source. Only a small range of properties are shown for wood, but it should be noted that broad range in wood properties exist based on wood type and configuration such as wood pellets or wood chips. These properties and composition are approximations that depend on the classification and region where it was produce, and are shown for comparison purposes. Although the H/C and heating value seem inversely proportional for the three coal classifications, the oxygen content decreases with increasing rank (anthracite coal is less oxidized than lignite and bituminous). Compared to gaseous and liquid fuels, wood and coal have substantially lower heating values and promote the highest carbon-based emissions. Like liquid fuels, the nature of heterogeneous combustion also adds challenges with solid-fuel combustion such as pulverization technologies, storage and soot radiation. Aluminum and magnesium were also added to compare heating values and energy density with the types of fuels discussed above. Although gas turbines do not currently run on metal fuels, it is interesting to note that their energy density surpasses conventional gaseous and liquid fuels and can potentially be considered as zero-carbon energy carriers containing the highest volumetric heat production when burned in air [14].

The gas turbine combustor needs to provide a safe, reliable and efficient operation across all engine conditions including transient events such as large load changes, startup, and shutdown. Combustor operability issues that arise from the combustion reactions and fuel–air interactions can inhibit engine operability. These issues involve auto-ignition, positioning the flame at a stable condition against flashback and blowout, and emission characteristics. The influence of fuel composition on these operability issues stem from the fuel's unique flammability characteristics such as its flammability limits, ignition behavior, chemical kinetics, and burning rate. Therefore, the fuel constituents and properties need to be understood to properly design and operate the gas turbine combustor to maximize engine operability.

A fuel–air mixture's flammability characteristics involve its flammability limits, ignitability limits, and spontaneous ignition. Understanding these characteristics is crucial in designing fuel injectors and determining combustor residence times especially for premixed systems. From a safety standpoint, it is also crucial to understand the mixture's flammability and ignition limits to avoid hot surface ignition from anywhere within the gas turbine package system. Kurz et al. [15] studied gas turbine safety in offshore operations where they analyzed possible failure scenarios from potential gas leaks leading to explosions due to auto-ignition.

Combustible mixtures are flammable within a limited range of composition, known as the flammability limits. Zabetakis [16] provides a clear illustration of the effects of composition and temperature on the mixture's flammability limits bounded by the mixture's lower flammability limit (LFL) and the upper flammability limit. In a gas turbine engine, the lower flammability limit is of primary concern since combustor operation is typically at global equivalence ratios less than unity. Figure 11 shows the effect of temperature on the LFL for various alkane hydrocarbons, and demonstrates an improvement of the lower limit with increasing molecular weight. Although not plotted in Fig. 11, the flammability limits are expanded significantly when adding hydrogen to the mixture which directly translates to an increase in the lean stability limit of a fuel–air mixture. This poses two key observations for H2 operation: (i) the engine will be able to operate at a lower air–fuel ratio before flameout and (ii) a greater auto-ignition and explosive risk is introduced with H₂ mixtures due to its widened flammability. The second observation is critical to safety and directly relates to the study done by Kurz et al. [15] because the lighter/smaller H₂ molecule has a greater tendency to diffuse through crevices that can potentially lead to a gas-leak explosion.

The LFL of a fuel–air mixture is essential in determining the available operability margin of a combustion system. Designing and controlling gas turbine engines to operate with sufficient margin from the flammability limit will eliminate instabilities leading up to a flame extinction. These instabilities include a combination or separate set of events due to incomplete combustion and thermoacoustic instabilities fed from local flame extinction–reignition phenomenon. In addition to providing a flammable fuel–air mixture, the flame must also be stabilized within the highly turbulent flow field of the engine which can promote blowout (lean extinction at LFL). A combustor's blowout characteristics are a function of the fuel's flammability range and the fuel's diffusive characteristics in addition to the turbulence intensity of the flow field.

The auto-ignition, or spontaneous ignition, characteristics of a fuel–air mixture are governed by its chemical kinetics which is directly related to the fuel composition. It is usually expressed as the minimum auto-ignition temperature, which is the lowest temperature at which a gas fuel or liquid fuel vapor can spontaneously ignite in air without the introduction of an ignition source at a given pressure. It can also be expressed as the ignition delay time, or auto-ignition delay time (AIDT), because it relates to the timescales of the combustion system. Lieuwen et al. [17] define AIDT as the time it takes for a fuel–oxidizer mixture to begin reaction at a given temperature and pressure. The AIDT of a mixture influences the amount of spark energy and fuel–air ratio needed for ignition, and is also a good measure when analyzing dormant passages within the fuel injector such as in dual-fuel applications. Figure 12 shows the effect of increasing the carbon content, or carbon chain length, of the hydrocarbon molecule on the minimum auto-ignition temperature. For heavier hydrocarbons, the flash point and auto-ignition temperature are used together to safely handle and reliably ignite the fuel–air mixtures as shown in Fig. 13 [18]. The auto-ignition characteristics need to be evaluated for each fuel since the chemical kinetic pathways depend on the constituents of the blend. There are numerous research groups that have focused on studying combustion chemistry parameters such as the AIDT using shock tubes and rapid-compression machines. Spectroscopy and other methods are applied to study the reaction pathways which are used to model global kinetic information such AIDT [19].

Lieuwen et al. [17] explain how fuel composition has a profound influence on auto-ignition, blowout, flashback, and combustion instability that stem from large variations in chemical time scales and flame propagation speeds between existing fuels even with similar heating values. Understanding these combustion parameters for each fuel candidate is crucial in ensuring proper flame stabilization. In contrast to a blowout where the flame is physically blown out of the combustor, a flashback occurs when the flame propagates in the upstream direction into the fuel injector's premixing passages. This can damage fuel injector components that are not designed to handle such high temperatures. This condition happens when the turbulent flame speed of the fuel–air mixture exceeds the flow velocity along a streamline. The turbulent flame speed of a mixture is a function of its laminar flame speed and turbulence intensity. Large differences in component diffusivities of the fuel can alter flame stretch effects as well as the local laminar flame speed. Hydrogen is a good example where its addition greatly alters combustion properties. As mentioned above, H₂ extends the mixture flammability and extinction stretch rate, but also introduces significantly different burning rates compared to other gases which make H2 mixtures more sensitive to changes in flow conditions. Even if local flow conditions within a boundary layer in a fuel injector can prevent flashback for conventional hydrocarbon fuel blends, H2 can exacerbate flashback propensity due to its dominating flame speed and transport properties and its effect on flame position. Therefore, the mixture's diffusive properties, which are dictated by the fuel composition, need to be considered in addition to chemical kinetic rates and flame propagation speeds.

Another major challenge in the design of combustion systems comes from combustion instabilities which are characterized by high-amplitude pressure oscillations. These pressure oscillations can be high and frequent enough to cause hardware damage. Oscillations from either the heat release rate, combustor acoustics or upstream flow field can couple to form a thermoacoustic feedback cycle. The upstream flow/mixture oscillations can potentially occur from either air fluctuations from the compressor, fuel fluctuations from the fuel system and injection, or a combination of both. These fuel–air oscillations directly affect the heat release rate downstream, and can potentially resonate natural frequencies of the combustor which exist in the form of longitudinal and transverse (radial and azimuthal) modes. These instabilities can range from low (<50 Hz) to high (>1000 Hz) frequencies characterized by different combustor modes, and can even cause flashback. Interaction between a combustor's acoustic modes and heat release oscillations adds or removes energy from the acoustic mode. Oscillations can be damped by numerous factors such as viscosity, heat transfer and sound radiation to name a few. Variation in fuel composition can affect the thermoacoustic feedback cycle by impacting the phase angle relationship between the heat release rate and pressure oscillations through both the convective and chemical time scales [20]. As explained earlier, the chemical time scale is a function of the mixture properties and kinetics. The convective time scale represents a residence time between the point of origin of the disturbance and flame location which depends on the fuel–air mixture's local flame speed. Because thermoacoustic instabilities are non-monotonic and depend on combustor geometry, there is no direct correlation between these instabilities and H/C.

The combustor operability issues discussed above can inhibit engine operability by limiting the region of stable, reliable and efficient burning. The location of these combustion phenomena and limits with respect to the design point can change depending on the engine conditions and fuel composition, and it is up to the designer to select appropriate reactant control measures to ensure operation with sufficient margin. It is also up to the operator to stay within the limits specified by the gas turbine OEM to avoid these operability issues to maximize engine life, performance, and minimize maintenance/repair downtime.

The system effects of fuel thermochemistry are evident in other related fossil fuel burning application studies. Alternative bio-fuels were examined by Rubie et al. [21] considering aviation applications. The study considered Algae, Camelina paraffinic kerosene (CSPK), and Jatropha paraffinic kerosene (JSPK) biofuels. All of which are close in molecular weight and H/C to jet A which they are intended to emulate. The study describes the creation of these synthesized fuels from feedstocks of algae, camelina, jatropha, and animal tallow. These fuels were used to “drop in” and supplement jet A in increasing mixed ratios. Performance differences were assessed considering simulation of a F404-GE-400 engine controlling to a fixed gas generator speed. The three different fuels showed only subtle changes in output performance, indicating all fuels as viable alternatives to jet A. In comparison, an industrial engine, as depicted in this paper, is more accepting of utilizing a wide variety of fuels since the engine is ground based, can fluctuate available power for most applications, and does not have the constraints of an aircraft mission profile. Bae and Kim [22] reviewed potential alternative fuels for automobile engine applications for both spark ignition and compression ignition engines. The fuels included compressed natural gas, hydrogen, liquefied petroleum gas and alcohol fuels for spark ignition engines, and biodiesel, di-methyl ether, and jet propellant-8 (JP-8) for compression ignition engines. These fuels were evaluated for their combustion properties such as octane and cetane number, physical properties that influenced spray/mixture formation for combustion, lower heating value, and engine compatibility. They emphasized the collaboration between the automotive engine manufacturers and the refinery industry toward more efficient and clean combustion engines. This is relatively similar to industrial gas turbine development where the engines and combustion systems are designed for greater fuel flexibility to accommodate by-product fuels and other low-quality fuels. For example, Gökalp and Lebas [23] emphasizes how industrial gases, which are defined as rejection from refineries or other chemical industrial processes, are great fuel candidates for industrial gas turbine engines. They also investigated other alternative fuels such as biogas and esters of vegetable oils, and evaluated them based on their physical/chemical characteristics.

The objective of the paper was to edify the reader about the systemic effects the fuel composition has on an industrial gas turbine engine. H/C is a distinctive attribute of the hydrocarbon fuel that as its value increases the fuel becomes lighter with increasing specific energy, resulting in higher output power and efficiency of the gas turbine when utilizing the fuel. However, the greater the H₂ proportion of the fuel leads to more pronounced H₂-combustion effects such as higher reactivity and larger differences in diffusivities between fuel components. On the other side of the spectrum, fuels with lower H/C tend to have lower ignition temperatures and added design and operational challenges such as stringent fuel treatment processes to remove particulates and providing atomization for efficient spray combustion. Therefore, the physical and chemical properties of the fuel blend must be considered when evaluating engine performance and operability.

When considering carbon emissions, it is easy to see burning wood comprising about ten carbon atoms for every one hydrogen atom or a 1:10 H/C will release much more carbon into the environment than methane at 4:1 H/C. Mankind has slowly moved toward decarbonizing the world's energy mix from millennia of burning wood, to burning coal in the last few centuries, then transitioning more to oil in the last century. More recently natural gas appears to be the bigger winner in fossil fuels as renewables, energy storage, and the H₂ economy start to take hold in this century.

The Authors would like to acknowledge the contributions of Dustin Truesdell, who provided performance data and reference information, and Priyank Saxena and Gareth Oskam, who provided combustion related advice and guidance.

AIDT =

auto-ignition delay time

C =

carbon–element

Cp =

specific heat in constant pressure

CED =

cold end drive—Solar Turbines engine designation

GPT =

gas producer turbine

H =

hydrogen–element

H/C =

hydrogen–carbon ratio

H/C_m =

hydrogen–carbon mass ratio

HED =

hot end drive—Solar Turbines engine designation

LFL =

lower flammability limit

LHV =

lower heating value

m =

mass flow

MM =

molecular mass

O =

oxygen–element

PT =

power turbine

SG =

specific gravity

T =

temperature

TRIT =

turbine rotor inlet temperature

W =

turbine power

WI =

Wobbe index

X =

# of moles–carbon

Y =

# of moles–hydrogen

Z =

# of moles–air

m =

mass