Search

Quantification of aleatoric uncertainties due to the inherent variabilities in operating conditions and fuel composition is essential for designing and improving premixers in dry low-emissions (DLE) combustion systems. Advanced stochastic simulation tools require a large number of evaluations in order to perform this type of uncertainty quantification (UQ) analysis. This task is computationally prohibitive using high-fidelity computational fluid dynamic (CFD) approaches such as large eddy simulation (LES). In this paper, we describe a novel and computationally efficient toolchain for stochastic modeling using minimal input from LES, to perform uncertainty and risk quantification of a DLE system. More specially, high-fidelity LES, chemical reactor network (CRN) model, beta mixture model, Bayesian inference and sequential Monte Carlo (SMC) are integrated into the toolchain. The methodology is applied to a practical premixer of low-emission combustion system with dimethyl ether (DME)/methane–air mixtures to simulate auto-ignition events at different engine conditions. First, the benchmark premixer is simulated using a set of LESs for a methane/air mixture at elevated pressure and temperature conditions. A partitioning approach is employed to generate a set of deterministic chemical reactor network (CRN) models from LES results. These CRN models are then solved at the volume-average conditions and validated by LES results. A mixture modeling approach using the expectation-method of moment (E-MM) is carried out to generate a set of beta mixture models and characterize uncertainties for LES-predicted temperature distributions. These beta mixture models and a normal distribution for DME volume fraction are used to simulate a set of stochastic CRN models. The Bayesian inference approach through SMC method is then implemented on the results of temperature distributions from stochastic CRN models to simulate the probability of auto-ignition in the benchmark premixer. The results present a very satisfactory performance for the stochastic toolchain to compute the auto-ignition propensity for a few events with a particular combination of inlet temperature and DME volume fraction. Characterization of these rare events is computationally prohibitive in the conventional deterministic methods such as high-fidelity LES.

The minimization of auto-ignition risk is critical to the design of premixers of high power aeroderivative gas turbines as an increased use of highly reactive future fuels (for example, hydrogen or higher hydrocarbons) is anticipated. Safety factors based on ignition delays of homogeneous mixtures are generally used to guide the choice of a residence time for a given premixer. However, auto-ignition chemistry under aeroderivative conditions is fast (0.5–2 ms) and can be initiated within typical premixer residence times. The analysis of what takes place in this short period necessarily involves the study of low-temperature auto-ignition precursor chemistry, but precursors can change with fuel and local reactivity. Chemical explosive modes (CEMs) are a natural alternative to study this as they can provide a measure for auto-ignition risk by considering the whole thermochemical state in the framework of an eigenvalue problem. When transport effects are included by coupling the evolution of the chemical explosive modes to turbulence, it is possible to obtain a measure of spatial auto-ignition risk where both chemical (e.g., ignition delay) and aerodynamic (e.g., local residence time) influences are unified. In this article, we describe a method that couples large eddy simulation (LES) to newly developed, reduced auto-ignition chemical kinetics to study auto-ignition precursors in an example premixer representative of real life geometric complexity. A blend of pure methane and di-methyl ether (DME), a common fuel used for experimental auto-ignition studies, was transported using the reduced mechanism (38 species/238 reactions) under engine conditions at increasing levels of DME concentrations until exothermic auto-ignition kernels were formed. The resolution of species profiles was ensured by using a thickened flame model where dynamic thickening was carried out with a flame sensor modified to work with multistage heat release. This paper is outlined as follows: First, a reduced mechanism is constructed and validated for modeling methane as well as DME auto-ignition. Second, sensitivity analysis is used to show the need for chemical explosive modes. Third, the thickened flame model modifications are described and then applied to an example premixer at 25 bar/890 K preheat. The chemical explosive mode analysis closely follows the large thermochemical changes in the premixer as a function of DME concentrations and identifies where the premixer is sensitive and flame anchoring is likely to occur.

Critical slowing down (CSD) is a phenomenon that is common to many complicated dynamical systems as they approach critical transitions/bifurcations. We demonstrate that pressure signals measured during the onset of thermoacoustic instabilities in a gas turbine engine test exhibit evidence of CSD well before the oscillation amplitude increases. CSD was detected through both the variance and the lag-1 auto-regressive coefficient in a rolling window of the pressure signal. Increasing trends in both metrics were quantified using Kendall's τ, and the robustness and statistical significance of the observed increases were confirmed. Changes in the CSD metrics could be detected several seconds prior to changes in the oscillation amplitude. Hence, real-time calculation of these metrics holds promise as early warning signals of impending thermoacoustic instabilities.

Finite-rate chemical effects at gas turbine conditions lead to incomplete combustion and well-known emissions issues. Although a thin flame front is preserved on an average, the instantaneous flame location can vary in thickness and location due to heat losses or imperfect mixing. Postflame phenomena (slow CO oxidation or thermal NO production) can be expected to be significantly influenced by turbulent eddy structures. Since typical gas turbine combustor calculations require insight into flame stabilization as well as pollutant formation, combustion models are required to be sensitive to the instantaneous and local flow conditions. Unfortunately, few models that adequately describe turbulence–chemistry interactions are tractable in the industrial context. A widely used model capable of employing finite-rate chemistry is the eddy dissipation concept (EDC) model of Magnussen. Its application in large eddy simulations (LES) is problematic mainly due to a strong sensitivity to the model constants, which were based on an isotropic cascade analysis in the Reynolds-averaged Navier–Stokes (RANS) context. The objectives of this paper are: (i) to formulate the EDC cascade idea in the context of LES; and (ii) to validate the model using experimental data consisting of velocity (particle image velocimetry (PIV) measurements) and major species (1D Raman measurements), at four axial locations in the near-burner region of a Siemens SGT-100 industrial gas turbine combustor.