The scientific goal of the CLEAN-Gas Project is to develop new experimental and numerical tools for improving natural gas combustion in innovative burners. It is important to underline that Natural Gas is of great interest to the European Energy policy due to its widespread availability and its environmental and technological benefits. Therefore, a deep understanding and high-level training in the experimental and numerical tools for investigating natural gas combustion in new burners are of upmost importance for future scientific and technological developments.
Forecasters predict that natural gas consumption in the EU will double over the next 25 years. European natural gas consumption currently represents approximately 17% of world consumption. European gas imports are expected to reach slightly over 80% of total consumption by 2030. To tackle this challenge, the EU is investing heavily in natural gas equipment, as demonstrated by the construction of the Nabucco pipeline in Turkey. The use of natural gas makes it possible to divide CO2 emissions nearly by 2 compared to coal. It also enables the use of gas turbines with an efficiency close to 50%. Natural gas is present in all sectors from companies/business to personal/private sector.
Natural gas is a fossil fuel whose energy conversion is mainly achieved by combustion. The combustion process induces two main side effects: the production of greenhouse gases (CO2) and the emission of pollutant species such as nitrogen oxides (NOx) and soot particles. Conventional techniques used to reduce these emissions, already low compared to usual fossil fuels, are often post-combustion treatments and they include CO2 storage, flue gases cleaned up by catalytic and non-catalytic conversions. Another solution is to act directly on the combustion process in order to limit pollutant emissions at the source while maximizing combustion efficiency. New processes are currently using this strategy, for example regenerative burners, flameless (MILD) combustion, combustion of highly diluted mixtures or oxy-combustion.
These processes, despite of the fact that they are already used in different industrial units, are still poorly understood and very difficult to transpose from one industry to another. Therefore, it is extremely important to develop academic and research studies on these new combustion processes to make best use of existing resources while limiting their environmental impact. These new processes are very different from existing technologies and constitute real technological breakthroughs.
The CLEAN-Gas project covered different topics that are relevant for the clean combustion of natural gas, from chemical kinetics of combustion to fluid dynamics and new technologies, both experimentally and theoretically. Considering the complex nature of the controlling pollutant formation and flame stability, the use of both experimental investigations and Computational Fluid Dynamics (CFD) proved to be essential for the development and implementation of such novel combustion technologies. In particular, the appropriate description of the interactions between the combustion process, and the system aerodynamics is crucial in order to develop innovative, clean and safe combustion systems.
Figure 1. CLEAN-Gas partners.
The project involved different academic partners and companies, belonging to different European countries, as described in Figure 1. The project contributed to the training of several ERSs, and improved the understanding of the combustion processes involving natural gas. The project partners worked on the development of detailed kinetic mechanism able to describe the combustion of natural gas in conventional and non conventional systems, including the formation of NOx and particulate emissions. Moreover, experimental and modelling activities addressed the combustion of natural gas in laminar and turbulent systems. The whole approach is based on a comprehensive investigation of different turbulent flames at different levels of complexity. The turbulent flames of increasing complexity vary in their nozzle design, fuel, and pressure conditions.
Kinetic model reduction and automatic optimization tools were developed, in order to facilitate the adoption of detailed chemistry inside turbulent 3D simulations. Different turbulent-chemistry interaction approaches were analysed and tested in comparison with Direct numerical Simulations (DNS) and with experimental results mostly in the context of Large Eddy simulations (LES). Finally, the numerical tools developed and validated inside the project were adopted for the simulation of industrial-scale combustion devices with the aim to improve the combustion process and reduce the formation of pollutant species.
The project involves the development of a comprehensive kinetic mechanism to model the pyrolysis, oxidation, and combustion of natural gas with predictive capabilities in a wide range of operating conditions. Special attention has been devoted to thermodynamic data and to the chemistry of pollutant species, particularly PAHs, carbonaceous particles (soot) and NOx. Several laminar sooting flames were modeled with laminarSMOKE CFD code with good agreement with experimental data (Figure 2 ad Figure 3) Similar validation was performed to validate the NOx sub-mechansism.
Figure 2. Sooting flame prediction using CFD simulations.
Figure 3. Comparison between measurements and simulation of soot volume fraction.
The kinetic mechanism has been improved also for unconventional combustion conditions, such as diluted MILD (Moderate or Intense Low oxygen Dilution) and oxy-fuel conditions, on the basis of the coupling of kinetic simulations with Uncertainty Quantification (UQ) and Principal Component Analysis (PCA). Project results demonstrated the applicability of the Kriging-PCA approach for the prediction of unexplored operating conditions of 1D and 2D flames and for parameter exploration of LES simulations. A new tool (OptiSMOKE++) was developed for kinetic mechanism reduction and optimization, resulting in a reduced mechanism which performs better with respect to non-conventional conditions.
The complex interactions between kinetics and turbulence were studied especially in MILD combustion regime conditions by developing simulation tools able to provide high-fidelity numerical experiments. Particular attention was devoted to the application/extension of the EDC/PaSR approach to Large Eddy Simulations. A comparison between different approaches based on EDC, PaSR allowed the development of a dynamic PaSR model and validation on lab-scale burners using both RANS and LES. Moreover, the LES methodology was extended to the Thickened Flame (TF-LES) model using a flame-wrinkling factor to further investigate flame/turbulence interactions. A new combustion tool relying on the coupling with both the Transported-PDF/FDF and the FPV/FGM based tabulated chemistry was developed for LES simulations. This model was successfully validated and applied on simple flames and oxy-flame configurations.
Since in many industrial devices lean-premixed flames are stabilized by a swirling flow, during the project several experiments studied the dynamics of flame stabilization and the flame response to flow disturbances, using a loudspeaker to generate the flow modulations. Thermo-acoustic instabilities, blow-off and flashback limits were experimentally characterized during the project. A specific burner was investigated to provide a reliable dataset of turbulent flames experiments for the validation of numerical simulation tools in the pressurized conditions of interest for gas turbine combustion.
Fig. 4. Flame light distribution for different configurations tested with different swirl intensity (S).
The experimental activity allowed the characterization of the flame response to incident acoustic waves. Large Eddy Simulations of the cold flow response to incident acoustic waves were also achieved. Post-processing techniques were developed to analyze the different velocity, heat release and pressure signals in the time and frequency domains, allowing the developments of scaling rules to interpret the observations. Different test rigs and combustors were used to measure the thermochemical states (temperatures, species concentrations) for a piloted turbulent lean premixed gas flames and flame structures were interpreted in terms of their combustion regime. Reactor network modelling was used to predict CO formation in close to reality combustor under consideration of measured temperature and residence times. Finally, a multi-physics LES tool has been used for the simulation of industrial-scale combustion devices, with the aim to study the combustion process. Particular attention was devoted to the role of radiative energy transfer and to the effect of different configurations (oxygen enrichment, CO2 dilution, high pressure) on the emissions from turbulent methane flames. Coupled simulations accounting for spectral properties of participating gases and of the quartz windows confining the flame have shown a great impact of radiative heat transfer on both flow and flame topology. Due to the large computational efforts in modeling the complex geometries of 3D real scale burners, a novel concept named virtual chemistry approach was developed and validated in comparison with the detailed kinetic mechanism first in ideal reactors and laminar flames. This approach was then adopted to perform the LES simulation of the Preccinsta burner using the developed reduced virtual chemistry mechanism.
Fig 5. LES simulation of Instantaneous field of Radiative Power.
Overall, project already resulted in more than 20 papers published on international peer-reviewed journals, and 31 presentations and posters at international conferences. Other papers prepared within the project are currently under review and will be published in the near future.
Main outcomes of the program are skilled people, scientific and transferable knowledge. All the ESRs completed their training and are finalizing their thesis in view of the PhD defenses. They are now well-trained experts in Natural Gas combustion and its applications in a general context of environmental needs. They master the necessary tools and methodologies from a cross approach perspective (theory, computation and experimentation), and are professionals able to address the main challenges in the combustion of natural gas.
The second outcome is scientific knowledge and expertise. The multi-scale (from micro-phenomenon to macro) and multi domain approach (chemistry to fluid mechanics) approach adopted inside the CLEAN-GAS project greatly contributed to the knowledge development in the field of natural gas combustion. Novel modeling approaches were developed both in the chemistry, fluidynamics analysis and in the simulation of the practical applications. A kinetic mechanism able to describe the combustion of natural gas and the formation of soot and NOx pollutants was developed and validated extensively. An automatic kinetic model reduction and optimization allowed to extend this kinetic mechanism to non conventional combustion conditions (i.e. diluted and oxy-fuel combustion and high pressure). Novel approaches to model turbulence-chemistry interaction were developed and validated in this project, thanks to the availability of LES solvers and new experimental results.
Publications, conference participations, and seminars allowed the ESRs to interact with the scientific and technical community involved in the exploitation of natural gas combustion. They also interacted with partner companies which hosted the students and advised the CELAN-GAS project partners about the industrial needs and perspective. Finally, a fourth outcome of the project is the creation of an international thematic network and community among T.I.M.E. Association.