|Ph.D Student||Patrick Arfi|
|Subject||Reduction of NOx Emissions from Gas Turbines Using Internal|
Exhaust Gas Recirculation
|Department||Department of Aerospace Engineering||Supervisor||Professor Emeritus Levy Yeshayahou|
Pollutant emissions from combustion systems have created increasing environmental concerns. Industrial gas turbine and especially aircraft engine emissions have potentially many different climatic effects such as atmospheric ozone production and destruction, changes in composition of greenhouse gases, alterations of cloud properties and coverage, etc. The main emitted pollutants in terms of environmental impacts are the oxides of nitrogen collectively named NOx. Throughout the troposphere and the lower stratosphere, where most of the subsonic aircraft fleet flies, NOx contribute to photochemical smog and acid rain, global warming as well as ozone formation. Stratospheric studies indicate that in the upper-stratosphere, where future hypersonic transports will be required to cruise, NOx may actively deplete ozone. To enhance thermal efficiency and therefore allow fuel savings and CO2 emission reductions, future gas turbine engines will utilize either air preheating or higher engine pressure ratios with consequently higher combustor temperatures. In conventional combustion chamber design this has an enormous effect on NOx formation. Hence, in order to overcome the conflict of interests between energy saving and pollutant emission reductions, alternative combustion concepts are needed. New modes of combustion have been recently introduced in gas turbines, including lean premixed combustion, staged combustion, catalytic combustion and rich-quench-lean combustion. The present study provides a critical review of the pros and cons of the current advanced combustion methods and technologies. It also focuses on a new promising combustion method, the flameless oxidation (FLOX). FLOX is based on large recirculation of hot combustion products allowing stable combustion in vitiated air. FLOX is characterized by a moderate and distributed temperature rise, no temperature peaks and hence allows low-NOx levels without impairing flame stability. The present investigation is concerned with its innovative application to an adiabatic combustor used in gas turbine engines. A new combustor design is proposed and the research and development plans outlined. Since combustion of fuel injected into hot combustion products of low oxygen content is a rather unknown area, FLOX is theoretically investigated. Emphasis is given to studying combustion chemistry of the fuel in air diluted with large quantities of hot combustion products and establishing stability limits of FLOX. Of special interest are the identifications of the shortcomings and improvements of Computational Fluid Dynamics (CFD) combustion models required to correctly predict characteristics and limitations of FLOX in practical combustion systems. Another objective is to determine the validity of the NOx emission models with regard to their applicability to this new combustion technology.
A numerical investigation in an ideal Perfectly Stirred Reactor (PSR), which is representative of the mixing patterns within the reacting fine structure of any turbulent flame and of the distributed turbulence-chemistry regime met in FLOX, was performed. It highlights the combined chemical and thermal effects of exhaust gas recirculation on flame stability and NOx formation. Parameters affecting the quenching time and especially influences of gas inlet temperature and composition are discussed. The limitations of global mechanisms and performance of reduced mechanisms used in CFD were studied under adiabatic and isothermal PSR conditions with regard to their capacity in reproducing finite-rate kinetic effects such as the local extinction phenomenon. Ignition and extinction patterns under conditions characteristics of FLOX were also studied to gain a better understanding of FLOX fundamentals. The numerical modeling of NOx emissions was also investigated. It was shown by PSR studies that proper predictions of NOx formation in advanced combustors requires sophisticated modeling approach including a detailed chemistry approach. It was demonstrated that the main points in emission modeling are that assumptions used in combustion models and lack of knowledge on intermediate species concentrations are responsible for the strong inaccuracy noticed with NOx emission models. Therefore improvement of the temperature predictions by accounting for extinction should remain the prime objective.
In the CFD investigation, emphasis was given to the turbulence-chemistry interaction in order to account for finite-rate kinetic effects, which were recognized as the main current limit of the development of combustors based on FLOX but also flue gas recirculation or fuel-staged combustion. A review of the current knowledge on turbulence-chemistry interactions and their modeling allowed obtaining a better physical insight of flame stretched induced extinction. It also allowed proposing and formulating a procedure for the computationally efficient simulation of extinction in turbulent flows accounting for vitiation effects. The extinction criterion is based on the comparison of the turbulent time scales giving the lifetime of the smallest eddies that can quench the flame, and the critical chemical time scale representing the minimum time necessary for the reaction to occur under well-mixed conditions. When the turbulent time scale characteristic of extinction is chosen to be the Kolmogorov scale, the extinction criterion becomes equivalent to the Klimov-Williams criterion. To account for the gas composition and inlet temperature effects on the chemical quenching time at a reasonable computation cost, a mixture freshness parameter was introduced. The extinction model combined to the Eddy Dissipation Combustion Model (EDCM) was tested against experimental data and other models. Experimental data were taken from the literature for the case of a sudden expansion chamber and a confined co-flow flame piloted by hot combustion products. They were also acquired from a specially built experimental set-up representing a confined co-flow flame stabilized by a bluff-body. The influences of chemistry description and combustion model assumptions were investigated numerically using the CFD tool and temperature and species mass fraction profiles were compared to experimental data. The implemented combustion model accounting for local extinction phenomenon due to flame straining and air vitiation, while far from being perfect, improved significantly temperature predictions by accounting partially for chemical kinetic effects. Among others it allows to predict flame lift-off. Further improvements may be obtained by the inclusion of intermediate species as CO and H2. The CFD study of NOx emission models showed that the use of combustion models based on fast-chemistry assumptions combined with the simplifying assumptions of O-O2 equilibrium for the NO-chemistry as applied in the majority of the current studies could be strongly misleading. Accounting for both local extinction in the combustion model and O-atom partial equilibrium in the emission model allowed noticeable improvements and reproduction of experimental trends. It was also pinpointed that determining absolute values of NOx emissions is not a realistic issue considering current combustion models. Finally, the application of the CFD models to a real practical combustor geometry and operating conditions illustrated the impact of the local extinction model on combustion performance predictions.
Theoretically, further research work directions are suggested. It includes implementation of reduced chemistry and transient effects in the CFD. From an engineering viewpoint, recommendations for further development of the FLOX combustor (FLOX-GT) were given. It was noticed that special care should be taken on the relative position of air and fuel inlet ports to optimize the design. By using the gained knowledge on combustion in highly preheated oxidant of low oxygen content and on NOx formation in combination with the theoretical investigation and the validated numerical tools, one can hopefully design more efficiently combustors with the required parameters for simultaneously optimizing low-NOx emissions, combustion efficiency and flame stability.