Ignition and Heat Release Behaviour of iso-Butanol and Gasoline Blended Fuels: An Experimental and Kinetic Modelling Study

The decarbonisation of transport and the introduction of further renewable energy sources are required to minimise the impacts of climate change, while meeting the energy needs of a developing global population. Introducing alternative fuels into existing and developing spark ignition (SI) engine technologies requires the thorough characterisation of the fuel’s combustion behaviour. The propensity of fuel to autoignite is a key property which limits SI engine performance through the development of engine knock. Autoignitive behaviour can be characterised by ignition delay time (IDT) measurements in rapid compression machines (RCM) or by measuring knocking behaviour within practical engines. RCMs provide an opportunity to study a fuels ignition behaviour at the fundamental level, the measurements of which often serve as a prediction for behaviour in more complex systems, such as SI engines. Through the application of both techniques, this work investigates the influence of iso-butanol blending on the combustion behaviour of gasoline (with particular focus given to the anti-knock properties of fuel blends), as well as assessing the validity of applying fundamental studies to predict practical engine level combustion behaviour. Accurate computational modelling provides an opportunity for the prediction of combustion behaviour quickly and cheaply when compared to experiments, facilitating the rapid optimisation of engine and fuel blend designs.
To enable the computational modelling of gasoline, surrogate fuels are required which replicate the target behaviours of the reference fuel, while minimising molecular complexity. The ability of a newly developed five component surrogate (5-C) to reproduce the autoignition behaviour of a research grade gasoline (RON 95 MON 86.6) is investigated within an RCM, at temperatures of 675-870 K, a pressure of 20 bar and equivalence ratios of 0.5 and 1.0, producing an excellent representation at stoichiometric conditions but displaying much lower reactivity than gasoline at lean conditions. When blended with iso-butanol (at 10, 30, 50 and 70% iso-butanol by volume), the representation of gasoline by 5-C continues to be generally good but at low temperatures (<770 K) and high iso-butanol concentrations (iB50/70), 5-C blends are considerably less reactive than gasoline blends. Upon investigation within a motored, skip-firing SI research engine, the 5-C continued to provide an accurate representation of gasoline’s normal and knocking combustion behaviour at spark advance timings of 2-10 CA° bTDC. Under blending with iso-butanol the surrogate continued to perform well but blends were observably less reactive at spark advance timings <8 CA° bTDC. Blends of 20-50% iso-butanol were found to be optimal for use in SI engines, providing considerable anti-knock benefits and comparable indicated power to gasoline. Correlations between RCM and engine measurements display the proficiency of fundamental measurements in predicting combustion behaviour within an engine at similar thermodynamic conditions.
Changes in the autoignition behaviour of 5-C due to blending with iso-butanol (5-70% iso-butanol) were studied experimentally within the RCM and computationally via chemical kinetic modelling. At low temperatures, iso-butanol generally reduces reactivity, suppressing the intensity of LTHR. As temperatures are increased, iso-butanol appears to suppress NTC behaviour and a cross-over in IDT measurements is observed between blends of 5 and 10% iso-butanol, wherein the 10% blend becomes the most reactive at intermediate to high temperatures. Modelling results largely failed to replicate complex blending behaviour and largely underpredicted IDTs in the NTC region. It is proposed that the model’s misrepresentation of LTHR behaviour is a cause for such global model failures, as evidenced by local OH, brute force enthalpy of formation and reaction A-factor sensitivity analyses which highlighted the importance of reactions and species of significance to first stage ignition and low temperature oxidation processes, in the determination of IDTs and characteristic LTHR properties. Minimising uncertainties in the thermodynamic properties of complex oxygenated species typical of low temperature oxidation would produce more accurate model predictions, as these uncertainties are currently large for many important species. The influence of these uncertainties on the parameters investigated in this study is substantial. Current computer models therefore cannot be effectively applied in the prediction of the combustion behaviour for gasoline/iso-butanol blends until these issues are resolved. Further studies of the species and key reactions identified in this research would help to improve current kinetic mechanisms.