1 Questions

1.1 Stoichiometric flame

Compute the adiabatic flame temperature at constant pressure of stoichiometric methane burning with air using three classes of assumptions and methods: Table 1: Methods to evaluate adiabatic flame temperatures Method Number of reactions Assumptions on 0 1 2 3 1 1 1 n Constant and equal Constant Variable Variable Technique Analytic Analytic Tables of enthalpy Code

The reference temperature for thermodynamic data is chosen at Repeat the exercice for two inlet temperatures: first

K. K and second (a value

typical of gas turbines) K. For method 2, Annex in 3 provides values of molar sensible enthalpies versus temperatures for all required species.

1.2 Lean flames

Compute the adiabatic flame temperature at constant pressure of lean methane-air flames as a function of equivalence ratio using method 1 in Table 1.

2 Solutions
2.1 Stoichiometric flame
Methane and air react through: (1)

with moles of for one mole of in air. Writing Eq. (1) implicitely assumes that all other chemical reactions are neglected. This is not a major problem to compute flame temperatures when they remain low enough. At higher temperatures, however, it leads to neglecting dissociations (for example molecules being broken into and ) and to large errors on adiabatic flame temperatures. These errors can only be suppressed by taking more species and reactions into account (see 2.1.4). The mass stoichiometric ratio is . The system goes from state 1 (unburnt) to state 2 (burnt) at constant pressure so that its total enthalpy is constant. The derivation is written here using molar quantities for convenience but it can also be written using mass quantities. The enthalpy of moles of a given species is: (2)

where is the molar heat capacity and is the heat of formation of one mole of species [Poinsot,2005]. In the present case, the total enthalpy of state 1 is:

(3)

while the total enthalpy of state 2 is :

(4)

Writing that leads to:

, using the fact that all pure species have zero formation enthalpies

(5)

where

is the molar heat of reaction. Eq. (5) could also have been written directly by stating that the difference between the sensible enthalpies of states 2 and 1 was the heat of reaction .1At

this stage must be obtained from tables found in textbooks [Williams, 1985; Kuo, 1986] or in thermodynamic tables [Stull,1971]. For the present flame: (6)

which means that burning one mole of

releases

kJ. Solving Eq. (5) gives the

adiabatic flame temperature . This can be simplified depending now on the form used for the enthalpy changes versus temperatures.

2.1.1 Method 0: analytic The easiest method to solve Eq. (5) is to assume that all heat capacities are constant and equal to a value . Eq. (5) becomes: (7)

or (8)

The main issue is then to choose Here, since (for example

! Which gas and at which temperature should be used ? of at a 'reasonable' temperature so that

dominates in fresh and burnt gas, the K) can be taken:

K when when K. This is far from the exact answer.

K and

2.1.2 Method 1: analytic A better method to solve for in Eq. (5) is to assume that the heat capacities are constant: they do not depend on temperature (but they still depend on the species type ). Eq. (5) becomes:

(9)

which is a simple linear equation in . In the present case, the issue becomes to choose values for the heat capacities which change significantly with temperature and have been assumed to be constant here. Some arbitrariness is necessary. Tables 2 (for and 3 (for K) show the influence of the temperature used to evaluate . For gases such as or K) on the

final flame temperature values of

remember that low temperature

must match usual thermodynamic predictions for diatomic gases:

with so that J/moleK. Table 2: Adiabatic flame temperature of stoichiometric methane-air flames assuming constant heat capacities and evaluating them at an arbitrary temperature . Cold

reactant temperature K. Note that for this case, since , the term is almost zero. The exact answer (to be determined in 2.1.4) is K without dissociations and K with dissociations.

K 298.15 1000

J/moleK 35,8 72,0

J/moleK 33,6 41,2

J/moleK 29,1 32,6

J/moleK 37,1 54,4

J/moleK 29,3 34,8

K 2734 2398

Table 3: Adiabatic flame temperature of stoichiometric methane-air flames assuming constant heat capacities and evaluating them at an arbitrary temperature . Cold

reactant temperature

K. The exact answer (to be determined in 2.1.4) is and K with dissociations.

K 298.15 K 1000 K

J/moleK 35,8 72,0

J/moleK 33,6 41,2

J/moleK 29,1 32,6

J/moleK 37,1 54,4

J/moleK 29,3 34,8

K 3070 K 2703 K

Obviously, this analytic method works but it depends too strongly on the choice of at which heat capacities are evaluated to provide a quantitative and precise result. Removing the assumption of constant heat capacities is needed and can be achieved as described in the next section.

2.1.3 Method 2: using Tables Solving Eq. (5) can be done without assuming that the heat capacities are constant if tables for the changes of sensible enthalpy are available. Such tables may also be found in textbooks or in specialized books. A part of such tables is given in 3: the table provides values of vs temperature for the main species. Calling at temperature , Eq. (5) can be recast as: the sensible molar enthalpy of species

(10)

which is an implicit equation for given below. Let us call:

. A simple iterative method using tables for

is

(11)

so that Eq. (10) becomes: (12)

If the reactant temperature is K. Table 4: Sensible enthalpy

K,

is zero so that the equation to solve is for temperatures from to

. Table 4 shows values of

for main species in methane-air combustion.

T (K)

2200 2300 2400 2500 2600 743 790 834 879 923

Obviously the value of when between and K gives 2.1.4. If the reactant temperature is of 3: so that

K is close to K. Linear interpolation K which is exactly the value obtained numerically in

K, the enthalpy of the fresh gases must be such that

is read from the table . Reading

and interpolating Table 4 shows that value obtained numerically in 2.1.4.

K is the solution. Again, this is exactly the

2.1.4 Method 3: using equilibrium codes Even though using thermodynamic tables removes ambiguities on the choice of heat capacities and allows to take into account variable with temperature and species type, methods presented in the previous section still assume that only one reaction is sufficient to describe the thermochemistry of the reaction. This is generally not true, especially at high temperatures where dissociations take place, break molecules and diminish temperature. To obtain precise results for such cases, equilibrium codes are needed. Such codes can be bought or obtained as freewares from certain laboratories. Here EQUIL was used to solve the previous problem (Table 5). Two sets of species:

A simple chemistry case with only as main species A complex chemistry case taking into account dissociations through the species

which are the species used in the GRI Mechanism [Frenklach,1995]. Note that the kinetic reactions are NOT needed for equilibrium computations: only the list of species is required. Table 5: Adiabatic flame temperature of stoichiometric methane-air flames using an equilibrium code (EQUIL of the CHEMKIN package) and all species of GRI Mechanism [Frenklach,1995]. Reactant temperature Final flame temperature Final flame temperature

Simple chemistry Complex chemistry 298.15 K 600 K 2327 K 2547 K 2225 K 2367 K

Results illustrate the importance of dissociations: when complex chemistry and dissociations are accounted for, the final flame temperature goes down from K to K for

K and from K to K for K. Dissociations are important and must be taken into account in the present example because the maximum temperatures are high (typically larger than K). However, for lean and/or cool flames, the effect of dissociation is much smaller and a simple estimate based on a limited number of species is sufficient.

2.2 Lean flames

If the equivalence ratio is less than unity, not enough the present written: moles will be available to burn

moles. The chemical reaction controlling methane-air combustion can then be (13)

for lean flames where

is a factor smaller than unity. The equivalence ratio of the reactant is (14)

so that reactants

is actually equal to the equivalence ratio is equal to the enthalpy of products

. Writing again that the enthalpy of leads to:

(15 )

Knowing

, this equation gives

. For the present case, we limit ourselves to the method 1, at K are used (see Table 2) so that

assuming constant heat capacities. The values of Eq. (15) becomes:

(16)

which gives

as a function of equivalence ratio: (17)

Figure 1: Variations of adiabatic flame temperature versus equivalence ratio for lean

methane-air flames. Constant heat capacities

Figure 1 displays variations of versus for the two cases stoichiometric flames, the adiabatic values obtained in Table 2 ( recovered.

and K) and 3 (

K. For K) are

3. Annex: enthalpy tables

Table 5: Enthalpy differences combustion ( T (K) 298.15 600 2200 2300 2400 2500 2600 2700 0 13,1 0 9,2 0 12,9

for main species in methane-air ).

0 10,5

0 8,9

142,7 66,8 103,5 83,1 63,4 152,4 70,6 109,6 88,4 67

162,1 74,4 115,8 93,7 70,6 172 78,3 122 99,1 74,3 78 81,6

181,9 82,3 128,1 104,5 191,9 86,1 134,2 110

References