Abstract
The technological development of renewable heating systems by combustion of renewable fuels for production of industrial heat is considered a holistic approach. Cheaper and easily transportable Biomass-derived green fuels, such as syngas mixture or producer gas, are reliable options that can replace the traditional fuels in such systems. In the current energy transition context, biomass-derived syngas (BDS) is often recognised as a fundamental path towards decreasing fossil fuel dependency and greenhouse gas emissions. However, hydrogen-containing BDS are prone to flame instability problems due to having higher syngas burning velocity and low energy density. In this project, more efforts are carried out to adopt BDS efficiently in industrial combustors stabilized by co-firing with H2 and CH4 or inert gas (CO2 or N2) dilutions by exploring in-depth knowledge of syngas burning velocity, flame temperature and gas emissions. To do so, a deeper understanding of the syngas combustion and emissions behaviour is explored for various fuel compositions, inlet flow rates, fuel-air mixture strengths, and thermal loads. The objective of this experimental and numerical study fits into this multi-task project framework: Firstly, a systematic numerical study is carried out to evaluate flame stability parameters, such as Laminar Burning Velocity (LBV) and peak LBV location (ΦLBV=max) and to present fresh insights on the effects of purposefully customized ternary syngas mixtures, CO/H2/CO2 and CO/H2/CH4. Wherein, chemical kinetic simulations of equimolar (CO: H2=1:1) forestry waste syngas were systematically carried out taking advantage of the open-source CANTERA solver using different kinetic models. Two detailed kinetic models i.e., newly released FFCM2, and USC mech II, are mainly implemented to report accurate flame parameters at 1 bar and different temperature levels (from 300 K up to 450 K) at a wide range of Equivalence Ratio (ER, Φ). The modified GRI-mech model is also used where needed to support and benchmark the results. On validating the results with experiments, FFCM-2 proved to be a good kinetic model for the considered syngas CO/H2/CH4 for mixtures containing a limited share of 30 % methane at a normal and moderately elevated temperature, 0.4 ≤ Φ ≤ 2.1. The USC mech II kinetic model remained consistent and performed very well for syngas mixtures CO/H2/CO2. Additionally, when composition analysis was conducted at different temperatures, the progressive CO2 dilution and CH4 addition of up to 30% reduced the peak LBV with 9% and 40% reductions, respectively and moved the peak LBV locations (ΦLBV=max) from rich towards lean ER conditions; however, only the latter effect was enhanced at the elevated initial temperature. Moreover, sensitivity analysi is reported to explore the most sensitive intermediate reactions relative to LBV. The numerical insights gained from premixed flame simulation provided intelligence about the ER setting for the syngas burner in experimental activity. Further work investigated the renewable heat via experimentally measuring flame temperature during the combustion of multi-component Producer Gas (PG), composed of 1.90% CH4, 18.50% H2, 8.85% CO2, 22.20% CO, 7.40% H2O, and 41.15% N2, obtained from wood pack aging waste gasification. Tests are conducted on an experimental set-up consisting of a rising co-current stationary fluidized-bed gasifier coupled with a purpose-built combustion test rig. Flame shape stability and temperature distribution are analysed using thermal imaging in the 7.5–14 µm spectral range, with a pixel-based correlation method. The flame emissivity is calibrated using N-type thin-wire thermocouples and adjusted on the Infrared (IR) thermal camera. Results show that the PG flame remained stable and uniformly distributed within the combustion chamber at an ER (Φ) of 0.85 and a thermal load (TL) of approximately 24 kW. However, at a lower ER of 0.43 and a thermal load of 10 kW, the flame exhibited instability, likely due to reduced flame temperatures, which lower reaction rates and weaken the chemical kinetics governing PG combustion. Additionally, high air-flow rates under this condition contributed to flame perturbation, detachment, and quenching. Significant flame instability and “neck thinning” are observed at ultra-lean conditions. Introducing a small amount of methane (up to 5%) stabilized the flame under these conditions, suggesting a potential strategy for maintaining flame stability during ultra-lean combustion (Φ = 0.43 and TL ∼10 kW). Moreover, flue gas emissions are monitored for 0.45 ≤ Φ ≤ 0.70, which varied from 53 ppm CO, 10% CO2 and 55 ppm NOx to 550 ppm CO, 15% CO2 and 275 ppm NOx. Data analysis highlighted the fuel-NOx production mechanism, as NOx emissions increased from 200 ppm to 275 ppm with a TL increase from 20 kW to 25 kW at a fixed ER condition. Nevertheless, the critical flame behaviour, such as propensity of flashback conditions, especially at ER (Φ) of 0.85, could not be explored in greater detail due to inherited optical access limitations of the experimental set-up. For this purpose, a 2D numerical premixed flame model is developed to perform the experimentally validated Computational Fluid Dynamics (CFD) investigation of the PG flame and optimize the premixed combustion process. The simulation setup is constructed using the CFD domain size derived from the existing experimental setup. The experimental validation of the PG flame shape and flame temperature is performed by capturing the flame using an IR thermal camera. As a research objective, the possibility of load flexibility and the extent of variation in ER are explored systematically by varying the TL and ER conditions until flashback occurs. It is observed that the Boundary Layer Flashback (BLF) occurred at a partial load condition of TL ∼15 kW. This BLF behaviour may be seen due to the entry of the flame front into a lower inlet velocity region adjacent to the burner wall. Therefore, the flame operating at TL ≥ 24 kW at the ER ∼ 0.85 condition is recommended for the currently designed setup. Moreover, the Hydrogen-enriched Producer Gas (HPG) fuel mixture is also prepared by adding a 1.22 m3 /h flow of hydrogen to increase H2 up to 30 %. The propensity of a flashback also increased at lower ER ∼ 0.65 at partial load conditions of HPG mixture, and only a nominal load operation, TL ≥ 28 kW is recommended for HPG mixture for future tests. A range of Safe Inlet Velocity (SIV) at corresponding Safe Thermal Load (STL) was quantified at different ER conditions. Moreover, the change of STL curves during combustion of PG and HPG fuels was plotted over a wide ER range in the case of H2-cofiring. Furthermore, NOx emission data were explored, revealing interesting results.
Throughout the multi-task framework, the several combustion research activities conducted in this thesis have strengthened the confidence in advancing, adopting, and upgrading the biomass syngas combustion technology, making it ready to integrate into boilers and industrial process heaters. The future work could focus on the optimization of renewable heat-based combustion systems through better control over the residual fraction of PG directed towards the burner. Moreover, optimising the sealing system and implementing the changes in the current design of the burner are part of near-future goals with the aim of running the tests with greater flexibility in terms of fuel composition and load modulation.