Crude oil refining is (after chemicals production) the second most energy intensive industry in advanced economies. For example, refining consumes 6% of the total energy used in the US. Current refinery technology is based on distillation for separation of the crude oil into fractions with varying molecular weights, followed by further reactions on some of these fractions (reforming, hydrotreating, cracking etc), which must then be further distilled. Distillation involves converting a large fraction of a liquid feed into a gas by boiling, so that compounds present in the feed can be separated by means of differences in their boiling points. In refining, distillation typically account for more than half of all the energy consumed, since the phase change on boiling requires significant energy input. One way of avoiding this large energy consumption would be to carry out fractionation in the liquid phase via a membrane. If a mixture of hydrocarbons is pressed by pressure against a membrane, and the membrane is permeable to only some of the materials, then we can separate the molecules that pass through the membrane from those that do not. This avoids the large energy injections of evaporation or distillation. Theoretical calculations show that the energy required for concentrating a mixture by membrane separation is less than 5% of the energy associated with distillation. Not surprisingly, people have been interested in using membranes to separate and concentrate liquids for some time. The majority of membrane separations to date are water based. A major success is in the area of desalination, where membranes are used to separate fresh water out of seawater. Membranes are not generally used for molecular separations in organic systems, because until recently there were few membranes stable in organic liquids. This has changed recently - research at Imperial College has developed membranes that are stable in most organic systems. These have been commercialised through an Imperial spin out company, Membrane Extraction Technology (MET) which was acquired by Evonik Industries on 1 March 2010. Evonik MET has made a substantial investment in a large scale membrane manufacturing facility in West London, delivering on the UK Government's vision for Manufacturing the Future. In many cases the required separation cannot be achieved in a single pass through a membrane, because the membrane does not discriminate highly enough between the different molecules that are present. In these cases, to achieve the required separation, the liquid can be processed through membranes multiple times. This arrangement of membranes is known as a membrane cascade. Given the advances that have been made in the development of membranes for organic systems and their application in membrane cascades, this project will research the use of membranes for refining crude oil. Membranes do not require boiling and condensation, and so can be operated at a single temperature. This will reduce the needs for heating and cooling, and so the associated heat losses. Thus we expect that isothermal refining with organic solvent nanofiltration membranes will significantly reduce the energy requirements for manufacturing fuels and lube products from crude oil. The project will work with a synthetic clean crude, made up to simulate the key hydrocarbon components of a real material. Experimental and simulation work will be used to design a membrane cascade to separate the synthetic crude; this cascade design will then be assembled and operated to prove the concept. Further simulations will then estimate what energy savings would result if isothermal refining were employed with a real crude. The project will work closely with partner Shell Gobal Solutions, who are a major company in oil refining and refinery technology.

Particulate Matter emissions are important for the health of both our planet and its population. Legislation on Particulate Matter (PM) is increasingly stringent and subject to the greatest change. However, Particulate Matter emissions are both the most difficult to measure and the most challenging to model. In simple terms, Particulate Matter is essentially soot (carbon) onto which species such as unburned hydrocarbons are adsorbed, and is intrinsic to all combustion processes.PM has been linked to global warming and serious epidemiological issues, and this has led to much regulation. Of greatest concern are the sub-micron particles that are invisible, yet it is these small particles that have the greatest deposition efficiency in the human respiratory system. While society is not yet able to replace combustion, technological developments can seek to minimise PM emissions.The complexity of measuring and modelling PM emissions means that modellers are dependent on published experimental data which can be old and incomplete. Furthermore, the modellers have no opportunity to specify the experiments. We have already collaborated on an informal basis, but without specific funding this has been very restricted. The integrated approach in this project will enable the modellers to specify the experiments, identify the most important measurements, and create a database that can be populated with the relevant experimental data. This will provide the modellers with immediate access to complete data.Oxford has a spark ignition engine with comprehensive optical access, and a range of burners: pre-mixed flat-flame (McKenna type), and co-flow (Santoro type, diffusion flame). Mass flow controllers allow us to vary the equivalence ratio of the core flow (pure fuel to the weak limit, with a choice of diluents) and the composition and flow of the annular flow (oxygen enriched or depleted air). The same fuels can be used in both the engine and burners, and in both cases the fuel composition can be controlled Measurement Capabilities - Oxford has a mix of proprietary and unique equipment for PM measurements. We can measure size distributions, mass loadings, composition, morphology and surface area. We also have a unique Differential Mobility Analyser that allows size segregation prior to PM characterisation. We have techniques for measuring temperature, as this has been identified by the modellers as of paramount importance. Most significantly, we will apply the novel technique - Laser Induced Grating Spectroscopy, LIGS temperature measurements of high accuracy and precision. In LIGS, the coherent, laser-like signal beam offers high discrimination against background scattering and luminosity to give a good signal-to-noise ratio in sooting flames. Potential exists for developing a transportable instrument for thermometry of flames in laboratory or technical combustion systems such as engines, gas turbines, incinerators etc.Computational modelling - Cambridge will create and improve computational models which describe combustion chemistry and soot particle formation. Combustion chemistry involves thousands of chemical reactions; the model must contain enough detail to describe the species which are important but must be concise enough to make numerical evaluation possible. Such models exist, but contain many parameters which need further refinement via experimental data. Modelling the formation and growth of soot particles is even more challenging. The chemical reactions between gas phase species and particle surfaces have to be combined with a population balance model to predict the particle mass and size. Eventually a detailed particle and chemistry model must be included in an engine model. These models are to be used to understand the mechanism of particle formation in an engine further and with this find operating modes which reduce particle formation and increase the efficiency of the engine.

This project proposes a novel concept of using nanofuels, pure metal nanoparticles (dry-fuel) or suspensions of nanoparticles in a liquid fuel (wet-fuel), as a future energy vector. The concept comprises three elements: production, utilization and regeneration of nanofuels. The utilization of thermal energy upon combustion of nanofuels is identified as the key element for the feasibility study. This project will use reciprocating internal combustion engines (ICEs) as a model system to assess the combustion process of nanofuels. Three identified potential nanofuels, silicon, aluminium and iron, in the form of wet-fuels and dry-fuels will be injected and combusted in two ICEs, and the engine performance including in-cylinder pressure, temperature and work output will be characterised. Key features of the experimental assessment including nanofuel formulation and injection, ignition and combustion of nanofuels, oxide particle capture and regeneration, and engine emission, wear and lubrication will be investigated. Our preliminary investigations indicate that the project is a visionary and innovative but risky application, worthy of EPSRC funding.

Laminar burning velocity measurements are needed for models that assist development of clean and efficient combustion in engines, boilers and furnaces, and for the validation of laminar burning velocity models. Our laminar burning velocity measurements are made in a spherical vessel with central ignition, so that a spherical flame front forms and propagates radially. During combustion the pressure rises and the unburned gas ahead of the flame front is compressed isentropically. So, from a single experiment, laminar burning velocity data are obtained for a sequence of linked temperatures and pressures. By varying the initial temperature and pressure the effects of pressure and temperature can be decoupled, and correlations generated for the effect of temperature and pressure on the laminar burning velocity. The schlieren system can detect the onset of cellularity (even when the flame is larger than the 40 mm diameter windows) so that we can avoid using data that violates our smooth flame front assumption. This builds on earlier work at Oxford over 16 years that has led to 3 innovations:+ Free-fall experiments to eliminate the effect of buoyancy.+ A multi-zone combustion model for data analysis, so that the effect of dissociation and the temperature gradient in the burned gas (typically 500 K) is incorporated into the analysis of flame front position and pressure rise.+ The use of 'real residuals' by retaining part of the previous combustion event as residuals, as opposed to the conventional approach of using a fixed composition N2/CO2 mixture to represent the residuals.Our facility has a comprehensive LabView interface for setting-up the experimental conditions and data logging. This system also ensures that condensation of either fuel or water vapour (in the residuals or as a diluent) is avoided. A schlieren system with a high speed video camera records early flame growth and cellularity (the departure from a smooth flame front, if it occurs). The experimental data are analysed by MATLAB routines that incorporate: image processing (of the schlieren system data), a multi-zone combustion model, and experimental pressure data; the code also combines data from multiple experiments in order to generate correlations for the laminar burning velocity. Initial conditions can be up to 450 K and 4 bar (final pressure limit of 35 bar), with combustion data obtained up to 30 bar and an unburned gas temperature of 650 K. Liquid fuels (or diluents such as water) can be added by a Hamilton precision glass syringe which is controlled by a syringe actuator. This facility and software are readily adaptable for testing different fuels.The combustion of fuels from renewable sources and their performance when combined with conventional fuels is very important. In 2008 the UK crude oil consumption was 78.7 Mt (~20% gasoline) - BP Statistical Review of World Energy; June 2009. EU legislation requires bio-fuel to become a minimum 5.75% of the total fuel consumption in 2010. Gasoline vehicles can mostly operate on a 10% ethanol 90% gasoline (E10) blend with no adverse effects. But, to exploit the potentially higher octane rating of E10 and its different combustion in engines, laminar burning velocity data for ethanol and its mixtures are needed. Ethanol is mostly simply made from the fermentation of sugars, but competition with food use means that second generation or cellulosic-ethanol needs to be exploited. Ethanol has been produced from cellulose for over 100 years, but there is now a rapid increase in the commercialisation of the process (http://en.wikipedia.org/wiki/Cellulosic_ethanol). Gaseous fuels from renewable sources depend on the processing route. The anaerobic digestion of waste (by mesophilic bacteria) produces biogas (which is 60-70%CH4, and 40-30%CO2), whilst pyrolysis of waste or biomass produces syngas (a partial oxidation process that gives typically 40% CO, 25% H2, 20% H2O, 15% CO2).

The transport sector accounts for a significant part of carbon emissions worldwide and in the UK about 20% of CO2 emissions are attributed to road transportation. Consequently, the need to mitigate the greenhouse effect of CO2 and reduce vehicle exhaust emissions has provided the driving force for developing cleaner more efficient vehicle powertrains and environmentally friendly fuels. Reducing consumption of petroleum-derived fuels has become one of the top priorities in the 21st century. As substitutes of the internal combustion engine have yet to overcome technical challenges to attain significant utilisation in the transport sector, the compression ignition diesel engine remains a very attractive powertrain option due to its high thermal efficiency. The International Energy Agency estimates that biofuels can grow to as much as 30% of the world's road transport fuel mix by 2050. Such fuels will include biodiesel and synthetic diesel fuels. In the same frame, alcohols such as bio-ethanol produced from non-food sources with reduced production costs and low CO2 emissions have been proposed as alternative fuels for direct blending with diesel, biodiesel or synthetic diesel. According to Shell, our industrial partner, ethanol made from Brazilian sugar cane produces around 70% lower CO2 emissions from production to use compared to gasoline. Therefore, the potential of ethanol-diesel blends (e-diesel) as alternative fuel for low carbon advanced diesel engines of today and tomorrow has become very important. The use of bio-derived fuel blends such as e-diesel also offers the benefit of compatibility with existing infrastructure. On the other hand, as the complexity of the properties of fuel blends increases, new phenomena that affect engine performance occur in the engine combustion chamber. Thus, improved scientific understanding is essential to overcome potential issues and/or gain potential benefits. One key phenomenon is the micro-explosion of multi-component fuels that is exceedingly possible to occur during spray atomisation and combustion in the case of fuel blends with difference of physical properties among the different fuels in the mixture. The micro-explosion of a miscible multi-component fuel droplet is due to the difference of volatility and boiling point among the different components. For an immiscible multi-component fuel droplet (emulsion droplet as routinely termed), the likelihood of micro-explosion will considerably increase if the lower-boiling-point component cannot dissolve in the mixture and disperse as micro-droplets inside the fuel droplet, such as in the case of e-diesel as the volume fraction of bioethanol increases. Micro-explosion in diesel engines has potentially significant implications on engine performance, combustion and emissions. The phenomenon offers a unique potential in optimising charge preparation in spray combustion systems, making the demands and design of fuel atomisation devices potentially more flexible. However, despite the potentially significant impact, fundamental understanding of the micro-explosion of fuel blends in diesel engines is still lacking. In the present study, we propose for the first time to carry out frontier research using both experimental and modelling techniques to investigate systematically the micro-explosion phenomenon and its effects on spray atomisation and combustion and emissions under realistic diesel engine conditions. The proposed research exploits the creativity that is highly likely to occur at the active interfaces and with the close collaboration between the experimental and modelling researchers with support from Shell which is one of the world's biggest distributors of biofuels. The research outcomes will be disseminated at top international conferences and journals. Development of this science base is vital for the UK to lead the world in advanced technologies of clean, efficient engines and sustainable low CO2 fuels today and tomorrow.