Challenge on Combustion

Abstract:

Combustion is known as a reaction between fuel and oxidizer to generate heat and power which provide us with a more convenient life such as a moving vehicle, boilers and electricity. However, pollutant by combustion is a major environmental problem that directly relates to health hazards.Therefore, understanding of turbulent combustion is needed to decrease pollution and this is a real challenge. The production of contaminants from combustion,concerns about pollutant emissions, such as smoke and oxides of nitrogen, apart from this formation of soot is another critical issue in combustion which has to be reduced because it can cause a wide array of serious health issues, including respiratory issues, shortness of breath, bronchitis, asthma, stroke, heart attack, cancer, and premature death. In order to tackle the issue we require knowledge of trace species in combustion processes. These species are produced by finite-rate chemical kinetic mechanisms that involve large numbers of elementary steps. Now the computer simulations are a truly important research tool to support the theory of combustion and understand phenomena from experiments. It is hoped that theory, experiment and simulation can allow a better control and understanding of combustion. Therfore, combustion simulation is performed on the combustor model and variation in the mass fraction of the different species (CO2, H2O, CH4, N2, O2, NOx emissions & Soot formation) in the simulation using line probes at different locations of the combustor is plotted. After basic simulation we have observed that Nox or pollutants and soot is formed in the combustor. Such formation has harmful effects on the environment and humans. The stringent government norms also demand the least formation of Nox and soot and to satisfy those requirements, we need to check the effect of adding the water in the fuel. The water content in the fuel is varied from 5% to 30% by mole and observed the effect of it on the results(line plots and contours). For this, a parametric study approach is adopted.

The combustion simulaton type used here is Non-primixed combustion. The reaction rate of pollutant formation in combustor is determined by both chemical kinetics and turbulence and the mixing of chemical species and combustion of gaseous fuel in combustor, burning methane (CH4) in air. The combustion is simulated by steady-state governing equations with pressure based solver and the effect of turbulence on the mean flow field was accounted for using the standard k–є turbulence model. The combustion is modeled using single-step(one- step) reaction mechanism and NOx model assuming conversion of the fuel to CO2, H2O, NO and NO2 etc.( i.e Species transport model in ansys fluent). In this model volumetric reaction is enabled so that reactions takes place only in the volume of the combustor. Moreover, inlet diffusion is enabled because when we are using pressure based solver, diffuision is not calculated at the inlet, therefore need to be calculated. Along with it, diffuision energy source is enabled.The reaction rates are modeled using eddy-dissipation model. Methane-air is defined as mixture material in fluent for the first case and then methane-air-2step as a mixture material for the second case which allows to add water in the fuel and one-step soot model to calculate the soot formation. The flow parameters like velocity,temperature, mass fraction is defined prior to the simulation. The variation of mass fraction of each species were numerically found inside the combustion chamber during combustion process and plotted on plane(XY) contour in post processing. On adding water in methane it was found that mass fraction of soot and pollutant No. inside the combustion chamber has reduced to some extent and this decrease in value is observed on every single rise in percentage(5%-30%) of water content added in the fuel by mole.

Theory:

Soot Formation:

Soot is a deep black powdery or flaky substance which largely consists of amorphous carbon formed due to incomplete burning of organic matter.It is more properly restricted to the product of the gas-phase combustion process. Formation of soot can take place in all kinds of practical combustion systems, especially in systems which operate on a diffusion flame concept. In technical combustion processes, soot usually forms at temperatures from 1000 to 2500°C. Soot particle's mean diameter is usually a few hundred Angstroms.These soot particles often stick together and form long branched or straight chain aggregates and their number densities are very similar even under widely varied conditions,a consequence of their formation mechanism. Soot particles are very good absorbers. It depends on their history (time, temperature, concentrations), how much of other substances they absorb until they reach the open atmosphere. The soot particles are a penalty for the entire world where they contribute to climate warming, ice melting and health hazard which causes various types of cancer and lung disease. Soot particles formed during combustion can be removed only by oxidation, by reactions with OH, O atoms or O2. Experiments show that this process is slow compared to the oxidation of hydrocarbon molecules and the necessary residence times for particle removal strongly increase with particle size.

Chemical Kinetics :

Basically, Chemical Kinetics is the description of the rate of a chemical reaction. This is the rate at which the reactants are transformed into products. Since a rate is a change in quantity that occurs with time, the change we are most concerned with is the change in the concentration of our contaminants into new chemical compounds

Simply,

• Fuel + air = Products (CO, CO2)------------>Global Reaction
• However, in reality , this global reaction is not the only reaction that take place.

Chemical Kinetics involves the experimental study of reaction rates in order to infer about the kinetic mechanisms for chemical conversion of reactants into products . For any given chemical reaction, the mechanism refers to the sequence of elementary steps by which overall chemical change occurs and an elementary step refers to the passing of a reactant or reaction complex through a single-transition state to a chemical form with a defined and detectable lifetime.This can be inferred as elementary reaction. If chemical conversion of reactants to products involves more than one elementary step, then the chemical structures that exist after each elementary step preceding final product formation are defined as reaction intermediate in common words when fuel decomposes, some part of fuel combine with O2, as explained in below figure. The whole stuff which occurs between the reactant to product is termed as Reaction mechanism and the study of this Reaction mechanism is Chemical Kinetics.

Figure: Skeletal structure of Combustion chamber

So, fluent uses a chemical kinetic solver which tries to simulate the elementary reactions taking place between Intermediate Species and the different path that the system takes to predict the final product. These reaction pathways are the function of operating conditions and function of time,meaning, how far the simulation is taking place and how long the combustion is taking place, depending on it final product species will be completely different. That is why this approach is very predictable, if reaction mechanism and pathways are computed accurately, then this process can give accurate results of thermodynamics property after the combustion process.

Assumed Reaction in Combustion Chamber:

CH4 + 2(O2 +3.76 N2) -------> CO2 + 2H2O + 7.52N2

Importance of Chemical Kinetics:

The reaction pathways determine:

1. When the combustion starts?(i.e ignition delay)- as elementary reactions are very crucial to start combustion that basically determine how the temperature is going to increase in computational domain, which ultimately effects ignition delay.
2. Heat release rate( rate of work done).
3. Can able to compute desired emission( How and how much pollutants are produced?)

Note: While we are simulating combustion of a any fuel, we have to make sure that the Reaction Mechanism should contain respective Elementary reactions and Elementary species, otherwise results will be wrong and even performance number will be wrong. For example, if we are simulating a low-temperature combustion we need to have the reactions which take place at low temperature, else the prediction will be completely different.

Elementary reaction types:

1. Bio-molecular Reactions(Most common in combustion)

• A + B ⇌ C + D

2. Uni - Molecular Reaction

• A ⇌B (Isomerization)
• A ⇌ B + C (Decomposition)

Ex: O2 è O + O

3.Termolecular Reactions

• A + B + C ⇌ C + M

(M = third body species acts as a catalyst)

Note: If we know how fast these reactions take place, then we can predict the reaction pathway. It means that if we design the structure ( skeletal of combustion chamber) of the reaction mechanism along with elementary reactions and elementary species in it and give as input to the solver then we can able to solve the emissions and thermodynamics quantities associated with it. However, there will be 100’s and 1000’s of elementary reactions and elementary species, for this we need a computer program. In order to do this we need, operating conditions – pressure, temperature and composition of the system and also Rate of reaction.

Rates of reactions for multistep mechanism

Here

• [O2] = concentration of oxygen, similarly for [H2],[H2O], etc
• Kf1= rate of forward reaction
• Kr2= rate of reverse reaction

Kf1 and Kf2, both are the constant of proportionality of O.D.E

All the above different O.D.E’s are called the system of coupled O.D.E. So, in theses O.D.E’s we have to find the concentration of each molecule ( [O2], [H2], [H2O]…etc) at each time, so this can be solved by Numerically Integrating entire O.D.E system.

Working of Basic Combustion Solver:

In some reaction mechanisms there are 1000’s of species and 5000-10000’s of reactions, so this kind of mechanism are not solved by pen and paper, but there needs to be computer program, that uses compact notations by combustion solver like Cantara, Kin( a Ansys product), converged CFD etc.

Compact notation:

Here;

• V’ji = for reactants
• V”ji = for products
• V’ji matrix represents Psychometric Coefficient matrix, which basically keep the track of no.of moles of species ‘j’ in reaction ‘i’ in reactant side and similarly with (V”ji)

Vji = V”ji - V’ji

Therfore, it is can be written as

Where

qi = net reaction rate which calculate for each reaction

[Xj] = concentration per unit volume of each and every species

Production rate ( d[Xj]/dt):

Here;

Wj = Production rate ( d[Xj]/dt)

L = No of reaction

By this we are able to get into a generic form which can be fed into computer and the whole system can be integrated by any integration package for example Mat lab(using ODE 45), Cantra.

But, ODE systems are stiff, it means that some species changes rapidly fast while few species slow, which means that it indirectly effects the numerical integration ‘dt’(time step), picking it will be critical, for example if dt is too big , system become unstable and if dt is small, it takes ever to reach a particular answer, therefore Cantara, mat lab support stiff system in reacting system but, if it system takes explicit integration scheme by adopting for example Euler integration method then it would be not work, hence Implicit ( backward integration) method is preferable.

Sample Review of thermal data file and Mechanism file:

Thermal data file:

• Simple thermal data file contain simple polynomial function which can calculate enthalpy,entropy, Specfic Heat etc

Mechanism file:

Here;

M = Third body ,

However when the the third body is present, then net reaction rate will be

where,

takes the value 2.40 or 15.40 or 2.00 which is show in above mech file, which acts a multiplier that enhance the reaction.

Calculation of Kfi and Kri:

A = Pre Exponential factor

T = Operating Temperature

Rc = Characteristic Gas constant

E = Activation Energy

• S and H are the entropy and Enthalpy , which can be calculated from therm file

Primary Classification of possible Combustion Simulation:

1. Based on Mixing

• Non Premixed Combustion ( Direct Injection or Late Injection)
• Premixed Combustion(Carburetor,Burner)
• Partiallypre mixed (Example: IC engine where ignition takesd place earlybut not very early)

2. Based on Phase

• Fluid Phase (Volumetric Reactions)
• Wall ( Surface Reactions)
• Particles ( Surface Reaction-Coal combustion)
• Porous Region ( Example: After treatment System)

Basics of Combustion Simulatin:

In CFD generally,we need to solve for

• Mass
• Momentum
• Energy Equations
• Turbulence Equation(K, K-epsilon etc)
• Additionally in combustion we have Species Transport Equation(i.e)

Here;

Yi = Mass fraction, 'i' is No.of species.

Diffusion:

Diffusion is a combination of Laminar adn Turbulence Diffusion. Typically for engineering applications( IC engines , Gas turbines, Large Burners) Turbulent diffusion is so high that laminar diffusion can be neglected. So, to do this ,we use fick's Law to model the diffusion, which is basically called as Dilute approximation

What is Dilute approximation?

During combustion, behaviour of gas diffusion is different because diffusion co-efficient are different with each other. For example, CO2 diffuses in O2,N2, CO2 as ambient fluid, therefore we have three differnt coefficient for CO2 w.r.t O2,N2,CO2, hence total 3x3 total 9 different coefficient, then we can assume for combustion simulation 100's of species which has many no.of diffusion co-efficient and if the simulation is totally tubulent we can ignore laminar diffuision,to calculate such only turbulent coefficient we use dilute approximation which eliminate laminar diffusion coefficient. In Dilute approximation, it analyze each and every species's diffusion coefficient and how that species diffuses into that entire mixture. For example it evaluate species 'i' , how it diffuses with other species. Instead , if we look how each and every species diffuses differently in the presence of a different species, we need multicomponent diffusion properties which is very accurate and this is required when the flow is purely laminar(example laminar flames), where fick's law is not valid and we need to adopt multicomponent diffusion properties, that then cost us high too.

Combustion Source calculation:

1. Finite Rate Chemistry: It takes into account of chemical kinetics and calculate rate of production of each and every species to solve complete OD system which can cost high and the process is slow as well

2. Mixing based Model:

• There are many classification of this model, which uses turbulence of the Air and fuel mixer to ignite with the help of K,Epsilon models etc, that calculate time scale factor(Ex: X sec) after which combustion starts during mixing . this type of phenomena is called as Turbulence Chemistry Interaction model, in other words it can be called as mixed is burnt. It means that these model doesn't require ignition source and also faster, but require user experience (when to use these models).
• One such type of model is Eddy Dissipation Model Which uses K and epsilon to calculate time for each reaction. This model is used when have inlet and assume a steady state tempeature distribution at outlet.

3. Finite Rate Eddy Dissipation Model: Combination of tboth finite rate chemistry and eddy Dissipation model. It is used to prevent instantaneous burning. It is helpful when you need ignition delay, because in eddy dissipation model sometimes in a cell when Air-fuel mix combination immediately starts and if Air-fuel mix ratio is good then even before mixing combustion takes place.

Simulation Process:

Geometry:

Combuster Chamber is modeled in any modelling software or in space-claim of Ansys Fluent and Transfered into 2D Axi-symmetric geometry in which we are assuming that the flow is axi symmetry, which means, from the below figure consider Y-Z plane, if we go along the circumference of the circle of a particular slice, their is no change in properties( rate of change of velocity or temperature or other gradients etc with respect to angle(teta) is zero, that is called Axi-symmetry Approximation.

Complete Chamber:

2D Geometry:

2D Geometry(zoom):

2D-Geometry which is revolved to some angle with a zoom image:

Meshing:

• Element Type : Triangular
• Element Size : 5e-3 m
• No of Elements : 39141
• No of Nodes : 20021

SOLVING:

Solver:

• Type : Pressure-based
• Time : Steady
• 2D space : Axi-symmetric

Model:

• Viscous : K-epsilon(standard)
• Species : Species Transport
• Mixture Material : case 1-Methane+ air

case 2 - Methane+air+2step

• Turbulance- chemistry : Eddy Dissipation

Boundary Conditions:

	Velocity(m/s2)	Temperature(K)	Mass Fraction(CH4)	Mass Fraction(O2)	Pressure(pascal)
Air- inlet	0.5	300	Case 1: 0 Case 2: 0	Case 1: 0.23 Case 2: Parametric	-
Fuel Inlet	80	300	Case 1: 1 Case 2 : Parametric	Case 1: 0 Case 2 : Parametric	-
Outlet	-	300	-	-	0

• After set up of all the conditions, as shown mentioned above, the solution is initialized and run for some iterations till residuals plot reaches to steady state.

RESULTS:

The results are observed on a plane with variation in mass fraction of species and other parmeters, by ploting line probes across the plane along with graphs w.r.t to the line probes where Y distance is taken on the X-axis and the mass fraction, temperature or other parameters variation on Y-axis . The below figures representing the plane with line probes on it at varies points in X direction and the counters of mass fraction & other parametrs along with graphs.

Case-1:

Case-2:

Parametric Study

Note: 0% H2O addition resultscan be taken from case 1.

• Addition of 5% H2O :

• Addition of 20% H2O :

• Addition of 30% H2O :

On adding water to methane at fuel inlet, it has been seen a decrease in soot and pollutant No. in terms of mass fraction which can be noticed by plotting the graphs shown below by taking maximum value of MF soot and MF pollutant No into account.

Conclusion: Therefore, it can be concluded that addition of water in the fuel can reduce the soot formation and pollutant No. content to some extent and on furthur addition it may effect the performation of combustion too- from previous research data.