16.4.5 Modeling Coal Combustion Using the Non-Premixed Model

If your model involves coal combustion, the fuel and secondary stream compositions can be input in one of several ways. You can use a single mixture fraction (fuel stream) to represent the coal, defining the fuel composition as a mixture of volatiles and char (solid carbon). Alternatively, you can use two mixture fractions (fuel and secondary streams), defining the volatiles and char separately. In two-mixture-fraction models for coal combustion, the fuel stream represents the char and the secondary stream represents volatiles. This section describes the modeling options and special input procedures for coal combustion models using the non-premixed approach.

There are three options for coal combustion:

• When coal is the only fuel in the system, you can model the coal using two mixture fractions, where the primary stream represents the char and the secondary stream represents the volatiles. Generally, the char stream composition is defined as 100% C(s). The volatile stream composition is defined by selecting appropriate species and setting their mole or mass fractions. Alternatively, you can use the empirical method (input of atom fractions) for defining these compositions.

  Using two mixture fractions to model coal combustion is more accurate than using one mixture fraction as the volatile and char streams are modeled separately. However, the two-mixture-fraction model incurs significant additional computational expense since the multi-dimensional PDF integrations are performed at run-time.

• When coal is the only fuel in the system, you can choose to model the coal using a single mixture fraction (the fuel stream). When this approach is adopted, the fuel composition you define includes both volatile species and char. Char is typically represented by including C(s) in the species list. You can define the fuel stream composition by selecting appropriate species and setting their mole fractions, or by using the empirical method (input of atom fractions). Definition of the composition is described in detail below.

  Using a single mixture fraction for coal combustion is less accurate than using two mixture fractions. However, convergence in ANSYS FLUENT should be substantially faster than the two-mixture-fraction model.

• When coal is used with another (gaseous or liquid) fuel of different composition, you must model the coal with one mixture fraction and use a second mixture fraction to represent the second (gaseous or liquid) fuel. The stream associated with the coal composition is defined as detailed below for single-mixture-fraction models.

Defining the Coal Composition: Single-Mixture-Fraction Models

When coal is modeled using a single mixture fraction (the fuel stream), the fuel stream composition can be input using the conventional approach or the empirical fuel approach.

• Conventional approach:

  To use the conventional approach, you will need to define the mixture of species in the coal and their mole or mass fractions in the fuel stream. Use the Boundary tab in the Species Model dialog box to input the list of species (e.g., C $_3$H $_8$, CH $_4$, CO, CO $_2$, C(s)) that approximate the coal composition, and their mole or mass fractions.

  Note that C(s) is used to represent the char content of the coal. For example, consider a coal that has a molar composition of 40% volatiles and 60% char on a dry ash free (DAF) basis. Assuming the volatiles can be represented by an equimolar mixture of C $_3$H $_8$ and CO, the fuel stream composition defined in the Boundary tab would be C $_3$H $_8$=0.2, CO = 0.2, and C(s)=0.60. Note that the coal composition should always be defined on an ash-free basis, even if ash will be considered in the ANSYS FLUENT calculation.

  To define ash properties, go to the Create/Edit Materials dialog box and select combusting-particle as the Material Type.

  The following table illustrates the conversion from a typical mass-based proximate analysis to the species fraction inputs required by ANSYS FLUENT. Note that the conversion requires that you make an assumption regarding the species representing the volatiles. Here, the volatiles are assumed to exist as an equimolar mix of propane and carbon monoxide.

  
  

  Proximate Analysis	Weight %	Mass Fraction (DAF)	Moles (DAF)	Mole Fraction (DAF)
  Volatiles - C $_3$H $_8$ - CO Fixed Carbon (C(s)) Ash	30 60 10	0.2035 0.1295 0.667 -	0.004625 0.004625 0.05558 -	0.07134 0.07134 0.85732 -
  (Total)			0.06483	1.0

  
  
  
  

  Moisture in the coal can be considered by adding it in the fuel composition as liquid water, H $_2$O(l). The moisture can also be defined as water vapor, H $_2$O, provided that the corresponding latent heat is included in the discrete phase material inputs in ANSYS FLUENT. If the liquid water is used as a boundary species, it should be removed from the list of excluded species (see Section  16.5.1).

  Note that if water is included in the coal, the water release is not modeled as evaporation, which is typically the case in the wet combustion model, described in Section  23.3.2.

• Empirical fuel approach:

  To use the empirical approach, enable the Empirical Fuel Stream option in the Chemistry tab. This method is ideal if you have an elemental analysis of the coal.

  In the Chemistry tab, input the lower heating value and mean specific heat of the coal. ANSYS FLUENT will use these inputs to determine the mole fractions of the chemical species you have included in the system. Then, in the Boundary tab, define the atom fractions of C, H, N, S, and O in the fuel stream.

Note that for both of these composition input methods, you should take care to distinguish atomic carbon, C, from solid carbon, C(s). Atomic carbon should only be selected if you are using the empirical fuel input method.

See Section  16.4.5 for details about further inputs for modeling coal combustion.

Defining the Coal Composition: Two-Mixture-Fraction Models

You can model coal using the two mixture fractions model, where the primary stream represents the char and the secondary stream represents the volatiles.

As in single-mixture-fraction cases, the fuel stream and secondary stream compositions in a two-mixture-fraction case can be input using either the conventional approach or the empirical fuel approach.

• Conventional approach:

  To use the conventional approach, you will need to define the mixture of species in the coal and their mole or mass fractions in the fuel and secondary streams.

  Use the Boundary tab of Species Model dialog box to define the mole or mass fractions of volatile species in the secondary stream (e.g., C $_3$H $_8$, CH $_4$, CO, CO $_2$, C(s)). Next, define the mole or mass fractions of species used to represent the char. Generally, you will input 100% C(s) for the fuel stream.

• Empirical fuel approach:

  To use the empirical fuel approach, enable the Empirical Secondary Stream option in the Chemistry tab for the volatile (secondary) stream. This method is ideal if you have an elemental analysis of the coal.

  In the Chemistry tab, input the lower heating value and mean specific heat of the coal. Then, in the Boundary tab, define the mole or mass fractions of species used to represent the char. Generally, you will input 100% C(s) for the fuel stream. Finally, define the atom fractions of C, H, N, S, and O in the volatiles. ANSYS FLUENT will use these inputs to determine the mole fractions of the chemical species you have included in the system. For example, consider coal with the following DAF (dry ash free) data and elemental analysis:

  
  

  Proximate Analysis	Wt % (dry)	Wt % (DAF)
  Volatiles	28	30.4
  Char (C(s))	64	69.6
  Ash	8	-

  
  
  
  

  Element	Wt % (DAF)	Wt % (DAF)
  C	89.3	89.3
  H	5.0	5.0
  O	3.4	3.4
  N	1.5	2.3
  S	0.8	-

  
  

  (Note that in the final column, for modeling simplicity, the sulfur content of the coal has been combined into the nitrogen mass fraction.)

  You can combine the proximate and ultimate analysis data to yield the following elemental composition of the volatile stream:

  
  

  Element	Mass	Mass Fraction	Moles	Mole Fraction
  C	(89.3 - 69.6)	0.65	5.4	0.24
  H	5.0	0.16	16	0.70
  O	3.4	0.11	0.7	0.03
  N	2.3	0.08	0.6	0.03
  Total	30.4		22.7

  
  

  This adjusted composition is used to define the secondary stream (volatile) composition.

  The lower heating value of the volatiles can be computed from the known heating value of the coal and the char (DAF):

  • LCV $_{{\rm coal, DAF}}$ = 35.3 MJ/kg

  • LCV $_{{\rm char, DAF}}$ = 32.9 MJ/kg

  You can compute the heating value of the volatiles as

  
  

  
  
  

  or

  
  

  
  
  

Note that for both of these composition input methods, you should take care to distinguish atomic carbon, C, from solid carbon, C(s). Atomic carbon should only be selected if you are using the empirical fuel input method.

Additional Coal Modeling Inputs in ANSYS FLUENT

Within ANSYS FLUENT, the DPM coal combustion simulation is defined as usual when the non-premixed combustion model is selected. The air (oxidizer) inlets are defined as having a mixture fraction value of zero. No gas phase fuel inlets will be included and the sole source of fuel will come from the coal devolatilization and char burnout. The coal particles are defined as injections using the Set Injection Properties dialog box in the usual way, and physical properties for the coal material are specified as described in Section  23.5. You should keep in mind the following issues when defining injections and discrete-phase material properties for coal materials:

• In the Set Injection Properties dialog box, you will specify for the Oxidizing Species one of the components of the oxidizer stream. This species concentration field will be used to calculate the diffusion-controlled driving force in the char burnout law (if applicable), and is $O_2$ by default.

  The specification of the char and volatile streams differs depending on the type of model you are defining:

  • If the coal is modeled using a single mixture fraction, the gas phase species representing the volatiles and the char combustion are represented by the mixture fraction used by the non-premixed combustion model.

  • If the coal is modeled using two mixture fractions, rather than specifying a destination species for the volatiles and char, you will instead specify the Devolatilizing Stream (as secondary) and the Char Stream (as primary).

  • If the coal is modeled using one mixture fraction, and another fuel is modeled using a second mixture fraction, you should specify the stream representing the coal as both the Devolatilizing Stream and the Char Stream.

• In the Create/Edit Materials dialog box, Vaporization Temperature should be set equal to the fuel inlet temperature. This temperature controls the onset of the devolatilization process. The fuel inlet temperature that you define in the Boundary tab of the Species Model dialog box should be set to the temperature at which you want to initiate devolatilization. This way, the look-up tables will include the appropriate temperature range for your process.

• In the Create/Edit Materials dialog box, Volatile Component Fraction and Combustible Fraction should be set to values that are consistent with the coal composition used to define the fuel (and secondary) stream composition.

• Also in the Create/Edit Materials dialog box, you will be prompted for the Burnout Stoichiometric Ratio and for the Latent Heat. The Burnout Stoichiometric Ratio is used in the calculation of the diffusion controlled burnout rate but has no other impact on the system chemistry when the non-premixed combustion model is used. The Burnout Stoichiometric Ratio is the mass of oxidant required per mass of char. The default value of 1.33 assumes that C(s) is oxidized by O $_2$ to yield CO. The Latent Heat input determines the heat required to generate the vapor phase volatiles defined in the non-premixed system chemistry. You can usually set this value to zero when the non-premixed model is used, since your definition of volatile species will have been based on the overall heating value of the coal. However, if the coal composition includes the water content, the latent heat should be set as follows:
  • Set latent heat to zero if the water content of the coal has been defined as H2O(L). In this case, the system chemistry will include the latent heat required to vaporize the liquid water.

  • Set latent heat to the value for water (2.25 $\times$ 10 $^{6}$ J/kg), adjusted by the mass loading of water in the volatiles, if the water content of the coal has been defined using water vapor, H2O. In this case, the water content you defined will be evolved along with the other species in the coal but the system chemistry does not include the latent heat effect.

• The Density you define for the coal in the Create/Edit Materials dialog box should be the apparent density, including ash content.

• You will not be asked to define the Heat of Reaction for Burnout for the char combustion.

Postprocessing Non-Premixed Models of Coal Combustion

ANSYS FLUENT reports the rate of volatile release from the coal using the DPM Evaporation/Devolatilization postprocessing variable. The rate of char burnout is reported in the DPM Burnout variable.

The Coal Calculator

The Coal Calculator dialog box automates the calculations described above for setting up a coal case from the proximate and ultimate analyses.

Figure 16.4.2: The Coal Calculator Dialog Box

The inputs to the Coal Calculator dialog box are:

1.   Coal Proximate Analysis, which is the mass fraction of Volatile, Fixed Carbon, Ash and Moisture in the coal.

2.   Coal Ultimate Anaysis, which is the mass fraction of atomic C, H, O, N and optionally S, in the Dry-Ash-Free (DAF) coal.

3.   The option to use a Secondary Stream. If enabled, the two mixture fraction model will be set with the primary stream representing char as $C<s>$, and an empirical secondary stream representing the volatiles.

4.   The Coal Particle Material Name. A DPM Combusting Particle Material will be created with this name. The default name is coal-particle.

5.   The Coal As-Received HCV (Higher Calorific Value).

6.   The High Temperature Volatile Yield. Enhanced devolatization at higher temperatures can cause the volatile yield to exceed the proximate analysis fraction. To model this, the actual volatile fraction used is calculated as that specified in the Proximate Analysis input multiplied by the High Temperature Volatile Yield. The actual Fixed Carbon fraction is then calculated as one minus the sum of the actual Volatile, Ash, and Moisture fractions.

7.    Fraction of N in Char (DAF). This input is used in calculating the split of atomic nitrogen for the Fuel NOx model.

When the OK button is clicked, ANSYS FLUENT makes the following changes:

(a)   The empirical fuel atomic compositions in the Boundary tab are set, and the Non-Adiabatic model is enabled as required for DPM. The empirical fuel (DAF volatile) Lower Calorific Value ( $LCV_{vol}$) is calculated as follows. First the DAF LCV of the coal is computed as,

where $Moisture$ and $Ash$ are the proximate moisture and ash fractions, $H_{ar}$ is the ultimate $H$ fraction, $W_{H_2O}$ and $W_H$ are the molecular weight of water and atomic hydrogen, respectively, and $h^{latent}_{H_2O}$ is the latent heat of water.

$LCV_{vol}$ is calculated from $LCV^{DAF}_{coal}$ using,

where $FixedCarbon$ and $Volatile$ are the proximate fixed carbon and volatile fractions, respectively.

(b)   A combusting particle material is created with Volatile Component Fraction and Combustible Fraction calculated from the ultimate and proximate analyses. The Discrete Phase Model (DPM) is enabled.

(c)   For the Fuel NOx model, the char N conversion is set to NO, and the Fuel NOx Volatile and Char mass fractions are set according to the ultimate and proximate compositions. Note that even though some of the Fuel NOx parameters are changed, the Fuel NOx model itself is not enabled.

After the Coal Calculator has set up the relevant models, you must build the PDF Table by clicking Calculate PDF Table in the Table tab. You will also need to create injections if you have not done this yet. After converging your coal combustion case, you may want to enable the NOx model for post-processing nitrogen-oxide pollutants.