An Investigation of Lean Blowout of Gaseous Fuel Alternatives to Natural Gas
Karalus, Megan Frances
MetadataShow full item record
This work examines lean premixed flame stability for multi-component fuel mixtures to support fuel flexibility for industrial combustors. A single Jet Stirred Reactor (JSR), a generic recirculation stabilized combustor, along with gaseous fuels of hydrogen, methane, and hydrogen/methane blends are chosen for the study. Experimental data on blowout are collected and a series of models are used to understand the mechanism of extinction in this recirculation-stabilized flame environment. By studying this more generic combustor, the aim is to develop generalizable results and methodologies for understanding and predicting lean blowout of multicomponent fuels. Experimental data approaching blowout are taken for fuels of pure hydrogen, pure methane, and hydrogen/methane blends in 10% by volume increments. The data relate inlet equivalence ratios to experimentally measured temperatures for each fuel approaching blowout and reveal the final blowout condition for each fuel. These blowout data are obtained by holding the air flow rate constant and decreasing the fuel flow rate until the flame is extinguished. Doing so holds the flow field and turbulence parameters approximately constant as blowout is approached. The reactor is stabilized to lower equivalence ratios and temperatures as the percentage of hydrogen in the fuel increases. In order to gain insight on the mechanism controlling blowout, two dimensional, axisymmetric computational fluid dynamic (CFD) simulations are carried out for the lean premixed combustion of both hydrogen and methane as the fuel. Hydrogen requires only 9 species to fully describe its chemistry. Therefore, the detailed mechanism of Li <italic>et al.</italic> is chosen for the hydrogen simulations. Methane combustion is described by the full GRI-3.0 chemical mechanism with 35 species. To facilitate reasonable computational times a skeletal mechanism of 22 species is developed from GRI-3.0 using the Directed Relation Graph method developed by Lu and Law. The CFD simulations for both hydrogen and methane combustion are run similarly to the experiments. The fuel flow rate is reduced until the CFD model no longer produces a burning solution. Contour plots from the CFD model illustrate the evolution of the flow-field, temperature profiles, and flame structure within the JSR as blowout is approached for both fuels. The modeling suggests that lean blowout in the JSR does not occur in a spatially homogeneous condition, but rather under a zonal structure. Analysis of the models from the perspective of a combusting fluid particle traveling through the jet, into the recirculation zone, and then entraining back into the jet suggests that the blowout condition is dependent on the development of the pool of radicals. The flame remains stable as long as the radical pool develops significantly enough to achieve ignition before the hypothetical combusting fluid particle is re-entrained. As the fuel flow decreases, the induction period increases and the ignition event is pushed further around the recirculation zone. Eventually, the induction period becomes so long that the ignition is incomplete at the point where the recirculating gas is entrained. This threshold leads to overall flame extinction. Two Chemical Reactor Network (CRN) models are developed using the flow field and reaction fields from the detailed CFD models in an attempt to capture the bulk of the physical processes responsible for flame stability. The single Plug Flow Reactor (PFR) model follows the concept of the hypothetical combusting fluid particle and assumes that only convective transport is responsible for stability. This model matches hydrogen blowout well, reproducing the ignition event and the development of the pool of radicals before re-entrainment. While the single PFR model with the UCSD chemical mechanism does predict the blowout temperature across the full range of methane/hydrogen fuel blends well, it fails to adequately predict blowout equivalence ratio for fuels with high methane concentrations. A two PFR model is subsequently developed in which the core jet region (of constant mass flow) exchanges mass with the recirculation region through turbulent diffusive transport. Entrainment of flow by jet action is confined entirely to the recirculation region, represented by the exhaust of the recirculation PFR being convectively re-entrained at its entrance. The two PFR model performs about as well as the single PFR model in predicting blowout for hydrogen in the JSR and shows significant improvement over the single PFR model in both following the experimental data approaching blowout, and predicting the blowout condition for methane. In fact the two PFR model shows good agreement with both equivalence ratio and temperature at blowout across the full range of hydrogen/methane blends. Regardless of the chemical mechanism applied, or whether we consider transport by convection only as in the single PFR model, or transport by both convection and diffusion as in the two PFR model, the story regarding the onset of blowout remains the same and is consistent with that given by CFD as well: the key to the stable operation of the reactor is the ignition event in the recirculation zone, resulting in the development of the radical pool. For pure hydrogen combustion as the fuel flow rate is reduced and the reactor moves towards blowout the destruction of the fuel slows and spreads, and the development of the radical pool moves further around the recirculation zone. The radical pool must develop (i.e. ignition must occur) before re-entrainment or the reactor will extinguish. For methane we similarly see the destruction of methane spread, and the net production of CO, and subsequently the net production of OH move further around the recirculation zone until the re-entrainment of radicals can no longer sustain the combustion. For methane, transport of the CO and radicals through turbulent diffusion appears to be a controlling process in this ignition event. The ignition event for hydrogen, on the other hand, is affected very little by the inclusion of diffusive transport of radicals. This is most likely due to the fact that the breakdown of hydrogen directly produces an H radical that feeds the chain propagating reaction, however the direct breakdown of methane has no such feedback. It is only in the destruction of methane intermediates that the H radical needed to feed the chain propagating reaction is produced.
- Mechanical engineering