Study of Vitiated Turbulent Combustion for Low-Emission High-Efficiency Hybrid Energy Systems

Lead Research Organisation: University of Southampton
Department Name: Faculty of Engineering & the Environment

Abstract

The search for zero- or low-emission, high-efficiency energy systems is becoming increasingly important and urgent, as energy security and sustainable development have become one of the top priorities of the 21st century. An emerging technology combining a solid oxide fuel cell (SOFC) and a gas turbine (GT) systems promises to dramatically increase the overall energy efficiency and significantly reduce the level of harmful emissions. Moreover, a lower overall cost can be achieved. Equally importantly, as an SOFC system operates within a high temperature range, a wide variety of hydrocarbon fuels can be utilized directly without pre-reforming, increasing the life-cycle efficiency and versatility of such a hybrid system. In a hybrid SOFC-GT system, the unspent exhaust fuel and high-grade heat from an SOFC system are utilized in the GT combustor, often in combination with a fresh stream of fuel-oxidizer mixture. The main challenge is to burn the exhaust fuel efficiently, when the reactant mixture is highly diluted with water steam (H2O) and carbon dioxide (CO2). As the mixture composition varies both spatially and temporally, combustion can take place in non-premixed and premixed modes in various parts of the combustor. It is also likely to have a mixed mode, called triple flames or edge flames, which exhibits features in between non-premixed and premixed modes. Local flame extinction and auto-ignition are expected to be significant. Such a scenario presents theoretical and modelling challenges, as well as difficulties for the design of the combustor. Diluted combustion under turbulent conditions similar to those in a hybrid SOFC-GT system must be systematically studied to put the construction of the next generation low-emission high-efficiency power systems on a firmer scientific foundation.Given the complexity of the complete problem under concern, the proposed research will focus on generic, fundamental issues related to turbulent combustion in the GT combustor. Direct numerical simulation (DNS), theoretical modelling and analysis will be conducted on a vitiated methane-air flame with a coflow. This base configuration is chosen to match that used in a series of experimental studies by a group at UC Berkeley on lifted flame stabilization in gas turbine combustors, so that useful experimental data is available for comparison. The study will consists of four main parts: (1) DNS of vitiated methane-air flames using multi-step systematically reduced chemical kinetics, greatly extending the capability of DNS as a predictive tool; (2) Parametric studies with various combinations of H2O and/or CO2 dilution and a varying degree of premixing; (3) Evaluation and further development of theories and models for partially premixed flames; and (4) Critical assessment of performance and operational issues related to hybrid SOFC-GT systems. The DNS will utilize national high-end computing (HEC) facilities HPCx and HECToR. The research will involve collaboration with three research groups in the USA. The results will be disseminated in international conferences and journals as well as at meetings and the website of the Consortium on Computational Combustion for Engineering Applications, funded by the EPSRC grant No. EP/D080223/1 (2006-2009).

Publications

10 25 50
 
Description In this project, combustion diluted by water droplets/mist/steam was investigated in detail, which is a highly complex multi-physics, multi-scale, non-linear and unsteady problem. The study employed advanced numerical simulations, mainly direct numerical simulation (DNS) and large-eddy simulation (LES), to gain fundamental understanding of the complicated interactions involved. A methodology was developed, in which turbulent gas combustion was described in the Eulerian frame while the liquid diluent phase was formulated in the Lagrangian frame. Compared with the more widely used Eulerian - Eulerian formulation, the present methodology was physically more truthful but computationally more demanding. The resulting governing equations together with appropriate initial and boundary conditions were then solved with high-order numerical schemes, which was executed on the national massively parallel computers HPCx and HECToR. A wide range of physical and numerical parameters were tested to mimic different scenarios in practice. Some of the DNS conducted of turbulent combustion with water mist as a diluent included around 17 million evaporating droplets, which is the largest numerical simulation of its kind to our knowledge. Such large-scale simulations were able to provide better statistical quantities as well as information about instantaneous interactions among turbulence, combustion, heat transfer and gas-liquid phase changes. The simulation results were carefully analysed to scrutinise: (1) the dynamics of the dispersed phase consisting of evaporating droplets; (2) the effects of combustion dilution on flame extinction and re-ignition; (3) the co-existence of non-premixed and premixed flames; (4) the various mechanisms of combustion suppression; (5) the impact of adding an inert diluent to a fuel-oxidiser system on combustion theory and modelling; and finally (6) the implications of diluted combustion for a hybrid SOFC-GT system.
Exploitation Route Findings from the present study are of great value to R & D of fire suppression systems using water mist as well as water/steam injection technologies for reducing NOx emissions from gas turbine combustors and internal combustion engines.The research outcomes were extensively disseminated in journals, at international conferences and UK research meetings that were attended by people from both university and industry. The work formed a key part of the UK Consortium on Computational Combustion for Engineering Applications (http://www.coccfea.ac.uk). The training element was particularly successful, with the postdoctoral researcher, Dr. Jun Xia, on the project becoming a lecturer at Brunel University at the end of the project.

A significant follow-on work was the development of a lattice Boltzmann method (LBM) for multiphase flow simulation, which involved the work of additional PhD students. LBM, unlike DNS or LES, is a mesoscopic kinetic approach. An LBM methodology was developed for multiphase, multi-species flow simulation. The flow around droplets was fully resolved and droplets collisions were included, which was a much improved representation of the real physical picture in the above applications. The method has great potential for accurately simulating multiphase combustion as well as flow, which would be explored in a follow-on research project (EPSRC grant No. EP/I000801/1).
Sectors Aerospace

Defence and Marine

Chemicals

Construction

Energy

Environment

Manufacturing

including Industrial Biotechology

Transport

 
Description EPSRC Standard Grant
Amount £267,459 (GBP)
Funding ID EP/I016570/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 08/2011 
End 09/2014
 
Description EPSRC Standard Grant
Amount £426,844 (GIP)
Funding ID EP/G034281/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 11/2009 
End 04/2013
 
Description EPSRC Standard Grant
Amount £97,845 (GBP)
Funding ID EP/I000801/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 08/2010 
End 02/2012
 
Description EU 7th Framework Programme
Amount € 300,000 (EUR)
Funding ID No. 246772 
Organisation European Union 
Sector Public
Country European Union (EU)
Start 06/2011 
End 06/2015
 
Description The Royal Society Newton International Fellowship
Amount £100,000 (GBP)
Funding ID NF110280 
Organisation The Royal Society 
Sector Charity/Non Profit
Country United Kingdom
Start 02/2012 
End 06/2016