Brief introduction to combustion

Laser-based combustion diagnostics 2009 Brief introduction to combustion 1. Background Practical combustion processes are generally very complex since the chemistry involves hundreds of reactions, the fluid flow propagation most often is three-dimensional, and physical processes such as diffusion and radiation have a large impact on the flame conditions. Due to this complexity combustion is a demanding task for researchers, both from theoretical and experimental point of view. The progress in combustion research has been rapid the last decades. The development of computers with higher capacities has implied that numerical calculations can treat a larger number of chemical reactions in the reaction schemes, and that the calculated flow fields can be treated with higher spatial resolution. Also regarding experimental measurements there has been a fast progress led by the development of techniques based on lasers. With laser radiation interacting with molecules in the combustion process, signals from scattering and fluorescence can be analyzed to get information on, for example, temperatures and species concentrations in the interaction point. One important advantage of laser techniques is thus that the measurement can be made without disturbing the combustion process. Other advantages are the possibility of performing measurements with high spatial as well as temporal resolution. To be able to extract information with laser diagnostic techniques, some basic knowledge about combustion is needed. The purpose of the present text is to give you a basic knowledge on combustion of importance for the application of laser diagnostics to combustion systems. 2. Introduction Combustion processes have been a foundation for human activities and development for thousands of years. To be able to control fire was of fundamental importance for the survival of human beings. Also today, we have a society where combustion is of central importance, nowadays mainly for energy production, industrial processes, and transportation. Actually, more than 90% of the energy use in the world can be related to combustion processes, and over 80% to combustion from fossil fuels. This is illustrated in Figure 1. 20,49% 6,37% 2,16% 35,19% Coal Oil Gas Nuclear Hydro Biomass/Waste Other renewables 10,5% 0,51% 24,79% Fig. 1 Sources for the global energy use 2004.1 1 IEA, World Energy Outlook 2006, The International Energy Agency, 2006, available at http://www.iea.org/ 1 Laser-based combustion diagnostics 2009 Since fossil fuels are not renewable and pollutants from combustion processes contribute to environmental problems such as acid rain, smog, and the greenhouse effect there are clear motivations to decrease the use of combustion. It is thus important to increase the fundamental understanding of the processes in combustion. It can be seen in Figure 2 that the energy use in the world continuously grows, and with negative environmental effects in mind, the question is how to replace combustion by other energy sources. Other solutions, such as fuel cells, are promising but will not be economically competitive for decades. The debate about environmental consequences of using nuclear power have led to a decreased interest in further development and constructions of new nuclear power plants, and at the same time world-wide old plants are closed down. The energy from wind can contribute on a local scale but will never be a main source of energy. The energy from the sun continuously flows towards the earth but to be a strong competitor the economic aspects must be more favorable. Fig. 2 A forecast of the global energy demand by energy source.2 The quantity on the y-axis (Mtoe) designates Million ton of oil equivalents. It is quite clear that although combustion has negative environmental effects, it will be the main source of energy for decades. More efficient combustion technology will be developed resulting in lower emissions driven by more severe regulations. Hand in hand with higher consumption of fossil fuels we can expect that worse qualities of oil and coal must be used, for instance, with higher sulfur content. We can also expect that there will be large differences between industrialized countries using modern technology and developing countries, which will need energy but have limited resources of thinking of ecological and environmental consequences. These countries being more industrialized will have an increasing energy demand, and will use easy ways of getting such. Often the solution is combustion, and often not especially efficient and clean. Research and development in combustion will have important tasks: • To reduce emissions of pollutants, such as nitrogen oxides, sulfur oxides, soot particles, unburned hydrocarbons, etc. • To improve efficiency. Higher efficiency will lead to less fuel consumption, and fuel consumption is directly related to the emission of carbon dioxide. Often an improved efficiency also leads to lower pollutant concentrations. • To find alternative fuels and technologies. Often there is a trade-off between the use of such alternatives and economy. For example, hydrogen combustion is very clean since it only produces water, but the cost of hydrogen production has to be accounted for as well. 2 IEA, World Energy Outlook 2008, The International Energy Agency, 2008, available at http://www.iea.org/ 2 Laser-based combustion diagnostics 2009 3. What is combustion? Combustion takes place in a flame. It can for example be a candle light, a Bunsen burner flame or a propagating flame in a spark-ignition engine. Flames are characterized by: • • • • Exothermic reactions where the chemically bound energy is released through chemical reactions. Oxidation processes, where most often air (or more correctly the oxygen in the air) is used as an oxidizer. High temperatures. Temperatures are often above 2000 K in the combustion products, when they are heated by the liberated energy. Radiation. For example a candle light has a yellow colour due to the temperature radiation from hot soot particles. The blue-green colour is due to emission from the chemically excited radicals C2 and CH. The most obvious classification of flames is in premixed flames and diffusion flames. In a premixed flame the fuel and oxidant are mixed on a molecular level before combustion occurs. In a diffusion flame the fuel and the oxidant are initially separated and do not mix until the moment of combustion. This will be discussed more in detail further on. Another way of characterizing flames is to divide them into propagating flames and stationary flames. A propagating flame moves towards the unburned gas, and an example of such a situation is a moving flame front in a spark-ignition engine. Information about the combustion process must then be captured instantaneously or on a statistical basis. A stationary flame burns at a fixed place on some kind of flame holder. A stationary flame can be stable i.e. the physical parameters at a spatial position are constant in time which means that measurements using for example laser diagnostics both can be averaged over a long time, and also can be repeated at a later time. On such a burner the gas flow is matched to the burning velocity of the flame, which has an opposite direction, as shown in Figure 3. Fundamental work on combustion problems is often made in stable stationary flames with simplified geometries and well-adjusted initial conditions. Fundamental research on combustion is normally also done in laminar flames. In such flames the volume elements of the gas have constant flow direction in time. However, in real combustion applications flames are generally turbulent, where the flow field can be characterized by a mean flow velocity but where the volume elements of the gas have different flow directions, also in time. Often high turbulence velocity is demanded since the combustion velocity is increased and thereby the combustion can be finished in a shorter time. This is advantageous in processes where the time for the combustion to be completed is limited, such as in spark-ignition engines. 4. Premixed flames 4.1 Flame structure A premixed flame can be divided into the following zones; unburned gas zone, preheat zone, reaction zone, and product zone. This is illustrated in Figure 3. In the preheat zone, the gas mixture is heated by heat conduction from the reaction zone and only a small amount of heat is released by chemical reactions. As the molecules flow towards the reaction zone, the temperature increases which also means that the reaction rates increase. The separation of the preheat zone and the reaction zone is often defined as the position at which there is an inflexion point in the temperature profile, as shown in Figure 4. A stable stationary flame is achieved when the burning velocity, S, matches the gas flow velocity, v. 3 Laser-based combustion diagnostics 2009 Fig. 3 General structure of a premixed Fig. 4 Temperature profile of a premixed laminar flame on a porous-plug burner. flame on a porous-plug burner. The exothermic reactions mainly occur in the reaction zone. The thickness of such a zone is on the order of some hundred microns for an atmospheric-pressure hydrocarbon flame, and the thickness is inversely proportional to pressure. The fast release of energy in such a narrow region leads to a very steep temperature gradient. After the reaction zone most reactions have occurred and the product zone is reached. A reaction zone is often called a flame front, and the product zone is also called "burned-gas zone" or "post-flame zone". The flame on the burner shown in Figure 3 is said to be one-dimensional. In each specific volume element on a certain height above the burner, the conditions regarding temperatures and concentrations are the same. The only change of these parameters is as a function of height above the burner. Thus, by measuring at different heights in such a flame it is possible to follow the combustion process. 4.2 Equivalence ratio In a premixed flame the relation between the amount of fuel and oxidant can be regulated, and this ratio also affects the composition in the product gas from a combustion process. This relationship is called the stoichiometry of the flame. Let us choose a certain fuel, for example ethane. Assume complete combustion to water and carbon dioxide. We can then calculate that there is a need of 3.5 moles oxygen for each mole of ethane, i.e. 1 C2H6 + 3.5 O2 → 2 CO2 + 3 H2O If a mixture is prepared with a molar ratio of 1:3.5 between ethane and oxygen, then we have a stoichiometric mixture. We can also say that we have an equivalence ratio Φ=1. The definition of the equivalence ratio is Φ= ( # moles ( # moles fuel / # moles oxygen ) stoichiometric mixture fuel / # moles oxygen )real mixture Assuming that we have a mixture of ethane and oxygen with a ratio on molar basis of 3:7, it means that the equivalence ratio is Φ = (3/7) / (1/3.5) = 1.5 A flame with a mixture having Φ>1 is called (fuel-)rich, while a flame with Φ<1 is called (fuel-)lean. Practical combustion processes take place between fuel and air. Air consists of 4 Laser-based combustion diagnostics 2009 ~21% oxygen, ~78% nitrogen, and ~1% argon. For simplicity this mixture is approximated to 21% oxygen and 79% nitrogen, since argon as well as nitrogen is inert. The molar relation between nitrogen and oxygen will then be (1-0.21)/0.21=3.76. Thus a stoichiometric mixture of ethane and air will, following the example above, contain 3.5 moles of oxygen and 3.5⋅3.76=13.16 moles of nitrogen for each mole of ethane. The ethane concentration will be 1/(1+3.5+13.16) = 5.7%. The composition of the product gas depends on the equivalence ratio, as shown in Figure 5. It is clearly demonstrated that all of the oxygen is not used for the combustion in fuel-lean flames. It partly survives into the product gases. Another observation is that fuelrich flames have rather high concentrations of carbon monoxide and hydrogen in the product gas. Finally, at stoichiometric conditions, not only carbon dioxide and water are formed. You can also observe concentrations in the range 0.005 - 0.015 of oxygen, carbon monoxide and hydrogen. This is in fact a result of chemical equilibria that can be related to a maximization of the entropy of the thermodynamic system. Fig. 5 Mole fraction of CO, H2, and O2 versus equivalence ratio for premixed ethane/air flames. Figure 6 shows the corresponding temperature profile. Maximum temperature (∼2250 K) is at an equivalence ratio slightly above 1. This behavior is typical for many hydrocarbon flames. Fig. 6 Adiabatic flame temperature vs equivalence ratio in premixed ethane/air flames. 5 Laser-based combustion diagnostics 2009 4.3 Flame characteristics A premixed flame can be characterised in terms of its adiabatic flame temperature, flammability limits, and burning velocity. Some of these data are presented in Table 1. Table 1 Characteristics for premixed flames of different fuel/air mixtures. Adiabatic flame Maximum laminar Fuel/oxidant mixture temperature (K) burning velocity (m/s) Methane (CH4) / air 2222 0.45 Ethane (C2H6) / air 2244 0.40 Propane (C3H8) / air 2250 0.43 Ethene (C2H4) / air 2375 0.75 Acetylene (C2H2) / air 2513 1.58 Hydrogen (H2) / air 2380 3.1 Methane (CH4) / oxygen 3010 4.5 Acetylene (C2H2) / oxygen 3431 11.4 Hydrogen (H2) / oxygen 3083 11.0 The adiabatic flame temperature is the maximum temperature that a mixture can reach from given conditions of initial temperature and pressure, i.e. without any losses in form of cooling and radiation. From the table it can be seen that common hydrocarbon fuels burning with air have adiabatic flame temperatures in the range 2200 - 2300 K. When the air is replaced by oxygen the temperature is generally increased to over 3000 K. The main reason is that the released energy does not need to heat up the inert nitrogen in the product gas. The maximum laminar burning velocity is the propagation velocity for a planar reaction zone towards the unburned gases. As can be seen in the table, hydrocarbon/air mixtures normally have burning velocities of around half a meter per second. This velocity is strongly dependent on the temperature, and that is the main reason why fuel mixtures burning with oxygen have burning velocities that are 5-10 times higher. The flame front moves faster at a higher temperature because of faster chemical reactions and an increased radical diffusion. The flammability limit expresses a range of stoichimetries in which it is possible to establish a flame front that propagates through unburned gas. This is shown for various mixtures in Figure 7. Fig. 7 Flammabilty limits for different fuel-oxidant mixtures. 6 Laser-based combustion diagnostics 2009 4.4 Flame chemistry The previously mentioned reaction, 1 C2H6 + 3.5 O2 → 2 CO2 + 3 H2O, is a so-called global reaction that has no connection to a real reaction. When looking at the combustion of ethane (or another hydrocarbon molecule), the first reaction that occurs is the abstraction of a hydrogen atom from the molecule by a radical. As an example the first reaction can be C2H6 + O → C2H5 + OH. This is an example of an elementary reaction, i.e. a reaction that occurs in reality between the colliding partners. Even for a simpler system, like the combustion of methane, a detailed mechanism may include ~150 reactions. A detailed mechanism for combustion of hydrogen with oxygen normally utilizes 19 reactions, as shown in Table 2. The temperature-dependent reaction rate constant (k) for a chemical reaction is given by the equation: k = AT n exp ( − Ea RT ) , where A designates the pre-exponential factor, n the temperature exponent, and Ea the activation energy. This equation is a modification of the classical Arrhenius equation, k = A ' T exp ( − Ea RT ) , whose two-parameter (A’ and Ea) representation sometimes is inadequate in combustion chemistry where rate constants have to be accurately determined over a very wide temperature range. Table 2 Reaction mechanism for the H2/O2 system with pre-exponential factor (A), temperature exponent (n) and activation energy (Ea) for each elementary reaction. By looking at the reactions in the table it is clear that combustion of hydrogen with oxygen to form water also includes formation of intermediate products such as O and H atoms, OH, H2O2, and HO2. It should be emphasized that, from a laser-diagnostic point of view, it is not 7 Laser-based combustion diagnostics 2009 only of interest to probe reactants and products of the global reaction, i.e. H2, O2 and H2O. Very often it is actually more interesting to investigate the intermediate species formed in the reaction zone because this region is the most important region from a chemical-kinetic point of view. In Figure 8a it can be observed how the reaction of methane and oxygen, along a reaction coordinate, turns into products, which are not merely carbon dioxide and water. The reaction zone is the region where the heat release, q, occurs. After the reaction zone, where the species may not be in chemical equilibrium, the concentrations of different species approach their equilibrium concentrations. Note that there is a substantial amount of carbon monoxide also at stoichiometric condition. In Figure 8b some reaction intermediates are shown from the same flame conditions as for the profiles shown in Figure 8a. It can be seen that there are radicals existing far into the product zone (H, O, OH), but also those that are present almost exclusively in the reaction zone (CH3, CH2O). The HO2 radical can be seen prior to the reaction zone, and the reason is that hydrogen atoms diffuse fast from the reaction zone towards the flow and reacts with oxygen to produce HO2. The most important reaction for premixed flames is, however, the reaction H + O2 → OH + O. This reaction dominates the flame propagation of premixed flames. (a) (b) Fig. 8 Major (a) and minor (b) species concentrations in a premixed stoichiometric methane/air flame. 8 Laser-based combustion diagnostics 2009 5. Radiation Radiation is a characteristic feature of flames. A lot of the radiation occurs in the infrared region and has its origin mainly in carbon dioxide and water. But when looking at for example a candle light, the most obvious radiation is the yellow light from the central part of the flame. The reason for this radiation is soot particles emitting temperature radiation which approximately has a spectral distribution according to the Planck distribution law. Another well-known type of radiation from a candle light is the blue light from outer regions at lower positions in the flame. This blue radiation is a characteristic feature of a reaction zone. A spectrum recorded in the reaction zone of a flame can appear as in Figure 9, with different emission bands from OH, CH, and C2. (The spectral regions are approximately: UV < 400 nm, violet 400 nm - 440 nm; blue 440 nm - 500 nm and green 500 nm - 570 nm.) Fig. 9 Emission spectrum from the reaction zone of a premixed acetylene/air flame. Since the thermal excitation to excited electronic states is very weak for most species even at flame temperatures, this is not the way excited radicals are produced. Some reactions of the reactive species are sufficiently exothermic for the products to be formed in excited electronic states. The emission after excitation by chemical reactions is called chemiluminescence. For C2, the three emission bands originate from the same excited electronic state. The reason why the bands yet occur at very different wavelengths is that the transitions originate from different vibrational states. Excited OH radicals are formed from the reaction CH + O2 → CO + OH*, where the star superscript designates electronic excitation. The observed OH chemiluminescence hence corresponds to the deexcitation of OH* to its ground state, in which the excess energy is lost as radiation. This process can be described as OH* → OH + hc/λ , where hc/λ corresponds to the energy of the radiation (h is Planck’s constant, c is the speed of light, and λ is the wavelength of the emitted radiation, i.e. 310 nm (see Fig. 9)). Excited CH and C2 radicals, are formed from C2 + OH → CO + CH* and CH2 + C → C2* + H2 , respectively. 9 Laser-based combustion diagnostics 2009 6. Diffusion flames In a diffusion flame the fuel and the oxidant are initially separated and do not mix until the moment of combustion. This is illustrated in Figure 10. Combustion takes place in zones where fuel and air meet if the temperature is high enough. This is a typical situation when combustion of liquids and solids are burned. If a fuel droplet burns the combustion takes place outside the droplet in a layer between vaporized fuel from the droplet and oxygen from the air diffusing towards each other. Fig. 10 Schematic illustration of a laminar diffusion flame. The variations in concentrations of different species are shown in Figure 11. In Figure 11a it can be seen that there is a position where methane and oxygen both have very low concentrations. This is the reaction zone. Naturally, the water concentration has its highest value here as well. The nitrogen does not react and thereby it diffuses further towards the central axis of the flame. In Figure 11b additional curves are shown. The carbon dioxide profiles follow the water profiles as expected. Note that the profiles for carbon monoxide and hydrogen peaks at lateral positions of 5-6 mm, which is closer to the central axis than the profile for carbon dioxide. Also, a curve showing the equivalence ratio has been indicated in the figure (note the logarithmic scale). In diffusion flames there is no general equivalence ratio for a flame, only a local value can be given at a certain spatial position in the flame. Thus, at the center of the flame, the local equivalence ratio is extremely high. Fig. 11 Concentration profiles in a methane/air diffusion flame. (a) Major species concentrations. (b) Major species concentrations and local equivalence ratios. 10