Intelligent instrumentation for assessment and monitoring of hydrogen blend fuels in domestic boilers

Significant reduction of greenhouse gas emissions (GHG) has become the utmost endeavour to achieve net-zero emissions by 2050. In the UK, domestic heating itself is responsible for 17% of the total GHG emissions, this is comparable to the contribution of all petrol and diesel cars (BEIS, January 2022). Therefore, the decarbonization of domestic heat is a big challenge. A sustainable route to reduce GHG is to replace natural gas (NG) with hydrogen (H2) since the combustion of H2 does not produce CO2. However, the challenge for H2 combustion is that its combustion characteristics substantially differ from NG (methane, CH4), e.g., its use affects combustion stability, heat release and NOx emission, and increases the combustion rate due to a higher H2 flame temperature. Various technological challenges are also associated with using pure H2 such as its production, safety, quick charge capability and low density, which limits its storage capabilities. At this transitional stage, a practical option is the use of higher H2 enriched fuel (i.e., more than 20% blend with NG), which would be a promising solution to lower the CO2 emission compared with other fossil fuels. However, the impacts of higher H2 enriched fuels on the widely used condensing heating boilers are not extensively studied and fully understood. The H2 enrichment leads to higher flame radicals such as OH*, CN*, CH* and C2*, higher combustion temperature and flame destabilisation, thus triggering higher NOx formation. The flame radicals are closely related to the combustion structure, temperature, heat release and pollution emissions. Moreover, domestic condensing boilers use premixed cylindrical/surface burners, and these burners produce an array of flames. It is extremely difficult to measure flame radical information in different depths of the array of flames using existing measurement systems. The development of an intelligent instrumentation system has, therefore, become indispensable to assess and monitor the flame radical emissions and NOx formation process at different depths of flames, thus facilitating an in-depth understanding of the combustion process of different H2/CH4 blends.
This project will develop and implement a new instrumentation system based on multi-spectral light field imaging to assess and monitor the flame radicals and temperatures with different H2/CH4 blends in domestic boilers. Light field image formation and depth reconstruction models will be developed to generate flame radical images at different depths for different spectral bands. The developed system will provide distinctive capabilities for characterising and quantifying the radical information and temperature profiles of a flame in a single exposure, simultaneously. The proposed project will also develop an intelligent data-driven model based on machine learning to predict NOx emission, thus, facilitating the improvement of domestic boiler performance. The relationships between flame radical characteristics and NOx emission will be established by conducting a series of experiments initially on a lab-scale test rig and then on commercial domestic boilers under different H2/CH4 blends and boiler settings.
The prototype system will also be tested on a gas turbine test rig to evaluate its wider applicability. Experiments will be conducted to investigate the characteristics of CO2, H2 and ammonia (NH3) blend combustion, thus providing an in-depth understanding of stability regions and NOx emission with different proportions of CO2/H2/NH3 in the blend.
The outcomes of this research will provide in-depth knowledge of the combustion characteristics of H2 blends, understanding of the boiler efficiency and pollutant formation process of domestic boilers. Once the system is developed, it will be used for the design of domestic boilers, and the engineering insights produced during the project could be used to develop a portable diagnostic tool for routine monitoring of blended-fuel boilers.