**8. Case study: W-t-E technology development with modern R&D computational tool**

The combustion, gasification or pyrolysis chamber (reactor) needs to be modeled in such way to assure best possible process conditions for the production of complete thermal conversion of waste. For such modeling mostly advanced computer based engineering tools are used. [18][21]

The thermal conversion process by using municipal solid waste as a fuel in W-t-E plant calls for detailed understanding these phenomena. First, this process depends on many input parameters like proximate and ultimate analyses, season of the year, primary and secondary inlet air velocity and second, on the output parameters such as temperature or mass flow rate of conversion products. The variability and mutual dependence of these parameters can be difficult to manage in practice. Another problem is how these parameters can be tuned to achieve the optimal conversion conditions with minimal pollutants emission during the plant design phase. To meet these goals, W-t-E plants are in the design phase investigated by using computational fluid dynamics (CFD) approach. The adequate variable input boundary conditions which are based on the real measurement are used and the whole computational work is updated with real plant geometry and the appropriate turbulence, combustion and heat transfer models. Different operating conditions are varied and conversion products are predicted and visualized.

CFD approach uses for description of conversion process in W-t-E a system of differential equations. Fluid mechanics of reacting flow is modeled with Reynolds Averaged Navier-Stokes equations (RANS), presented in the following form:

$$\frac{\partial \overline{\rho}}{\partial t} + \frac{\partial}{\partial \mathbf{x}\_{\dot{\jmath}}} \left( \overline{\rho} \overline{\nu}\_{\dot{\jmath}} \right) = 0 \tag{3}$$

Turbulent viscosity can be determined using various turbulent models to close-down the system of Reynolds' equations. The two-equation *k -* ε turbulent model is used for the purpose of the presented reacting flow modeling. Application of *k -* ε turbulent model in the modeling of reacting flows has already been proven by many authors as a very successful one. Turbulent

2

<sup>=</sup> (7)

Combustion of Municipal Solid Waste for Power Production

http://dx.doi.org/10.5772/55497

299

) and *ε* its dissipation (irreversible transfor‐

(10)

(11)

e

¯

*iυ* ′ *i*

*j k j j kj <sup>k</sup> k k <sup>I</sup>*

 h

*jj j*

*j i i*

 u u

 u u

*j ij*

*x xx*

1 2 *j i i*

é ù æ ö ¶ ¶ ¶ = +- ê ú ç ÷ ç ÷ ¶ ¶¶ ë û è ø

*I C C*

u

' '

used in the presented work are: *Cη* = 0,09; *C1* = 1,44; *C2* = 1,92; *σk* = 1 and *σε* = 1,3.

ru

*j ij*

Pr *t j p*

*<sup>η</sup> <sup>T</sup> h c*

*t j*

*x*

is the turbulent Prandtl number. *Cη*, *C1*, *C2*, *σk* and *σε* are constants, and their values

*k x xx k*

æ ö ¶ ¶ ¶ =+ - ç ÷ ç ÷ ¶ ¶¶ è ø

 h h

s

e

re

2

¶ = - ¶ (12)

 e

 r

in Eq. 5 is also defined with turbulent viscosity:

h

s

+ -+ = ê ú ç ÷ ¶¶ ¶ ¶ ê ú ë û è ø (8)

e

+ -+ = ê ú ç ÷ ¶¶ ¶ ¶ ê ú ë û è ø (9)

*I*

e

*<sup>k</sup> <sup>C</sup>*h

*t*

Local values of *k* and *ε* are computed using the following transport equations:

( ) ( ) *<sup>t</sup>*

( ) ( ) *<sup>t</sup> j*

u

 u e

*k t*

e

h

*t*

h

*I*

e

*jh* ¯′

*tx x x*

¶¶ ¶ ¶ é ù æ ö

ru

re

the source terms are modeled as:

Reynolds' enthalpy flux *ρυ* ′

where Pr*<sup>t</sup>*

*tx x x*

¶¶ ¶ ¶ é ù æ ö

h r

viscosity is computed using:

where *k* is turbulent kinetic energy – *k* =0.5(*υ* ′

mation of kinetic energy into internal energy).

$$\frac{\partial}{\partial t} \left( \overline{\rho} \overline{\nu}\_{\dot{\jmath}} \right) + \frac{\partial}{\partial \mathbf{x}\_{\dot{\jmath}}} \left( \overline{\rho} \overline{\nu}\_{\dot{\jmath}} \overline{\nu}\_{\dot{\imath}} \right) = -\frac{\partial p}{\partial \mathbf{x}\_{\dot{\imath}}} + \overline{f\_{\imath \dot{\imath}}} - \frac{\partial}{\partial \mathbf{x}\_{\dot{\jmath}}} \left( \overline{\tau}\_{\dot{\imath}\dot{\jmath}} + \overline{\rho \nu\_{\dot{\jmath}}' \nu\_{\dot{\imath}}'} \right) \tag{4}$$

$$\frac{\partial}{\partial t} \left( \overline{\rho} \overline{h} \right) + \frac{\partial}{\partial \mathbf{x}\_{\dot{j}}} \left( \overline{\rho} \overline{\nu} \overline{\nu} \overline{h} \right) - \frac{\partial p}{\partial t} + \frac{\partial}{\partial \mathbf{x}\_{\dot{j}}} \left( \overline{q}\_{\dot{j}} + \overline{\rho \nu \acute{\prime} h^{\prime}} \right) = \overline{I\_{T}} \tag{5}$$

Reynolds' stresses (*ρυ*' *<sup>j</sup> <sup>υ</sup>*' ¯ *<sup>i</sup>* ) are modelled by the introduction of turbulent viscosity *η<sup>t</sup>* :

$$
\overline{\rho \nu\_{\rangle}^{\prime} \nu\_{\ \dot{i}}^{\prime}} = \frac{2}{3} \delta\_{\dot{i}\!\!/} \left( \rho k + \eta\_{t} \frac{\partial \nu\_{k}}{\partial \mathbf{x}\_{k}} \right) - \eta\_{t} \left( \frac{\partial \nu\_{\ \dot{i}}}{\partial \mathbf{x}\_{\ \dot{j}}} + \frac{\partial \nu\_{\ \dot{j}}}{\partial \mathbf{x}\_{\ \dot{i}}} \right) \tag{6}
$$

Turbulent viscosity can be determined using various turbulent models to close-down the system of Reynolds' equations. The two-equation *k -* ε turbulent model is used for the purpose of the presented reacting flow modeling. Application of *k -* ε turbulent model in the modeling of reacting flows has already been proven by many authors as a very successful one. Turbulent viscosity is computed using:

$$
\eta\_t = \rho \mathbf{C}\_\eta \frac{k^2}{\varepsilon} \tag{7}
$$

where *k* is turbulent kinetic energy – *k* =0.5(*υ* ′ *iυ* ′ *i* ¯ ) and *ε* its dissipation (irreversible transfor‐ mation of kinetic energy into internal energy).

Local values of *k* and *ε* are computed using the following transport equations:

$$\frac{\partial}{\partial t}(\rho k) + \frac{\partial}{\partial \mathbf{x}\_j}(\mathbf{\bar{\sigma}}\_j k) - \frac{\partial}{\partial \mathbf{x}\_j} \left[ \left( \eta + \frac{\eta\_t}{\sigma\_k} \right) \frac{\partial k}{\partial \mathbf{x}\_j} \right] = I\_k \tag{8}$$

$$\frac{\partial}{\partial t}(\rho \varepsilon) + \frac{\partial}{\partial \mathbf{x}\_{\dot{\boldsymbol{\beta}}}} (\overline{\mathbf{v}}\_{\dot{\boldsymbol{\beta}}} \varepsilon) - \frac{\partial}{\partial \mathbf{x}\_{\dot{\boldsymbol{\beta}}}} \left[ \left( \eta + \frac{\eta\_{t}}{\sigma\_{\varepsilon}} \right) \frac{\partial \varepsilon}{\partial \mathbf{x}\_{\dot{\boldsymbol{\beta}}}} \right] = I\_{\varepsilon} \tag{9}$$

the source terms are modeled as:

**8. Case study: W-t-E technology development with modern R&D**

The combustion, gasification or pyrolysis chamber (reactor) needs to be modeled in such way to assure best possible process conditions for the production of complete thermal conversion of waste. For such modeling mostly advanced computer based engineering tools

The thermal conversion process by using municipal solid waste as a fuel in W-t-E plant calls for detailed understanding these phenomena. First, this process depends on many input parameters like proximate and ultimate analyses, season of the year, primary and secondary inlet air velocity and second, on the output parameters such as temperature or mass flow rate of conversion products. The variability and mutual dependence of these parameters can be difficult to manage in practice. Another problem is how these parameters can be tuned to achieve the optimal conversion conditions with minimal pollutants emission during the plant design phase. To meet these goals, W-t-E plants are in the design phase investigated by using computational fluid dynamics (CFD) approach. The adequate variable input boundary conditions which are based on the real measurement are used and the whole computational work is updated with real plant geometry and the appropriate turbulence, combustion and heat transfer models. Different operating conditions are varied and conversion products are

CFD approach uses for description of conversion process in W-t-E a system of differential equations. Fluid mechanics of reacting flow is modeled with Reynolds Averaged Navier-

( ) 0 *<sup>j</sup>*

u

t ru u

ru

*<sup>i</sup>* ) are modelled by the introduction of turbulent viscosity *η<sup>t</sup>*

*j k i*

 u

*k ji*

*x xx*

 h

¶ ¶¶ ç ÷ è ø è ø

+ =- + - + ¢ ¢ ¶ ¶¶ (4)

+ -+ + = ¢ ¢ ¶ ¶¶ (5)

u

¶ (3)

:

(6)

ru

( *<sup>j</sup>*) ( *j i*) *i ij j i* ( ) *j ij*

( ) ( *<sup>j</sup>* ) ( *jj T* ) *j j <sup>p</sup> h h q hI t x tx*

u

æ ö ¶ ¶æ ö ¶ ¢ = + -+ ç ÷ ç ÷

¶ ¶ ¶

*<sup>j</sup> t x*

*<sup>p</sup> <sup>f</sup> tx xx*

¶ ¶ ¶

*j i ij t t*

*k*

 ru u

> ru

2 ' 3

 d r h

¶ + =

¶r

¶

Stokes equations (RANS), presented in the following form:

**computational tool**

298 Advances in Internal Combustion Engines and Fuel Technologies

are used. [18][21]

predicted and visualized.

¶

¶

ru

¶

¶

Reynolds' stresses (*ρυ*' *<sup>j</sup> <sup>υ</sup>*' ¯

r

ru u

$$I\_k = \eta\_t \left( \frac{\partial \overline{\boldsymbol{\sigma}}\_i}{\partial \boldsymbol{\alpha}\_j} + \frac{\partial \overline{\boldsymbol{\sigma}}\_j}{\partial \boldsymbol{\alpha}\_i} \right) \frac{\partial \overline{\boldsymbol{\sigma}}\_i}{\partial \boldsymbol{\alpha}\_j} - \rho \varepsilon \tag{10}$$

$$I\_{\varepsilon} = \mathbf{C}\_{1} \frac{\varepsilon}{k} \bigg[ \eta\_{t} \bigg( \frac{\partial \overline{\upsilon}\_{i}}{\partial \mathbf{x}\_{j}} + \frac{\partial \overline{\upsilon}\_{j}}{\partial \mathbf{x}\_{i}} \bigg) \frac{\partial \overline{\upsilon}\_{i}}{\partial \mathbf{x}\_{j}} \bigg] - \mathbf{C}\_{2} \rho \frac{\varepsilon^{2}}{k} \tag{11}$$
 
$$\begin{aligned} \overline{\rho \nu \overline{\upsilon}^{\prime} \mu^{\prime}} \text{ in Eq. 5 is also defined with turbulent viscosity:} \end{aligned}$$

Reynolds' enthalpy flux *ρυ* ′ in Eq. 5 is also defined with turbulent viscosity:

$$\overline{\rho \boldsymbol{\nu}^\* \boldsymbol{h}^\*} = -\frac{\eta\_t}{\mathbf{Pr}\_t} \boldsymbol{c}\_p \frac{\boldsymbol{\mathcal{O}} \boldsymbol{T}}{\boldsymbol{\mathcal{O}} \boldsymbol{x}\_j} \tag{12}$$

where Pr*<sup>t</sup>* is the turbulent Prandtl number. *Cη*, *C1*, *C2*, *σk* and *σε* are constants, and their values used in the presented work are: *Cη* = 0,09; *C1* = 1,44; *C2* = 1,92; *σk* = 1 and *σε* = 1,3.

Advection – diffusive equation of mass species (*ξk*) of the component *k* has due to Reynolds' averaging, an additional term called turbulent mass species flux:

$$\overline{\rho \xi\_k' \nu\_j'} = \frac{\eta\_t}{\mathbf{S} \mathbf{c}\_t} \frac{\partial \xi\_k}{\partial \mathbf{x}\_j} \tag{13}$$

which is written in following form of general chemical reaction:

n

**Figure 11.** The 2D engineering plan and photos within built combustion chamber

secondary combustion chamber can be also seen.

applied.

phase of the new one.

after plant was built.

1 1

*k k*

*N N*

*kb*

*k k k k*

Where *ν 'k* and *ν ''<sup>k</sup>* designate the stoichiometric coefficients of component *k* for reactants and products, respectively. Chemical reaction rate *Rk* in Eq. 19 is calculated by appropriate combustion models. It has to be pointed out that nowadays many turbulent combustion models are in practical use. Their application depends on the type of combustion (diffusion, kinetic, mixed), fuel type (solid, liquid, gaseous) and combustion device (furnace, boiler, engine). Most of models include various empirical constants which need to be individually determined case by case. In this case, on the base of best practice recommendations and its references [1] for this kind of combustion the Eddy Dissipation Combustion Model should be

With CFD approach the combustion processes can be predicted and the operating conditions with combustion chamber design can be optimized in existing W-t-E or in the design project

Figure 11 shows 2D engineering plan view of W-t-E plant. On this base the W-t-E was built and operates with RDF. The photos of primary combustion chamber on Figure 11 were taken

Figure 11 shows grate details in the primary combustion chamber with waste input and the secondary combustion chamber with secondary and tertiary air inlet. Moreover, the exit of the

The 3D geometry plan on the base of engineering plans in real measure was drown (Figure 12). Each dimension was marked on the plan with corresponding input dimensions which can be varied. In this way each dimension is easy and quickly modified and the entire construction can be modified and redrawn and further steps like mesh creation or design optimization is

 *X X* ¾¾® ¬¾¾ = =

, *<sup>k</sup> <sup>f</sup>*

 n

å å ¢ ¢¢ (20)

Combustion of Municipal Solid Waste for Power Production

http://dx.doi.org/10.5772/55497

301

and can be modeled with turbulent viscosity using the *k-ε* model. The complete advection – diffusive mass species equation is:

$$\frac{\partial}{\partial t} \left( \overline{\rho \boldsymbol{\varepsilon}\_k} \right) + \frac{\partial}{\partial \boldsymbol{\alpha}\_j} \left( \overline{\rho \boldsymbol{\nu}\_j \boldsymbol{\xi}\_k} \right) - \frac{\partial}{\partial \boldsymbol{\alpha}\_j} \left| \left( \rho \boldsymbol{D}\_k + \frac{\eta\_t}{\mathrm{Sc}\_t} \right) \frac{\partial \boldsymbol{\xi}\_k}{\partial \boldsymbol{\alpha}\_j} \right| = \overline{\boldsymbol{I}\_{\boldsymbol{\xi}\_k}} \tag{14}$$

where Sc*<sup>t</sup>* is the turbulent Schmidt number and *Dk* molecular diffusion coefficient of component *k*. With the new term:

$$
\Gamma\_{k,eff} = \rho D\_k + \frac{\eta\_t}{\mathbf{Sc}\_t} = \Gamma\_k + \frac{\eta\_t}{\mathbf{Sc}\_t} \tag{15}
$$

the Eq. 14 can be rewritten as:

$$\frac{\partial}{\partial t} \left( \overline{\rho \boldsymbol{\varepsilon}\_k^{\varepsilon}} \right) + \frac{\partial}{\partial \boldsymbol{\varepsilon}\_j} \left( \overline{\rho \boldsymbol{\nu}\_j \boldsymbol{\varepsilon}\_k^{\varepsilon}} \right) - \frac{\partial}{\partial \boldsymbol{\varepsilon}\_j} \left( \boldsymbol{\Gamma}\_{k, \text{eff}} \frac{\partial \boldsymbol{\varepsilon}\_k^{\varepsilon}}{\partial \boldsymbol{\varepsilon}\_j} \right) = \overline{\boldsymbol{I}\_{\boldsymbol{\xi}\_k}} \tag{16}$$

Source terms of energy and mass species transport equations are computed by the following two equations where *ωk* is computed by the turbulent combustion model:

$$\overline{I\_T} = -\sum\_{k=1}^{N} \Delta H\_{f,k}^o \overline{o\_k} \tag{17}$$

$$
\overline{I\_{\xi\_k}} = \mathcal{M}\_k \overline{o\_k} \tag{18}
$$

whereΔ *H°f,k* is the standard heat formation and *Mk* the molecular mass of the component *k*. In Eq. 17 and Eq. 18 the *ωk* stands for the formation/consumption rate of component *k* and is defined by the following expression:

$$
\overline{\rho}\_{k} = \frac{d\left\lbrack \overline{X}\_{k} \right\rbrack}{dt} = \left(\nu\_{k}'' - \nu\_{k}'\right)\overline{R}\_{k} \tag{19}
$$

which is written in following form of general chemical reaction:

Advection – diffusive equation of mass species (*ξk*) of the component *k* has due to Reynolds'

Sc *t k*

( ) ( ) Sc *<sup>k</sup>*

, Sc Sc

( ) ( ) , *<sup>k</sup>*

*j jj*

Source terms of energy and mass species transport equations are computed by the following

,

w

whereΔ *H°f,k* is the standard heat formation and *Mk* the molecular mass of the component *k*. In Eq. 17 and Eq. 18 the *ωk* stands for the formation/consumption rate of component *k* and is

> ( ) *<sup>k</sup> k k kk*

 nn *R*

*d X*

w

*dt*

é ù

w

1 *<sup>N</sup> <sup>o</sup> T fk k k I H*

> *<sup>k</sup> k k I M* x=

=

+ -G = ç ÷

*k j k k eff*

¶¶ ¶¶ ç ÷

æ ö ¶¶ ¶ ¶

*t x xx*

 ru x

two equations where *ωk* is computed by the turbulent combustion model:

h

*k eff k k*

 G = + =G + r

*D*

*j j tj*

 r

h x

*t j x*

and can be modeled with turbulent viscosity using the *k-ε* model. The complete advection –

¶ ¢ ¢ <sup>=</sup> ¶ (13)

x

(15)

ë û è ø (14)

*t k*

 x

*D I*

h

is the turbulent Schmidt number and *Dk* molecular diffusion coefficient of component

*t t*

 h

*k*

x

è ø

*I*

x

=- D å (17)

ë û = = -¢¢ ¢ (19)

(18)

(16)

*t t*

*k j*

rx u

*k j k k*

 ru x

*tx x x*

 ¶¶ ¶ é ù æ ö ¶ + - += ê ú ç ÷ ¶¶ ¶ ¶ ê ú

averaging, an additional term called turbulent mass species flux:

diffusive mass species equation is:

where Sc*<sup>t</sup>*

*k*. With the new term:

the Eq. 14 can be rewritten as:

defined by the following expression:

rx

300 Advances in Internal Combustion Engines and Fuel Technologies

rx

$$\sum\_{k=1}^{N} \nu\_k' X\_k \quad \xleftarrow{k\_f} \quad \sum\_{k=1}^{N} \nu\_k'' X\_{k'} \tag{20}$$

Where *ν 'k* and *ν ''<sup>k</sup>* designate the stoichiometric coefficients of component *k* for reactants and products, respectively. Chemical reaction rate *Rk* in Eq. 19 is calculated by appropriate combustion models. It has to be pointed out that nowadays many turbulent combustion models are in practical use. Their application depends on the type of combustion (diffusion, kinetic, mixed), fuel type (solid, liquid, gaseous) and combustion device (furnace, boiler, engine). Most of models include various empirical constants which need to be individually determined case by case. In this case, on the base of best practice recommendations and its references [1] for this kind of combustion the Eddy Dissipation Combustion Model should be applied.

With CFD approach the combustion processes can be predicted and the operating conditions with combustion chamber design can be optimized in existing W-t-E or in the design project phase of the new one.

**Figure 11.** The 2D engineering plan and photos within built combustion chamber

Figure 11 shows 2D engineering plan view of W-t-E plant. On this base the W-t-E was built and operates with RDF. The photos of primary combustion chamber on Figure 11 were taken after plant was built.

Figure 11 shows grate details in the primary combustion chamber with waste input and the secondary combustion chamber with secondary and tertiary air inlet. Moreover, the exit of the secondary combustion chamber can be also seen.

The 3D geometry plan on the base of engineering plans in real measure was drown (Figure 12). Each dimension was marked on the plan with corresponding input dimensions which can be varied. In this way each dimension is easy and quickly modified and the entire construction can be modified and redrawn and further steps like mesh creation or design optimization is possible in real time. On this base, the mesh of 160,871 nodes and 810,978 elements (Figure 12) was created. It is very important that the mesh creation is designed optimally which means that the mesh is more dense in significant area like air input or when the combustion processes are very intensive such as in the primary and secondary combustion chamber. Due to these facts the optimal control volume size is needed and the remeshing iteration process is estab‐ lished to achieve the optimum mesh creation. That means that smaller control volume is applied where the combustion process is more intensive or at the reactants inlet of the W-t-E what is clearly seen in Figure 12.

**Figure 13.** Area definition in W-t-E

secondary air inlet

**Figure 14.** Simulation results for maximal ash temperature versus secondary air velocity and oxygen mass fraction in

Combustion of Municipal Solid Waste for Power Production

http://dx.doi.org/10.5772/55497

303

**Figure 12.** geometry plan of W-t-E with dimensions and geometry meshing

In addition the boundary conditions with entire combustion, radiation, particle tracking and other models with input and output parameters are set up and the solver is started to reach the convergence criteria like maximum number of iterations or residual target. These input parameters are operating conditions like intake velocities, temperatures, reactants mass flow rates, dimension values and the output parameters like temperatures, combustion products mass flow rate and other flue gas parameters. The boundary condition components of gaseous component are changeable and dependent on the distance of coordinate x. In this work the boundary conditions are set as a polynomial function of variable x:

$$\mathbf{f}\_k(\mathbf{x}) = \mathbf{a}\_k \mathbf{x}^3 + \mathbf{b}\_k \mathbf{x}^2 + \mathbf{c}\_k \mathbf{x} + \mathbf{d}\_k; \quad \mathbf{k} = 1...n; \text{ a, b, c, d} = \text{constants} \tag{21}$$

corresponding to the statistics of the local measurements of specific gaseous components along the grate [11][14][20][21][19].

Figure 13 shows the marked area for primary, secondary and tertiary air inlet, fuel inlet and flue gases outlet. In addition, special cross section on secondary combustion chamber on inlet (SecIn) and outlet (SecOut) were created to identify and to monitor the combustion products and other parameters in this significant area. In this way, the location of single parameter can be distinguished. The W-t-E operation optimization process was made by using design exploration which is a powerful tool for designing and understanding the analysis response of parts and assemblies.

**Figure 13.** Area definition in W-t-E

possible in real time. On this base, the mesh of 160,871 nodes and 810,978 elements (Figure 12) was created. It is very important that the mesh creation is designed optimally which means that the mesh is more dense in significant area like air input or when the combustion processes are very intensive such as in the primary and secondary combustion chamber. Due to these facts the optimal control volume size is needed and the remeshing iteration process is estab‐ lished to achieve the optimum mesh creation. That means that smaller control volume is applied where the combustion process is more intensive or at the reactants inlet of the W-t-E

In addition the boundary conditions with entire combustion, radiation, particle tracking and other models with input and output parameters are set up and the solver is started to reach the convergence criteria like maximum number of iterations or residual target. These input parameters are operating conditions like intake velocities, temperatures, reactants mass flow rates, dimension values and the output parameters like temperatures, combustion products mass flow rate and other flue gas parameters. The boundary condition components of gaseous component are changeable and dependent on the distance of coordinate x. In this work the

<sup>f</sup>*<sup>k</sup>* (*x*)= <sup>a</sup>*<sup>k</sup> <sup>x</sup>* <sup>3</sup> <sup>+</sup> <sup>b</sup>*<sup>k</sup> <sup>x</sup>* <sup>2</sup> <sup>+</sup> <sup>c</sup>*<sup>k</sup> <sup>x</sup>* <sup>+</sup> <sup>d</sup>*<sup>k</sup>* ; k= 1...*n*; a, b, c, d=constants (21)

corresponding to the statistics of the local measurements of specific gaseous components along

Figure 13 shows the marked area for primary, secondary and tertiary air inlet, fuel inlet and flue gases outlet. In addition, special cross section on secondary combustion chamber on inlet (SecIn) and outlet (SecOut) were created to identify and to monitor the combustion products and other parameters in this significant area. In this way, the location of single parameter can be distinguished. The W-t-E operation optimization process was made by using design exploration which is a powerful tool for designing and understanding the analysis response

what is clearly seen in Figure 12.

302 Advances in Internal Combustion Engines and Fuel Technologies

the grate [11][14][20][21][19].

of parts and assemblies.

**Figure 12.** geometry plan of W-t-E with dimensions and geometry meshing

boundary conditions are set as a polynomial function of variable x:

**Figure 14.** Simulation results for maximal ash temperature versus secondary air velocity and oxygen mass fraction in secondary air inlet

Figure 14 shows analyses that help us to determinate the interaction among maximal ash temperature versus secondary air velocity and oxygen mass flow rate at secondary air inlet. The maximal ash temperature from secondary air velocity from 27 m/s to 29 m/s increases rapidly and picked the maximum ash temperature at 1,850 K. On the other hand, there is no significant dependence of oxygen mass flow rate in region from 0.255 to 0.21. In this way we can predict and avoid the possible damages cause by fly ash flagging on boiler tubes.

Figure 15 show results of temperature field comparison by different operating condition with different oxygen mass flow rates in case of enriched oxygen combustion. The temperature in secondary combustion chamber increases when oxygen enriched air is used [4] and this phenomena is clearly seen by temperature comparison on this picture. On the other hand, we have to be sure that the maximum ash temperature was not exceeded the ash melting point and we have to avoid fly ash deposit on heat exchangers walls which can cause a great damage.

Figure 16 shows 3D ash temperature particle tracking through the W-t-E. The ash temperature changing through the W-t-E and it was picked in the secondary combustion chamber where the oxygen enhanced combustion is used. In addition, the ash temperature has fallen due to the wall cooling. It was found out when the flaying ash clashes into the walls the probability of ash deposit at these sections is high.

**Figure 16.** Ash temperature particle tracking and streamlines with velocity review

with appropriate inlet boundary conditions.

which is distributed to the citizens or industry.

**9. Conclusion**

be applied.

This must be taken into consideration when the residence time is calculated.

Streamlines with velocity review is shown in Figure 16. The majority of the stream takes the short way through W-t-E and the velocity becomes higher at the exit of the secondary chamber.

Combustion of Municipal Solid Waste for Power Production

http://dx.doi.org/10.5772/55497

305

As shortly presented in this chapter the CFD with additional optimization features is the most convenient tool to predict the optimal conditions which have to be achieved to achieve the thermal and environmental efficiency and never to endanger the safety of the W-t-E operation. With this tool the problems because can be avoided and the whole situation can be predicted

Waste presents a source of energy. The energy utilization is possible with the appropriate integrated waste management system and utilization of appropriate technologies within the legally permissible environmental impact. Such system can create power and heat or cold,

Future waste management is going to depend on W-t-E technologies for the high calorific part of the waste stream, not suitable for recycling. The energy in waste will be utilized as the energy prices are not only high but are in constant rise. But the decision making process for the technology selection should not stand only on presented energy efficiency of the technology, thus only full scale long term tested technologies with proven environmental impact should

**Figure 15.** Temperature field by different oxygen mass flow rate at secondary enriched air inlet

**Figure 16.** Ash temperature particle tracking and streamlines with velocity review

Streamlines with velocity review is shown in Figure 16. The majority of the stream takes the short way through W-t-E and the velocity becomes higher at the exit of the secondary chamber. This must be taken into consideration when the residence time is calculated.

As shortly presented in this chapter the CFD with additional optimization features is the most convenient tool to predict the optimal conditions which have to be achieved to achieve the thermal and environmental efficiency and never to endanger the safety of the W-t-E operation. With this tool the problems because can be avoided and the whole situation can be predicted with appropriate inlet boundary conditions.
