Numerical Study of a Microturbine Combustor
Master thesis
Permanent lenke
https://hdl.handle.net/1956/18244Utgivelsesdato
2018-06-21Metadata
Vis full innførselSamlinger
- Geophysical Institute [1222]
Sammendrag
Microturbines are small-scale gas turbine engines with a power range of 20 to 500 kW, typically used for off-grid power generation or mechanical drive scenarios. The technology also shows potential in combination with high-temperature fuel cell technology to provide high chemical to electrical energy conversion efficiency with very low emissions. A core component in the microturbine is the combustor where the fuel is oxidised with compressed air, releasing its chemical energy. A cost-efficient combustion chamber is needed that will burn the fuel with high efficiency and low formation of pollutants such as Carbon Monoxide (CO) and Nitrogen Oxides (NOx). The Department of Mechanical and Marine Engineering at Western Norway University of Applied Sciences has a microturbine that is used for demonstration purposes. The combustor is a simple can-type combustor fuelled by propane. The combustor is a very simple design with a straight perforated flame tube. It is assumed that this design can be improved with regards to combustion efficiency and lower emissions. This thesis contains numerical simulations of the selected microturbine combustor using a Computational Fluid Dynamics tool. The numerical algorithm solves the Navier-Stokes transport equations iteratively in the context of Reynolds-averaged Navier-Stokes paradigm. The K-Epsilon model is incorporated into the algorithm to account for the turbulence effects. The mean reaction rate term in the averaged species transport equations are closed using the Steady Laminar Flamelet model. The computational costs of this method are lower compared to other more complex approaches. The geometry of the investigated combustor mimics the non-premixed combustion mode where the fuel and oxidiser streams are segregated. The first numerical model represents the exact geometry of the installed combustor. The initial and boundary conditions are chosen from experimental data. Experimental data is also used for model validation. The results from the numerical simulation are provided for scalars such as temperature, velocity, unburnt fuel, CO and NOx. The results show that the selected combustor is a compromised design with excessive temperatures, incomplete combustion and high CO emissions. The second numerical computations represent a modified design of the first case, with the same initial and boundary conditions. The proposed model has a larger diameter with a complex geometry. A swirler device is introduced to improve the mixing between the fuel and the oxidiser, to enhance the combustion and to reduce emissions. The computed scalars of interest such as major and minor species and mean temperature obtained from the modified combustor are compared with the original combustor. This proposed model requires further validation with laboratory measurements. The model can also serve as a basis for further design improvements before a prototype is built.