Fuel Interchangeability for Lean Premixed Combustion in Gas Turbine Engines

In response to environmental concerns of NOx emissions, gas turbine manufacturers have developed engines that operate under lean, pre-mixed fuel and air conditions. While this has proven to reduce NOx emissions by lowering peak flame temperatures, it is not without its limitations as engines utilizing this technology are more susceptible to combustion dynamics. Although dependent on a number of mechanisms, changes in fuel composition can alter the dynamic response of a given combustion system. This is of particular interest as increases in demand of domestic natural gas have fueled efforts to utilize alternatives such as coal derived syngas, imported liquefied natural gas and hydrogen or hydrogen augmented fuels. However, prior to changing the fuel supply end-users need to understand how their system will respond. A variety of historical parameters have been utilized to determine fuel interchangeability such as Wobbe and Weaver Indices, however these parameters were never optimized for today’s engines operating under lean pre-mixed combustion. This paper provides a discussion of currently available parameters to describe fuel interchangeability. Through the analysis of the dynamic response of a lab-scale Rijke tube combustor operating on various fuel blends, it is shown that commonly used indices are inadequate for describing combustion specific phenomena.

Chemical Kinetic Analysis of a Flameless Gas Turbine Combustor

Flameless combustion (FC) is one of the most promising techniques of reducing harmful emissions from combustion systems. FC is a combustion phenomenon that takes place at low O2 concentration and high inlet reactant temperature. This unique combination results in a distributed combustion regime with a lower adiabatic flame temperature. The paper focuses on investigating the chemical kinetics of an prototype combustion chamber built at the university of Cincinnati with an aim of establishing flameless regime and demonstrating the applicability of FC to gas turbine engines. A Chemical reactor model (CRM) has been built for emulating the reactions within the combustor. The entire combustion chamber has been divided into appropriate number of Perfectly Stirred Reactors (PSRs) and Plug Flow Reactors (PFRs). The interconnections between these reactors and the residence times of these reactors are based on the PIV studies of the combustor flow field. The CRM model has then been used to predict the combustor emission profile for various equivalence ratios. The results obtained from CRM model show that the emission from the combustor are quite less at low equivalence ratios and have been found to be in reasonable agreement with experimental observations. The chemical kinetic analysis gives an insight on the role of vitiated combustion gases in suppressing the formation of pollutants within the combustion process.

Towards Improved Prediction of Aero-Engine Combustor Performance Using Large Eddy Simulations

The paper presents an assessment of large eddy simulation (LES) and conventional Reynolds averaged methods (RANS) for predicting aero-engine gas turbine combustor performance. The performance characteristic that is examined in detail is the radial burner outlet temperature (BOT) or fuel-air ratio profile. Several different combustor configurations, with variations in airflows, geometries, hole patterns and operating conditions are analyzed with both LES and RANS methods. It is seen that LES consistently produces a better match to radial profile as compared to RANS. To assess the predictive capability of LES as a design tool, pretest predictions of radial profile for a combustor configuration are also presented. Overall, the work presented indicates that LES is a more accurate tool and can be used with confidence to guide combustor design. This work is the first systematic assessment of LES versus RANS on industry-relevant aero-engine gas turbine combustors.

Review and Analysis of Bluff Body Flame Stabilization Data

This paper compiles and analyzes bluff body stabilized flame blowoff data from the literature. Many of these studies contain semi-empirical blowoff correlations that are, in essence, Damko¨hler number correlations of their data. This paper re-analyzes these data, utilizing various Damko¨hler number correlations based upon detailed kinetic modeling for determining chemical time scales. While the results from this compilation are similar to that deduced from many earlier studies, it demonstrates that a rather comprehensive data set taken over a large range of conditions can be correlated from “first-principles” based calculations that do not rely on empirical fits or adjustable constants (e.g., global activation energy or pressure exponents). The paper then discusses the implications of these results on understanding of blowoff. Near blowoff flames experience local extinction of the flame sheet, manifested as “holes” that form and convect downstream. However, local extinction is distinct from blowoff — in fact, under certain conditions the flame can apparently persist indefinitely with certain levels of local extinction. We hypothesize that simple Damko¨hler number correlations contain the essential physics describing this first stage of blowoff; i.e., they are correlations for the conditions where local extinction on the flame begins, but do not fundamentally describe the ultimate blowoff condition itself. However, such correlations are reasonably successful in correlating blowoff limits because the ultimate blowoff event appears to be correlated to some extent to the onset of this first stage.

Experimental Study of Combustion on a Micro Gas Turbine Combustor

A research program is in development in China as a demonstrator of combined cooling, heating and power system (CCHP). In this program, a micro gas turbine with net electrical output around 100kW is designed and developed. The combustor is designed for natural gas operation and oil fuel operation, respectively. In this paper, a prototype can combustor for the oil fuel was studied by the experiments. In this paper, the combustor was tested using the ambient pressure combustor test facility. The sensors were equipped to measure the combustion performance; the exhaust gas was sampled and analyzed by a gas analyzer device. From the tests and experiments, combustion efficiency, pattern factor at the exit, the surface temperature profile of the outer liner wall, the total pressure loss factor of the combustion chamber with and without burning, and the pollutants emission fraction at the combustor exit were obtained. It is also found that with increasing of the inlet temperature, the combustion efficiency and the total pressure loss factor increased, while the exit pattern factor coefficient reduced. The emissions of CO and unburned hydrogen carbon (UHC) significantly reduced, but the emission of NOx significantly increased.

Ignition Characteristics of a Fischer-Tropsch Synthetic Jet Fuel

Ignition delay times of a “real” synthetic jet fuel (S8) were measured using an atmospheric pressure flow reactor facility. Experiments were performed between 900 K and 1200 K at equivalence ratios from 0.5 to 1.5. Ignition delay time measurements were also performed with JP8 fuel for comparison. Liquid fuel was prevaporized to gaseous form in a preheated nitrogen environment before mixing with air in the premixing section, located at the entrance to the test section of the flow reactor. The experimental data show shorter ignition delay times for S8 fuel than for JP8 due to the absence of aromatic components in S8 fuel. However, the ignition delay time measurements indicate higher overall activation energy for S8 fuel than for JP8. A detailed surrogate kinetic model for S8 was developed by validating against the ignition delay times obtained in the present work. The chemical composition of S8 used in the experiments consisted of 99.7 vol% paraffins of which approximately 80 vol% was iso-paraffins and 20% n-paraffins. The detailed kinetic mechanism developed in the current work included n-decane and iso-octane as the surrogate components to model ignition characteristics of synthetic jet fuels. The detailed surrogate kinetic model has approximately 700 species and 2000 reactions. This kinetic mechanism represents a five-component surrogate mixture to model generic kerosene-type jets fuels, namely, n-decane (for n-paraffins), iso-octane (for iso-paraffins), n-propylcyclohexane (for naphthenes), n-propylbenzene (for aromatics) and decene (for olefins). The sensitivity of iso-paraffins on jet fuel ignition delay times was investigated using the detailed kinetic model. The amount of iso-paraffins present in the jet fuel has little effect on the ignition delay times in the high temperature oxidation regime. However, the presence of iso-paraffins in synthetic jet fuels can increase the ignition delay times by two orders of magnitude in the negative temperature (NTC) region between 700 K and 900 K, typical gas turbine conditions. This feature can have a favorable impact on preventing flashback caused by the premature autoignition of liquid fuels in lean premixed prevaporized (LPP) combustion systems.

Computational Fluid Dynamic Prediction of Noise From a Cold Turbulent Propane Jet

Numerical solutions of a turbulent jet flow are used to provide velocity information throughout a simple cold turbulent propane jet at a Reynolds number of 68,000. Predictions provided by the Reynolds-averaged Navier-Stokes simulations, based on a Reynolds stress turbulence model, are compared with experimental data available in the literature. The effect of the modelled inlet boundary conditions on the predicted flow field is described, and the discrepancy between the simulation results and experiment measurements is found to be less than the corresponding variations due to uncertainness in the experimental boundary conditions. In addition, these solutions are used as the basis for noise predictions for the jet based on Lighthill’s theory using the Goldstein broadband noise source formalization that postulates axisymmetric turbulence superposed on the mean flow. The latter model provides an aeroacoustic tool that is reasonable in identifying components or surfaces that generate significant amounts of noise, thereby providing opportunities for early design changes to aircraft and gas turbine components.

Skeletal Mechanism for C2H4 Combustion With PAH Formation

A semi-detailed kinetic mechanism with 100 species and 816 reactions for ethylene combustion including PAH formation was elaborated. The model includes the C2H5OH sub mechanism combustion as well. This mechanism has in view to be the base of further kinetic schemes of practical fuels (reference fuels). The mechanism was reduced to a skeletal model with 72 species and 580 reactions. The elaborated models were validated on experimental data bases of heat release as well as formation of polyaromatic hydrocarbons and soot in laminar premixed C2H4, C2H4 / C2H5OH flames taken from literature. The calculated ignition delay times, laminar flame speeds, as well as temporal profiles of small and large aromatics and also soot particles are in good agreement with experimental data obtained for pressures 1 – 5 bar, temperatures T0 = 1100 – 2300 K, fuel/oxygen equivalence ratio φ = 0.5 – 2.

Flame Transfer Matrices of a Premixed Flame and a Global Check With Modelling and Experiments

In this paper alternative ways to obtain the thermo-acoustical characteristics of perfectly premixed flames given by their flame transfer matrix (FTM) is investigated. In particular a model based data reduction procedure which greatly reduces the experimental effort and therefore enables to provide this flame data for many more operation points than previously possible is proposed and validated. It is shown how the acoustic pressure field measured from two forcing states using the multi microphone method leads to the determination of the direct experimental FTM. The next relatively simpler method shown is the hybrid method which is based on Rankine-Hugoniot relations and the experimental flame transfer function (FTF) from OH*-chemiluminescence measurements for heat release fluctuations. Later to obtain the FTM using a network model based on Rankine-Hugoniot relations and an n-τ-σ FTF model representing the flame by regression analysis of the acoustical measurements is presented. Experimental results for the direct experimental FTM and the hybrid FTM are compared with the model based result. The results indicate very good consistency between the direct, hybrid and model based techniques providing a global check of the methods/tools used for analysing the thermoacoustic mechanisms of flames.

Noise Generated by a Lifted Flame in a Vitiated Co-Flow

The acoustic field generated by a lifted flame is studied by a hybrid approach. First, Large Eddy Simulations (LES) are used to compute the flow and the acoustic sources. Next, an inhomogeneous wave equation is solved to obtain the resulting acoustic field. The flow computations show good agreement with experimental data. The dominant acoustic sources are found to be located in the ignition region and at the tip of the flame. The acoustic computations revealed the presence of low-frequency waves radiated in the far-field. The shape of the most energetic acoustic modes are identified by POD analysis to be axial modes.