Hydrogen Combustion Technologies

Task 1.3: Hydrogen Combustion

The overall objective of the task is to provide clean and efficient solutions for CO2 capture technologies based on combustion of hydrogen and hydrogen-rich synthetic gaseous fuels

Task 1.3 contributes to the overall objectives of BIGCCS through:

  1. Advancement in the fundamental understanding of the physical processes that characterize combustion of hydrogen, including flame propagation and pollutant formation, at both laminar and turbulent conditions, as opposed to combustion of more conventional gas turbine fuels (natural gas, fuel oil).
  2. Optimization of the hydrogen fuel injection method and of its mixing pattern within the oxidant stream, while still conserving operational safety and fuel efficiency.
  3. Acquisition of detailed knowledge about flame propagation and stabilization characteristics, with specific focus on flashback (upstream flame propagation against the direction of the main reactant’s flow), for hydrogen-air flames.
  4. The close collaboration with world-renowned combustion research groups at Sandia National Laboratories (Livermore, California), DLR (Stuttgart, Germany), TU Munich (Munich, Germany) and Princeton University (Princeton, New Jersey) strengthen considerably the scientific level and the international visibility of the research performed within the task.

Addressing point 1) is of fundamental importance to improve the confidence in the utilization of hydrogen-rich mixtures as main fuel for large-scale, clean and highly efficient power generation with minimal CO2 footprint. Solving the issues related to point 2) and 3) above will enable the deployment of pre-combustion CO2 capture schemes based on new, highly efficient (no fuel dilution with N2 required!) hydrogen combustion equipment for gas turbine engines.This is a necessary step within the pre-combustion route to CCS because using existing combustion equipment is presently hindered by the peculiar combustion properties of hydrogen (high reactivity, high diffusivity) compared to more traditional hydrocarbon fuels, for which existing equipment has been designed.

Task 1.3 achievements so far:

  1. Flashback of hydrogen-air turbulent flame in ducts is characterized, using Direct Numerical Simulation (SINTEF/Sandia NL) and high-definition laboratory experiments (TU Munich), a previously unknown feature of near-wall flame propagation is discovered and severe limitations of a widely used boundary layer flashback model from Lewis and Von Elbe are exposed.
  2. A novel analytic flame channel-flashback model is proposed and validated (SINTEF/Sandia NL).
  3. A new fuel injection device, based on a porous steel support, a H2-selective membrane and a ceramic diffusor layer, is being investigated, both numerically and experimentally, in order to solve well-known issues related to hydrogen injection through conventional fuel injection methods (nozzles) and the concentrated point sources of the highly reactive fuel species (SINTEF/DLR).
Fully developed turbulent channel flow

Channel flow turbulence at a friction Reynolds number of 180: flow is in the positive x-direction (from right to left), red iso-surfaces of the second invariant of the velocity gradient tensor (lambda2=-0.01) represent near-wall vortex structures.

 

Anchored v-flame case

Direct Numerical Simulation of a fuel rich (ϕ=1.5) H2-air premixed v-flame anchored in a turbulent channel flow. Gruber et. al., J Fluid Mech 658, 5-32 (2010).

 

Freely propagating flame case

Direct Numerical Simulation of a fuel lean (ϕ=0.55) H2-air premixed turbulent flame freely propagating upstream (flashback) in a turbulent channel flow. Gruber et. al., J Fluid Mech 709, 516-542 (2012).

 

Premixed and stratified upstream flame propagation

Upstream flame propagation in fully developed channel flow with premixed to stratified mixture transition: flow is in the positive x-direction (from left to right), flame moves upstream against the bulk flow (from right to left), red iso-surface marks the flame (T=1700K) and blue isosurfaces backflow pockets (u=0).

 

Premixed and stratified upstream flame propagation

Upstream flame propagation in fully developed channel flow with premixed to stratified mixture transition: flow is in the positive x-direction (from left to right), flame moves upstream against the bulk flow (from right to left), channel flow turbulence is visualized by isosurfaces of the second invariant of the velocity gradient tensor (lambda2) and the flame by the red isosurface (T=1700K).

 

Premixed and stratified upstream flame propagation

Upstream flame propagation in fully developed channel flow with premixed to stratified mixture transition: flow is in the positive x-direction (from left to right), flame moves upstream against the bulk flow (from right to left), red iso-surface marks the flame (T=1700K). Lateral view.

 

Premixed and stratified upstream flame propagation

Upstream flame propagation in fully developed channel flow with premixed to stratified mixture transition: flow is in the positive x-direction (from left to right), flame moves upstream against the bulk flow (from right to left), channel flow turbulence is visualized by isosurfaces of the second invariant of the velocity gradient tensor (lambda2) and the flame by the red isosurface (T=1700K). Lateral view.

 

Reactive jet-in-cross-flow case

Direct Numerical Simulation of non-premixed turbulent flame stabilization in the wake of a transverse hydrogen fuel jet in cross flow of air. Grout et. al., J Fluid Mech 706, 351-383 (2012).

 

Model validation

Validation of flame shape flashback model (red line) versus DNS data (black scatter plot). Gruber et. al., Proc Comb Inst, in press (2014).