Hydrogen utilisation in heat and power systems

February 13th, 2024

This project is focused on understanding the complexities of turbulent hydrogen combustion.

Project lead

Dr Jim Patel, jim.patel@csiro.au

Leader researchers

Mr Manahara Manatunga, manahara.manatunga@csiro.au

Challenge

Hydrogen combustion will play a key role in decarbonising critical Australian industries such as mineral processing and green steel production.

Understanding the intricacies of turbulent hydrogen combustion presents several challenges due to:

  1. Unique Properties of Hydrogen: Hydrogen has high flame temperatures, a large flammability range, high flame propagation speeds, high mass diffusivity, high thermal NOx, and the tendency to flashback and detonate. These properties make hydrogen combustion behaviour significantly different from that of traditional fuels, and more difficult to predict using current models optimised for hydrocarbon combustion.
  2. Effects of Turbulence: Turbulence can stretch and wrinkle the reaction layer, sometimes causing local flame quenching and leading to combustion instabilities. Understanding these effects requires detailed knowledge of both fluid dynamics and combustion chemistry.
  3. Coupling of Turbulence and Chemistry: The interactions between turbulence and chemistry are highly complex, multi-scaled and coupled. Turbulence affects the mixing and reaction rates, while the heat released by the reactions can affect the turbulence. This coupling makes the combustion process highly nonlinear and difficult to model.
  4. Numerical Modelling Challenges: Developing numerical models that can accurately predict turbulent hydrogen combustion is a significant challenge. These models need to account for the complex chemistry, the effects of turbulence, and the coupling between them. They also need to be computationally efficient to be practical for use in combustor design.
  5. Experimental Challenges: Measuring properties in turbulent hydrogen flames is difficult due to the high temperatures, high speeds, and small scales involved. Advanced diagnostic techniques are required, but these can be challenging to implement and interpret.

These challenges make understanding turbulent hydrogen combustion a complex but important task. Overcoming these challenges will require a combination of experimental studies, numerical modelling, and theoretical analysis. The ultimate goal is to develop accurate and efficient models that can be used to design safe and efficient hydrogen-fuelled combustors for heat and power applications.

What we are doing

This project is focused on understanding the complexities of turbulent hydrogen combustion. The research activities are divided into three main tasks:

1. Buoyancy effects on hydrogen diffusion flames

The project aims to understand how buoyancy influences the mixing characteristics and flame structure of hydrogen diffusion flames.

The study has found that in the far field region of the flame, buoyancy effects dominate inertial effects, causing the fuel to mix with the surrounding air more quickly and resulting in a shorter flame length. Furthermore, the study has shown that buoyancy predominantly affects fuel-air mixing, while its effect on chemical kinetics is inconsequential.

The study recommends that the buoyancy term be included in Computational Fluid Dynamics (CFD) calculations of hydrogen diffusion flames.

2. Reviewing the Laminar Flamelet Concept in diffusion flames

The Laminar Flamelet Concept is a commonly used modelling method to describe the turbulence-chemistry interaction in diffusion flames. This concept assumes that combustion occurs within a very thin reaction zone (smaller than the smallest turbulent scale, the Kolmogorov scale), and can be calculated using one-dimensional opposed flame flamelet equations in mixture fraction space.

However, the research has found that regions exist, at sufficient distance from the fuel inlet (nozzle), where unsteady effects control the evolution of the flame structure and need to be accounted for. The project is actively developing a way to include unsteady effects in hydrogen diffusion flame problems.

3. Differential diffusion in hydrogen flames

The traditional application of the Laminar Flamelet Concept assumes that all combustion species have equal diffusivities and that their mass diffusivities and thermal diffusivities are equal. While this assumption is valid for many hydrocarbon flames, is not suitable for combustion processes that involve hydrogen-rich fuels.

Hydrogen molecules (H­2) and hydrogen atoms (H) have a uniquely high mass diffusivity compared to their thermal diffusivity and this differential diffusivity plays an important role in flame structure evolution in turbulent hydrogen flames. This calls for the need to include a detailed definition of diffusive transport effects in predictive models designed for hydrogen combustion.

To overcome the limitation of the conventional flamelet method, an improved method has been developed within this project. This method calculates the ‘flamelets’ in physical space with detailed transport using a 1-dimensional opposed flow diffusion flame solver, and then imports them into the CFD software to create Probability Density Function (PDF) look-up tables. Large Eddy Simulations (LES) are then used to simulate the flow field and the PDF tables to account for detailed chemistry and transport. Preliminary results of temperature and mixture fraction distributions that have been obtained are included in the below animation.

This animation shows the Temperature, Mass Fraction of the OH radical (YOH) and the Mixture Fraction (Z) predictions of a hydrogen/nitrogen diffusion flame in air. The hydrogen/nitrogen mixture is issuing vertically from a fuel inlet with a circular cross section. This inlet is centred at the spatial location where The X-Coordinate and the Y-Coordinate are both 0 m. The Y-Coordinate indicates, in metres, the axial distance from the fuel inlet, and the X-Coordinate indicates the radial distance from the radial centre of the fuel inlet. The results are from a preliminary simulation carried out by the Gas Processing Team of CSIRO Energy, using the high-fidelity turbulence modelling technique Large Eddy Simulations (LES) coupled with the Steady Flamelet Model to model the evolution of the thermochemical state of the reacting flow.

The fundamental understanding gained from this project has supported the development of an ultra-low NOx hydrogen combustor.

Outcomes to date

In our study into Buoyancy Effects on Hydrogen Diffusion Flames, we have found that appropriate representation of buoyancy effects is critical when modelling hydrogen diffusion flames using CFD. Based on this study, we have recommended that, in the future, researchers appropriately account for buoyancy in their studies of such flames.

In our review of the application of the Laminar Flamelet Concept to hydrogen diffusion flames, we have demonstrated that the steady flamelet assumption only applies to part of the flame while the remaining area is affected by unsteady affects that need to be accounted for. While work is being carried out to include such effects to our existing flame model, we believe that the discovery of these unsteady effects informs the research community of their existence and the need to address them.

Lessons learned

The broad literature review done within this project, especially on the combustion kinetics and the NOx pathways has led to development of predictive numerical models that enabled the development of the ultra-low NOx hydrogen combustor.

Project finish date

August 2024

Relevant project publications

Effect of buoyancy on CFD prediction accuracy of hydrogen/methane jet diffusion flames. M Manatunga, FC Christo, J Sheehy, J Schluter, S Shelyag. Proceedings of the 23rd Australasian Fluid Mechanics Conference. https://www.afms.org.au/proceedings/23/Manatunga_et_al_2022.pdf

Influence of buoyancy on the mixing, flame structure, and production of NO in hydrogen diffusion flames. M Manatunga, FC Christo, J Schluter, S Shelyag. International Journal of Hydrogen Energy 57, 328-337 https://doi.org/10.1016/j.ijhydene.2024.01.041

In preparation:
  • Detailed review of the flamelet concept in modelling hydrogen diffusion flames. M Manatunga, FC Christo, J Schluter, S Shelyag. Planned submission to journal Combustion and Flame. Planned submission date 28/02/2024.
  • Flamelets calculations of hydrogen diffusion flames in physical space including detailed transport. Planned submission to journal Combustion and Flame. Planned submission date 30/04/2024.