Written by Richard Marsh, Cardiff University, Principal Investigator on UKCCSRC Call 1 Project Oxyfuel and exhaust gas recirculation processes in gas turbine combustion for improved carbon capture which completed in July 2014
What progress has taken place since the last blog?
We have been busy making some really exciting measurements. Our first job was to install an oxygen supply line into Cardiff University’s GTRC facility at Port Talbot. Oxygen is quite a hazardous substance since it reacts with almost any material, including metals, so we have undertaken a strict safety review and ensured that our oxygen supply complies with a number of safety regulations relating to this type of work. The new oxygen system is fully remote controlled, which means that we operate this via an interface in the control room, just as for air and fuel.
We have undertaken a series of combustion experiments at high pressure, which have provided very interesting and useful data in demonstrating how oxygen and carbon dioxide affect the flame in a gas turbine swirl burner. This data has been presented at an international gas turbine conference and we are finalising a research paper ready for publication in a quality peer-reviewed journal.
What testing has taken place?
A comprehensive test campaign was undertaken using a small-scale swirl burner at atmospheric pressure with a number of fuel, oxygen and diluent flow rates at around 5 kW and in some case to operate with pure oxyfuel (i.e. no diluent gas). This allowed us to produce operating maps, showing the effect of the different gases flowing into the flame zone. The data was then used to design a larger-scale series of tests where the burner performance could be studied in greater depth.
The large swirl burner has been operated at a number of power and pressure settings, up to 50kW and 3 bar pressure. We have introduced oxygen into the flame to give oxygen concentrations of almost 60% in some cases. Additionally we have added CO2 to the flame, which has allowed us to study the operational effect of simulated exhaust gas recirculation. Alongside this we have made gas analysis and chemiluminescence measurements to quantify what is happening in the flame in terms of heat release, location and products of combustion.
What have the results shown?
The results have been very enlightening. The effect of high concentrations of oxygen proved that the flame did not become difficult to operate due to the predicted increase in burning velocity. In fact the addition of excess oxygen behaved more like a diluent, thus extra oxygen began to replace the role of nitrogen in cooling and moderating the flame structure. Our results indicate that a 100% oxygen-methane flame (with no inert diluent) is entirely possible, but would need substantial amounts of excess oxygen to balance the flame in terms of chemical kinetics versus the flowfield. Hence, the study of exhaust gas recirculation becomes more critical as an engine developer would need to use up all of the (expensive) oxygen being produced. Gas analysis measurements showed that in all the cases tested the flame still operated satisfactorily, since all of the methane was burnt. However, an interesting discovery was that the dilution CO2 had the potential to dissociate, forming carbon monoxide in the exhaust. This might be a challenge for CO2 scrubbing systems and further research will be needed to understand this in more detail.
What will happen next?
Our results and findings have been published in a report to the UKCCSRC, plus we are formulating the key findings in a journal paper soon. We have also been successful in securing funding from the EPSRC to examine the effect of exhaust gas recirculation with Leeds and Edinburgh Universities, plus some key industrial partners. This 1.4 million pound project will allow us to extend the research and particularly focus on the dissociation phenomenon mentioned above.