Numerical Simulation of Rice Husk as an Alternative Fuel in a Precalciner

Abstract

To tackle the issues of high energy consumption, substantial carbon emission intensity in the cement industry, as well as under-utilization of agricultural waste, this study took an 8000 t/d cement production line at a plant in Indonesia as the research object. Using a Computational Fluid Dynamics (CFD)-based numerical method, the co-firing of pulverized coal with rice husk was simulated in both In-Line Calciner (ILC) and Separate-Line Calciner (SLC) precalciners. Four rice husk replacement levels (10%, 20%, 30%, and 40%) were evaluated in terms of temperature distribution, species concentration, raw meal calcination, and pollutant formation. The results indicate that increasing the rice husk ratio reduces the high-temperature region, lowers the peak temperature, and decreases overall thermal levels. The decomposition rate of CaCO₃ at the outlet of the ILC-type precalciner decreased from 81.11% to 75.32%, while that of the SLC-type precalciner fell from 93.27% to 88.50%. CO₂ and NO_X emissions were remarkably reduced, with the emission reduction effect positively correlated with the rice husk substitution ratio. Taking into account both decomposition rate requirements and emission reduction targets, it is recommended that the blending ratio of rice husk in ILC precalciners should be controlled at 10%, while for SLC precalciners, it can be increased to 40%. This provides a technical reference for low-carbon transformation and biomass resource utilization in the cement industry.

1. Introduction

Cement manufacturing represents 12–15% of total industrial energy use and roughly 7% of worldwide CO₂ emissions, of which fuel combustion accounts for about 40% [1,2]. Additionally, cement production generates significant amounts of NO_X, SO₂ and particulate matter [3], exacerbating air quality and ecological issues. Maintaining production capacity while reducing energy consumption and pollution emissions is paramount for the cement industry. Among the various strategies, the utilization of alternative fuels is widely regarded as one of the most promising technologies for coal conservation and emission reduction, significantly decreasing the reliance on fossil fuels and curbing greenhouse gas and pollutant emissions [4,5]. Among these alternative fuels, biomass stands out due to its renewability and carbon neutrality [6]. Firstly, biomass is characterized by its low carbon, low nitrogen and low pollution properties [7]. Secondly, it boasts a higher volatile content and a faster combustion rate [8]. Rice husk, a primary by-product of rice processing, is produced in significant quantities in agricultural giants like China. Historically, these rice husks were often incinerated or landfilled, resulting in resource wastage and environmental pollution. If rice husk and pulverized coal can be co-combusted, it can not only decrease coal usage and carbon emissions but also enhance the resource utilization rate of agricultural waste [9].

There has been extensive research and practical application in the cement industry regarding the use of rice husk as an alternative fuel. Yu et al. [10] conducted combustion tests on rice husk, coal, and their mixed fuel using the non-isothermal thermogravimetric method. They found that the ignition temperature of the mixed fuel was notably lower than that of pure coal, with significantly improved combustion characteristics and a higher degree of burnout. Li et al. [11] performed combustion experiments with sawdust, rice husk, corn straw and bituminous coal under simulated precalciner conditions, focusing on the analysis of NO_X generation and reduction characteristics. Ren et al. [12] used a cement project in Southeast Asia as an example to introduce the application of rice husk as a substitute for raw coal in cement clinker calcination. The results indicated that reducing coal usage had a manageable impact on clinker burning and clinker quality, with notable decreases in SO₂ and NO_X emissions from the system. Liang et al. [13] systematically analyzed the application effect of rice husk as a biomass alternative fuel in a 5000 t/d cement production line through industrial tests. They discovered that rice husk blend combustion is compatible with the thermal system of cement kilns. By replacing part of the pulverized coal, rice husk can effectively reduce carbon emissions from fossil fuels. Furthermore, rice husk blend combustion does not significantly negatively impact clinker performance, maintaining a stable mineral composition and ensuring that physical properties meet standards. Yang [14] chose a new dry process cement kiln production line as the project carrier. By adopting the precalciner direct combustion scheme and using rice husk as a biomass alternative fuel, the clinker output remained unchanged, while the standard coal consumption for the clinker decreased by 19.8%. Cheng et al. [15] proposed a low-temperature co-gasification technology for rice husk and waste tires. This technology reduced the silicon and sulfur content in the solid residue, minimized the risk of clinker crusting, and decreased NO_X emissions. El-Salamony et al. [16] tested a mixture of rice husk and refuse-derived fuel (RDF) as a coal substitute in an actual production line. They found that incorporating this mixture could decrease coal consumption by 17% and effectively reduce power consumption by 13%.

The selection of rice husk as the biomass fuel in this study was based on a systematic evaluation of locally available agricultural and forestry residues against criteria relevant to cement precalciner operation: annual availability, lower heating value, volatile content, ash content and composition, alkali/silica ratio, grindability, and compatibility with existing kiln systems. Most existing research on coal and rice husk co-combustion is at the laboratory scale, yet cement precalciners are large-scale reactors with complex coupled reactions (fuel combustion and raw meal decomposition), making lab studies insufficient to guide real production, and co-firing studies in precalciners remain scarce. To address this gap, this work integrates actual production data from thermal calibration at an Indonesian cement plant with CFD numerical simulations and performs a direct comparative CFD analysis of rice husk co-firing in both ILC and SLC precalciner configurations under identical fuel and operating conditions. By systematically evaluating four blending ratios (10–40 wt%), the study correlates the resulting temperature field, species field, raw meal decomposition, and NOx emissions with the blending ratio, and derives precalciner-specific optimal blending recommendations that balance production constraints with emission reduction targets, thereby providing technical guidance for practical operations.

2. Simulation Methods and Mathematical Models

2.1. Fundamental Conservation Equation

The mass transfer process of fluids such as fuel and raw material within precalciners is exceedingly complex, yet fluid flow must adhere to the conservation laws for mass, momentum, and energy [17]. The corresponding governing equations are as follows:

(1)

Mass conservation equation:

$${\displaystyle \frac{\partial \rho }{\partial \mathrm{t}}}+\nabla \times (\rho \stackrel{\mathrm{⃑}}{\mathrm{u}})=0$$

(1)

where $\mathrm{t}$represents time (s); $\rho$ denotes density (kg/m³); and $\stackrel{\mathrm{⃑}}{\mathrm{u}}$is the velocity vector (m/s).

(2)

Momentum conservation equation:

$${\displaystyle \frac{\partial (\rho {\overline{\mathrm{u}}}_{\mathrm{i}})}{\partial \mathrm{t}}}+{\displaystyle \frac{\partial (\rho {\overline{\mathrm{u}}}_{\mathrm{i}}{\overline{\mathrm{u}}}_{\mathrm{j}})}{\partial {\mathrm{x}}_{\mathrm{j}}}}=-{\displaystyle \frac{\partial \overline{\mathrm{p}}}{\partial {\mathrm{x}}_{\mathrm{i}}}}+{\displaystyle \frac{\partial }{\partial {\mathrm{x}}_{\mathrm{j}}}}[\mu ({\displaystyle \frac{\partial {\overline{\mathrm{u}}}_{\mathrm{i}}}{\partial {\mathrm{x}}_{\mathrm{j}}}}+{\displaystyle \frac{\partial {\overline{\mathrm{u}}}_{\mathrm{j}}}{\partial {\mathrm{x}}_{\mathrm{i}}}})]+{\displaystyle \frac{\partial \left(-\rho \overline{{\mathrm{u}}_{\mathrm{i}}^{‘}{\mathrm{u}}_{\mathrm{j}}^{‘}}\right)}{\partial {\mathrm{x}}_{\mathrm{j}}}}+{ \mathrm{S}}_{\mathrm{u}}$$

(2)

where ${\overline{\mathrm{u}}}_{\mathrm{i}}$ is the Reynolds-averaged (time-mean) velocity component (m/s); $\overline{\mathrm{p}}$ is the Reynolds-averaged pressure (Pa); $\mu$is the dynamic viscosity of the fluid (Pa·s); and ${\mathrm{S}}_{\mathrm{u}}$ is the user-defined or additional momentum source term (N/m³).

(3)

Energy conservation equation:

$$\rho {\overline{\mathrm{c}}}_{\mathrm{p}}\left({\displaystyle \frac{\partial \overline{\mathrm{T}}}{\partial \mathrm{t}}}+{\overline{ \mathrm{u}}}_{\mathrm{j}}\times {\displaystyle \frac{\partial \overline{\mathrm{T}}}{\partial {\mathrm{x}}_{\mathrm{j}}}}\right)={\displaystyle \frac{\partial }{\partial {\mathrm{x}}_{\mathrm{j}}}}\left({\mathrm{k}}_{\mathrm{eff}}{\displaystyle \frac{\partial \overline{\mathrm{T}}}{\partial {\mathrm{x}}_{\mathrm{j}}}}\right)+\Phi +{\mathrm{q}}_{\mathrm{rad}}+{\mathrm{q}}_{\mathrm{gen}}$$

(3)

where ${\overline{\mathrm{c}}}_{\mathrm{p}}$ is the time-averaged specific heat at constant pressure (J/(kg·K)); $\overline{\mathrm{T}}$ is the time-averaged temperature (K); ${\mathrm{k}}_{\mathrm{eff}}$ is the effective thermal conductivity (W/(m·K)); $\Phi$ denotes the viscous dissipation term (W/m³); ${\mathrm{q}}_{\mathrm{rad}}$ is the radiation heat source term (W/m³); and ${\mathrm{q}}_{\mathrm{gen}}$ is the volumetric heat source term (W/m³).

2.2. Turbulence and Particle Motion Model

Considering the complexity of fluid flow within the precalciner, including the presence of swirling and secondary flows, this study adopted the Realizable $\mathrm{k}\text{-}\epsilon$ model with swirl correction [18]. Here, $\mathrm{k}$ represents turbulent kinetic energy (m²/s²), while $\epsilon$ denotes the turbulent energy dissipation rate (m²/s²). The corresponding equation is as follows:

$${\displaystyle \frac{\partial \left(\rho \mathrm{k}\right)}{\partial \mathrm{t}}} + {\displaystyle \frac{\partial \left(\rho \mathrm{k}{\mathrm{u}}_{\mathrm{i}}\right)}{\partial {\mathrm{x}}_{\mathrm{i}}}} = {\displaystyle \frac{\partial }{ \partial {\mathrm{x}}_{\mathrm{j}}}}\left[\left(\mu +{\displaystyle \frac{{\mu }_{\mathrm{t}}}{{\sigma }_{\mathrm{k}}}}\right){\displaystyle \frac{\partial \mathrm{k}}{\partial {\mathrm{x}}_{\mathrm{j}}}}\right]+ {\mathrm{G}}_{\mathrm{k}} +{ \mathrm{G}}_{\mathrm{b} }- \rho \epsilon -{ \mathrm{Y}}_{\mathrm{M}} +{ \mathrm{S}}_{\mathrm{k}}$$

(4)

$${\displaystyle \frac{\partial \left(\rho \epsilon \right)}{\partial \mathrm{t}}}+{\displaystyle \frac{\partial \left(\rho \epsilon {\mathrm{u}}_{\mathrm{i}}\right)}{\partial {\mathrm{x}}_{\mathrm{i}}}}={\displaystyle \frac{\partial }{\partial {\mathrm{x}}_{\mathrm{j}}}}\left[\left(\mu +{\displaystyle \frac{{\mu }_{\mathrm{t}}}{{\sigma }_{\epsilon }}}\right){\displaystyle \frac{\partial \epsilon }{\partial {\mathrm{x}}_{\mathrm{j}}}}\right]-{\mathrm{C}}_{2\epsilon }\rho {\displaystyle \frac{{\epsilon }^2}{\mathrm{k}}}+ {\mathrm{C}}_{1\epsilon }{\displaystyle \frac{\epsilon }{\mathrm{k}}}\left({\mathrm{G}}_{\mathrm{k} }+{\mathrm{C}}_{3\epsilon }{\mathrm{G}}_{\mathrm{b}}\right)+{ \mathrm{S}}_{\epsilon }$$

(5)

where ${\mathrm{G}}_{\mathrm{b}}$ represents the turbulent kinetic energy induced by buoyancy (kg/(m·s³)), and ${\mathrm{G}}_{\mathrm{k}}$ denotes the turbulent kinetic energy driven by the average velocity gradient (kg/(m·s³)); ${\mathrm{Y}}_{\mathrm{M}}$ signifies the contribution of wave expansion to the total dissipation rate in compressible turbulence (kg/(m·s³)); ${\sigma }_{\mathrm{k}}$ and ${\sigma }_{\epsilon }$ are the turbulent Prandtl numbers for turbulent kinetic energy and dissipation rate, respectively; ${\mathrm{S}}_{\mathrm{k} }$and ${\mathrm{S}}_{\epsilon }$ are user-defined source terms (kg/(m·s³)); and ${\mathrm{C}}_{1\epsilon }$, ${\mathrm{C}}_{2\epsilon }$ and ${\mathrm{C}}_{3\epsilon }$ are empirical constants.

In this study, the discrete phase model (DPM) is employed to simulate the movement of particles within the precalciner. This model is founded on the stochastic trajectory model from a Lagrangian perspective to simulate the movement process of particles [19]. The motion equation for particles under the stochastic trajectory model is as follows:

$${\displaystyle \frac{\mathrm{d}{\mathrm{u}}_{\mathrm{p}} }{\mathrm{dt}}} ={\mathrm{F}}_{\mathrm{D}}\left(\overline{\mathrm{u}}-{\mathrm{u}}_{\mathrm{p}}\right) + {\displaystyle \frac{{\mathrm{g}}_{\mathrm{x}}\left({\rho }_{\mathrm{p}}-\rho \right)}{{\rho }_{\mathrm{p}}}} + {\mathrm{F}}_{\mathrm{x}}$$

(6)

where ${\mathrm{F}}_{\mathrm{D}}\left(\overline{\mathrm{u}} - {\mathrm{u}}_{\mathrm{p}}\right)$ represents the attractive force per unit mass on the particles (N/kg); $\overline{\mathrm{u}}$ represents the local Favre-averaged gas velocity component in the direction of particle motion, obtained from the Eulerian flow field solved by the turbulence model (m/s); ${\rho }_{\mathrm{p}}$ signifies the density of the particles (kg/m³); ${\mathrm{g}}_{\mathrm{x}}$ is the gravitational acceleration (m/s²); and ${\mathrm{F}}_{\mathrm{x}}$ represents the sum of all forces (N) acting on the particles, excluding gravity and the attractive force.

2.3. Particle Combustion Model and Radiation Heat Transfer Model

In this study, the kinetics/diffusion-limited model is used to simulate the char combustion process, and the combustion rate is controlled by oxygen diffusion and surface reaction kinetics [20]. The corresponding equation is as follows:

$${\displaystyle \frac{\mathrm{d}{\mathrm{m}}_{\mathrm{p}}}{\mathrm{dt}}} = {-\mathrm{A}}_{\mathrm{p}}{\mathrm{P}}_{\mathrm{OX}}{\displaystyle \frac{{\mathrm{D}}_0\mathrm{K}}{{\mathrm{D}}_0+\mathrm{K}}}$$

(7)

$${\mathrm{D}}_0={\mathrm{C}}_1{\displaystyle \frac{{\left[\left({\mathrm{T}}_{\mathrm{P} }+{\mathrm{T}}_{\infty }\right)/2\right]}^{0.75}}{{\mathrm{d}}_{\mathrm{p}}}}$$

(8)

$$\mathrm{K}={\mathrm{C}}_2{\mathrm{e}}^{-\left({\mathrm{E}}_2/\mathrm{R}{\mathrm{T}}_{\mathrm{p}}\right)}$$

(9)

where ${\mathrm{m}}_{\mathrm{p}}$ represents the mass of the fuel particles (kg); ${\mathrm{A}}_{\mathrm{p}}$ denotes the surface area of the particles (m²); ${\mathrm{P}}_{\mathrm{OX}}$ signifies the partial pressure of the gas surrounding the particles (Pa); ${\mathrm{d}}_{\mathrm{p}}$ is the particle size (m); ${\mathrm{T}}_{\infty }$ and ${\mathrm{T}}_{\mathrm{P}}$ represent the temperature of the surrounding gas and the initial particles (m), respectively; ${\mathrm{C}}_1$ is the diffusion rate constant; and ${\mathrm{C}}_2$ is the pre-exponential factor of the reaction rate.

The P1 radiation model is used to simulate high-temperature gas–solid radiation heat transfer, which dominates heat transfer in precalciners. The expression is as follows:

$${\mathrm{q}}_{\mathrm{r}} = -\Gamma \nabla \mathrm{G}$$

(10)

$$\Gamma ={\displaystyle \frac{1}{3\left(\mathrm{a}+{ \sigma }_{\mathrm{s}}\right) - \mathrm{C}{\sigma }_{\mathrm{s}}}}$$

(11)

$$\nabla \left(\Gamma \nabla \mathrm{G}\right) - \mathrm{aG}+4\mathrm{a}\sigma {\mathrm{T}}^4={\mathrm{S}}_{\mathrm{G}}$$

(12)

where ${\mathrm{q}}_{\mathrm{r}}$ is the radiative heat flux (W/m²); $\Gamma$ is the diffusion coefficient (m); $\mathrm{G}$ is the incident radiation (W/m²); a represents the absorption coefficient (m⁻¹); ${\sigma }_{\mathrm{s}}$ is the scattering coefficient (m⁻¹); $\mathrm{C}$ represents the coefficient of the linear anisotropic phase function; $\sigma$ is the Stefan–Boltzmann constant (5.67 × 10⁻⁸ W·m⁻²·K⁻⁴); and ${\mathrm{S}}_{\mathrm{G}}$ is a volumetric user-defined radiation source (W/m³).

2.4. Raw Material Decomposition

In this study, the Species Transport model and the Finite-Rate/Eddy-Dissipation model are employed to simulate the decomposition of the raw material (CaCO₃). The equations are as follows:

$${\displaystyle \frac{\partial }{{\partial }_{\mathrm{t}}}}\left(\rho {\mathrm{Y}}_{\mathrm{i}}\right) + \nabla \left(\rho \overrightarrow{\mathrm{v}}{\mathrm{Y}}_{\mathrm{i}}\right) = -\nabla {\mathrm{J}}_{\mathrm{i}} + {\mathrm{R}}_{\mathrm{i}} + {\mathrm{S}}_{\mathrm{i}}$$

(13)

where ${\mathrm{Y}}_{\mathrm{i}}$ represents the mass fraction of the substance $\mathrm{i}$; ${\mathrm{R}}_{\mathrm{i}}$represents the net production rate of the chemical reaction (kg/(m³·s)); ${\mathrm{S}}_{\mathrm{i}}$ denotes the additional generation rate caused by the discrete phase and user-defined source terms (kg/(m³·s)); and ${\mathrm{J}}_{\mathrm{i}}$ signifies the diffusion flux of the substance (kg/(m²·s)).

The decomposition reaction equation for CaCO₃ is expressed as follows:

$${\mathrm{CaCO}}_3\left(\mathrm{s}\right)\to \mathrm{CaO}\left(\mathrm{s}\right) + \mathrm{C}{\mathrm{O}}_2\left(\mathrm{g}\right) ∆\mathrm{H} = +178\mathrm{kJ}/\mathrm{mol}$$

(14)

The pre-exponential factor for the CaCO₃ decomposition reaction is 1 × 10⁷ s⁻¹, and the decomposition rate of raw material is defined as follows:

$$\omega = {\displaystyle \frac{{\mathrm{m}}_1{- \mathrm{m}}_2}{{\mathrm{m}}_1}}\times 100\%$$

(15)

where ${\mathrm{m}}_1$ represents the total mass flow of CaCO₃ at the inlet for raw materials (kg/s) and ${\mathrm{m}}_2$ represents the total mass flow of CaCO₃ at the outlet of the precalciner (kg/s).

2.5. NO_X Generation Model

In Fluent, there are three types of NO_X compounds: rapid NO_X, thermal NO_X, and fuel NO_X. Due to the minimal amount of rapid NO_X generated within the precalciner, this study solely focuses on the generation of thermal NO_X and fuel NO_X in the precalciner [21]. The formation of thermal NO_X primarily results from the oxidation reaction between nitrogen and oxygen in the air at elevated temperatures. The generation of fuel NO_X is a highly intricate process. In this study, we assume that during the formation of NO from volatile nitrogen (N) and coke nitrogen (N), the intermediate product is always HCN. Nitrogen present in coke can be directly converted to NO, whereas nitrogen in volatile matter must be converted to NO via HCN.

2.6. Solution Methods

ANSYS Fluent 2022 R1was used for calculations, and the control equations were discretized using the finite volume method. We have chosen the SIMPLE algorithm, which is based on pressure–velocity coupling, which allows for a more accurate consideration of the relationship between pressure and velocity. All discretization schemes are second-order upwind schemes.

2.7. Model Assumptions and Limitations

In the numerical framework, several simplifying assumptions are adopted that may influence the predictions at high rice husk substitution rates. For turbulence, the Realizable $\mathrm{k}\text{-}\epsilon$ model with standard wall functions is employed, which assumes isotropic eddy viscosity and may potentially underestimate mixing in strongly swirling regions, thereby affecting the burnout prediction. The discrete phase model (DPM) treats fuel and raw-meal particles as spherical with two-way coupling, neglecting particle shape changes and fragmentation. Devolatilization is not explicitly modeled; instead, instantaneous volatile release is assumed at the fuel inlet. This simplification tends to overestimate the early heat release and local temperature, and the effect becomes more pronounced at higher rice husk fractions owing to the fuel’s high volatile content. Char combustion is described by a kinetics/diffusion-limited model with constant activation energy and pre-exponential factor, which does not account for the progressive inhibition of char reactivity due to ash accumulation at high burnout, possibly leading to an overprediction of the overall burnout rate. Regarding NO_X formation, only fuel NO_X and thermal NO_X are considered, with HCN taken as the sole intermediate; prompt NO_X is ignored. Finally, radiation is modeled using the P1 approximation, neglecting local variations in radiative properties due to soot and ash particles.

3. Basic Parameters of the Study

3.1. Geometric Model and Meshing

In this study, the ILC and SLC precalciner systems of an 8000 t/d cement production line in actual operation at a cement plant were the focus. Figure 1 provides a simplified schematic of raw material flow in and out of the precalciner system. The raw material entering the ILC precalciner undergoes decomposition, passes through a fourth-stage preheater, and then enters the SLC precalciner, rather than directly proceeding to the rotary kiln.

Figure 2a depicts the geometric model of the SLC precalciner. With a total height of 25.68 m, this precalciner features a tertiary air pipe at the bottom, an RH inlet beside the pipe, a cone height of 4.16 m, and two symmetrically positioned coal injection pipes on both sides. The column stands at 20 m high, with two raw material pipes distributed at the lower part, forming an angle of 30°. The top serves as the outlet for waste gas and materials. Figure 2b illustrates the geometric model of the ILC precalciner. Standing at a total height of 25.12 m, the precalciner primarily consists of a lower cylinder, a cone, an upper cylinder, and a transverse outlet. Flue gas from the kiln tail enters the precalciner via the bottom of the lower cylinder. On one side of the lower cylinder is the rice husk (RH) inlet. The cone, standing at 2.8 m high, features two symmetrically positioned coal injection pipes on both sides. On both sides of the upper cylinder, there is a tertiary air pipe and a raw material pipe. The tertiary air pipe enters the precalciner radially. As the powder in the gas flow moves upward, it collides with the top and rebounds downward, thereby extending the powder’s residence time within the furnace and facilitating an improvement in the decomposition rate.

In this study, the geometric models of the ILC and SLC precalciners were meshed into hexahedral structured grids. The hexahedral structured grid elements were arranged regularly, resulting in high-quality grids that enhance calculation efficiency and facilitate convergence. Grid independence verification ensured that the results were independent of grid number. Five grid sizes were selected: 483,300, 544,300, 625,700, 711,800, and 875,500 for the ILC precalciner and 718,800, 883,200, 932,000, 1,039,080, and 1,116,580 for the SLC precalciner.

Table 1. ILC precalciner outlet temperature comparison.

Number of Grids	483,300	544,300	625,700	711,800	875,500
Simulated temperature (K)	1157	1155	1154	1158	1159
Actual temperature (K)	1134

and

Table 2. SLC precalciner outlet temperature comparison.

Number of Grids	718,800	883,200	932,000	1,039,080	1,116,580
Simulated temperature (K)	1174	1169	1175	1167	1170
Actual temperature (K)	1160

compare the simulated and measured outlet temperatures. The listed outlet temperature error of less than 1% indicates grid independence. After comprehensive consideration, 625,700 grids were chosen for the ILC precalciner and 1,039,080 grids for the SLC precalciner in the subsequent simulation research. The minimum value of Determinant 2 × 2 × 2 for all grid qualities was 0.45, and the minimum angle was 27°, satisfying the requirements for calculation accuracy.

3.2. Material Characteristics and Boundary Factors

Table 3. Proximate and ultimate analyses of fuels.

Item	Proximate Analysis (wt%)				Ultimate Analysis (wt%)					Qnet, Ar (MJ/kg)
	Mad	Aad	Vad	FCad	Cad	Had	Oad	Nad	Sad
Coal	19.24	10.28	37.58	32.90	49.20	3.11	17.34	0.74	0.09	17.96
RH	6.88	15.38	63.03	14.71	38.48	5.29	37.89	0.33	0.02	14.76

presents the proximate and ultimate analyses of the pulverized coal and rice husk utilized in the cement production line. The high ash content (15.38 wt%) and silica-rich composition of the rice husk have important implications for precalciner operation. During combustion, the ash may form low-melting alkali silicates that increase the risk of slagging on refractory surfaces and fouling in downstream ducts. The silica can also alter the clinker mineralogy if carried into the rotary kiln in significant quantities. From a combustion kinetics perspective, the ash layer that forms around burning char particles acts as a diffusion barrier, potentially reducing the char burnout rate at high substitution levels—an effect not captured by the simple kinetics/diffusion model used in this study. Furthermore, the high volatile content, while beneficial for ignition, may lead to early oxygen depletion in the fuel-rich zones, influencing NO_X reduction through reburning mechanisms. These aspects imply that the optimal blending ratios derived in this work should be viewed as upper limits from a thermal and decomposition standpoint, with ash-related constraints possibly requiring lower substitution rates in plants susceptible to build-up problems.

Table 4. Chemical composition of the raw meals.

Component	SiO2	Al2O3	Fe2O3	CaO	MgO	Loss on Ignition
%	13.17	3.97	2.29	43.28	0.96	35.49

outlines the chemical composition of the raw material employed in the cement production line.

Based on actual production and measurement data, the boundary conditions utilized in the simulation were determined, which are outlined in

Table 5. Boundary conditions of ILC precalciner.

Boundary	Tertiary Air	Klin Exhaust Gas	Coal	RH	Raw	Outlet
Temperature (K)	1113	1418	323	323	989	1134
Pressure (Pa)	−839	−300	19,361	−780	4051	−776
Velocity	21.4 m/s	30 m/s	3.6 kg/s	0.4 kg/s	50 kg/s
Turbulent intensity (%)	5	10	5	5	5	5
Hydraulic diameter (m)	1.7	3.7	0.254	0.542	1	4.765

and

Table 6. Boundary conditions of SLC precalciner.

Boundary	Tertiary Air	Coal	RH	Raw-1	Raw-2	Outlet
Temperature (K)	1177	333	333	1035	1104	1160
Pressure (Pa)	−510	22,668	−860	5506	−2376	1923
Velocity	23 m/s	5.4 kg/s	0.6 kg/s	27 kg/s	27 kg/s
Turbulent intensity (%)	10	5	5	5	5	10
Hydraulic diameter (m)	3	0.254	0.542	1	1	4.756

The pressures are the gauge pressures at their respective inlets. Turbulence intensity was determined based on industrial inlet pipe characteristics and standard CFD setup for cement precalciners. Hydraulic diameter is the equivalent diameter of a circular duct that has the same cross-sectional area and perimeter as the actual non-circular inlet. Klin exhaust gas refers to the high-temperature flue gas coming from the rotary kiln into the ILC precalciner, which provides additional heat and turbulence for combustion and raw material decomposition.

. By maintaining the total fuel mass flow constant and varying the mass flow rates of the pulverized coal and rice husk according to different blending ratios of rice husk, four sets of operating conditions were established (with rice husk proportions of 10%, 20%, 30%, and 40%, respectively).

3.3. Model Verification

The actual rice husk blending ratio in both the ILC and SLC precalciners was 10%, and the model validation was carried out under this operating condition. To evaluate the reliability of the simulation, the predicted outlet values were compared with the measured data obtained from the thermal calibration of the precalciner, which is summarized in

Table 7. Verification of ILC precalciner.

	Simulation	Measured	Error (%)
Temperature (K)	1154	1134	2.91
CO2 content (%)	21.8	22.6	3.54
O2 content (%)	1.81	1.87	3.21

and

Table 8. Verification of SLC precalciner.

	Simulation	Measured	Error (%)
Temperature (K)	1167	1160	0.60
CO2 content (%)	26.8	26.1	2.68
O2 content (%)	3.52	3.41	3.23

. The deviations between the simulated and measured values are below 5%, which is considered acceptable for engineering applications, thereby confirming the reliability of the simulation results.

Validating the internal distributions of temperature and species is challenging because industrial precalciners are not equipped with spatially resolved measurement ports. As an indirect validation, the predicted temperature and CO₂ fields in the TTF precalciner for the base case were compared with the CFD results reported by Wei and Kao [21] for a similar configuration. The overall pattern—a high-temperature region near the fuel inlet, a rapid temperature drop upon raw meal introduction, and a gradual temperature increase in the upper column—agrees qualitatively with their findings. This consistency, together with the quantitative agreement of the outlet parameters (Table 7 and Table 8), provides confidence that the simulation adequately captures the principal flow, combustion, and decomposition characteristics relevant to the parametric study.

4. Simulation Results and Discussion

4.1. Influence on the Temperature Field

Figure 3 illustrates the temperature contour cloud map on the Z = 0 section within the ILC precalciner at varying rice husk blending ratios. High-temperature kiln tail flue gas enters the main combustion zone from the bottom. Upon entering the furnace, pulverized coal and rice husk swiftly release volatile matter, which reacts violently with fixed carbon, generating substantial heat and rapidly decomposing the raw material. Because of the precalciner’s internal structure, the high-temperature gas tends to distribute to the left, leading to higher temperatures on the left and lower temperatures on the right in the middle and upper sections of the furnace. As the rice husk blending ratio increases, the extent of the high-temperature zone noticeably shrinks. Figure 4 displays the average gas temperature at various heights within the ILC precalciner across different rice husk blending ratios. The graph demonstrates a consistent temperature trend within the furnace regardless of the rice husk blending ratio. With an increase in rice husk blending ratio, the overall precalciner temperature drops significantly due to the low calorific value of rice husk. Between heights of 5–8 m, the introduction of raw material and tertiary air causes a sharp decrease in the gas’s average temperature. Beyond a height of 8 m, the precalciner’s average gas temperature stabilizes gradually. Ultimately, the furnace top temperature rises as the powder is obstructed at the top, prompting further combustion of unburned coke.

Figure 5 illustrates the average gas temperature along the SLC precalciner’s height under different rice husk blending ratios. The bottom of the precalciner serves as the inlet for tertiary air, with rice husk entering the main combustion zone alongside it and undergoing intense combustion. Simultaneously, raw material enters from the cone top, rapidly absorbs heat, and decomposes significantly, causing a sharp drop in the average gas temperature. Subsequently, as the volatile components of the fuel are depleted, the residual coke continues to burn and release heat, leading to a slight increase in furnace temperature with height.

When the rice husk blending ratios are 10%, 20%, 30%, and 40%, the average gas temperatures at the outlet of the ILC precalciner are 1154 K, 1152 K, 1150 K and 1149 K, respectively. For the SLC precalciner, the outlet temperatures are 1167 K, 1161 K, 1159 K and 1155 K, respectively. The observed reduction in peak temperature and shrinking of the high-temperature zone with increasing rice husk fraction can be attributed to two coupled effects. First, rice husk has a lower calorific value and a higher volatile content than pulverized coal, leading to a more distributed heat release and lower local flame temperatures. Second, the rapid devolatilization of rice husk consumes oxygen early in the main combustion zone, slightly delaying char burnout and spreading the heat release over a larger volume, which further suppresses peak temperatures. This temperature reduction directly lowers the CaCO₃ decomposition rate because the endothermic calcination reaction is strongly temperature-dependent and, in the ILC precalciner with its shorter residence time, the effect is more pronounced.

4.2. Influence on Raw Material Decomposition

Figure 6 and Figure 7 respectively depict the average mass fraction of CaCO₃ and CaO along the SLC precalciner’s height under different rice husk blending ratios. Raw material entering the main combustion zone causes a sharp rise in the CaCO₃ mass fraction at a 4–5 m height. Subsequently, the mass fraction of CaCO₃ drops sharply. This is attributed to intense heat exchange, which rapidly decomposes the raw material, generating a substantial amount of CaO and CO₂. Simultaneously, the influx of tertiary air induces a dilution effect, resulting in a decrease in the mass fraction of CaO. Following this, the mass fraction of CaCO₃ persistently decreases, while that of CaO persistently increases, indicating continuous decomposition of the raw material after its entry into the furnace. Furthermore, as the proportion of rice husk replacing pulverized coal increases, the overall mass fraction of CaCO₃ in the precalciner gradually rises.

Figure 8 and Figure 9 respectively illustrate the average mass fraction of CaCO₃ and CaO at varying heights in the SLC precalciner when blending different ratios of rice husk. Tertiary air carries a substantial amount of O₂ into the furnace from the bottom, triggering intense combustion of the fuel and releasing a large amount of heat. The raw material outside the raw material bundle undergoes rapid endothermic decomposition. At approximately 12 m, the mass fraction of CaCO₃ spikes due to raw material agglomeration without sufficient diffusion, concurrently causing a sudden decrease in the corresponding mass fraction of CaO.

In addition, as the proportion of rice husk replacing pulverized coal increases, the overall mass fraction of CaCO₃ in both ILC and SLC precalciners gradually rises. Simultaneously, the overall mass fraction of CaO exhibits a gradual downward trend. The rising proportion of rice husk replacing pulverized coal leads to a gradual decrease in the overall temperature within the furnace, subsequently reducing the decomposition rate of the raw materials. The calculation results indicate that when the rice husk blending ratio is 10%, 20%, 30%, and 40%, the decomposition rates of CaCO₃ at the outlet of the ILC precalciner are 81.11%, 79.25%, 77.36% and 75.32%, respectively. For the SLC precalciner, the corresponding rates are 93.27%, 92.98%, 91.58% and 88.50%. Considering that the decomposition rate of the raw materials at the outlet of the precalciner should be maintained between 85% and 95% in practical production, a rice husk replacement ratio of 10% or less is recommended for the ILC precalciner, while a ratio of 40% is suggested for the SLC precalciner.

The observed decrease in precalciner exit temperature and CaCO₃ decomposition rate with increasing rice husk fraction has direct implications for downstream clinker quality and kiln stability. A precalciner exit with a decomposition rate below 85% increases the thermal load on the rotary kiln, potentially requiring a higher kiln burning zone temperature to complete calcination and form the required alite (C₃S) phase. This additional thermal stress can accelerate brick wear and increase fuel consumption in the kiln, partially offsetting the substitution benefits. Moreover, the lower and more uniform precalciner temperatures may reduce the formation of reactive CaO with a high specific surface area, which could affect the subsequent combination with SiO₂ and the development of early clinker strength. From an operational standpoint, the 40% rice husk case in the SLC precalciner still meets the decomposition target, but operators should monitor the kiln current and free lime content to ensure stable burning conditions. These considerations reinforce the need to view the recommended blending ratios not only from a decomposition and emission perspective but also considering the overall pyro-processing system balance.

4.3. Influence on Composition Fields

Figure 10 and Figure 11 depict the distribution of O₂ and CO₂ on the Z = 0 section within the ILC precalciner. It is evident from Figure 11 that the high-concentration area of O₂ is primarily located near the tertiary air inlet, with the O₂ concentration gradually decreasing as the height increases. In the main combustion zone, the intense combustion of a substantial amount of pulverized coal rapidly consumes O₂, leading to a swift decline in its concentration. Simultaneously, coal combustion and raw material decomposition release a considerable amount of CO₂, resulting in a high CO₂ concentration in this region, as illustrated in Figure 12. As the height increases further, O₂ continues to slowly decrease due to the slow combustion of residual unburned coke in the pulverized coal, which continuously consumes O₂. Accompanied by the combustion of unburned coke and the ongoing decomposition of raw material, CO₂ is continuously generated, further enriching it in the upper region. Due to the biased distribution of high-temperature gas towards the left side of the furnace, the decomposition rate of the raw material on the left side is higher, leading to a higher CO₂ concentration.

Figure 12 and Figure 13 respectively present the average mass fraction of O₂ and CO₂ along the height of the SLC precalciner. These figures clearly demonstrate that as the proportion of rice husk replacing pulverized coal increases, the production of CO₂ significantly decreases. This phenomenon can be attributed to the fact that rice husk, compared to pulverized coal, has a higher oxygen content and lower carbon content, and thus requires less O₂ and releases less CO₂ during combustion. Additionally, the decreased rate of raw material decomposition also results in reduced CO₂ emissions.

Figure 14 and Figure 15 respectively depict the NO distribution on the Z = 0 section in the ILC and SLC precalciners under varying rice husk blending ratios. It is evident from the figures that the local high-concentration areas of NO are primarily situated near the fuel inlet. Within this region, intense combustion of fuel enriched with O₂ leads to the generation of a substantial amount of fuel-type NO_X and a smaller quantity of thermal-type NO_X, resulting in a high NO concentration. Above the main combustion zone, as the volatile matter and residual nitrogen in the fixed carbon continue to oxidize, the NO concentration steadily increases with height. Notably, as the proportion of rice husk replacing pulverized coal increases, the overall NO concentration in the precalciner gradually decreases. This phenomenon occurs because rice husk contains a lower nitrogen content compared to pulverized coal, thereby producing less fuel-type NO_X during combustion. Additionally, elevating the rice husk blending ratio significantly lowers the maximum temperature within the furnace, further reducing the generation of thermal-type NO_X.

In the SLC precalciner (optimal rice husk ratio: 40%), compared with the 10% blending ratio, the NOx emission is reduced by 16.0%, and the CO₂ emission is reduced by 4.7% under the optimal 40% rice husk blending condition. For the ILC precalciner, the recommended blending ratio is limited to 10% to ensure an adequate decomposition rate. Compared with a 40% blending ratio, the NOx reduction potential can reach 21.7%.

5. Conclusions

This study, utilizing actual production data from a cement plant in Indonesia, investigated the mixed combustion of pulverized coal and rice husk in a precalciner, employing computational fluid dynamics methods.

5.1. Verified Findings

The blending proportion of rice husk significantly impacts the temperature distribution within the precalciner. As the rice husk blending ratio increases, the high-temperature zone shrinks, peak temperatures decline, and the overall furnace temperature gradually decreases. Specifically, the average gas temperature at the outlet of the ILC precalciner dropped from 1154 K to 1149 K, while that of the SLC precalciner decreased from 1167 K to 1155 K. The blending proportion of rice husk reduced the decomposition rate of the raw material in the precalciner.

5.2. Operational Recommendations

When the rice husk blending ratios were 10%, 20%, 30% and 40%, the CaCO₃ decomposition rates at the outlet of the ILC precalciner were 81.11%, 79.25%, 77.36% and 75.32%, respectively; for the SLC precalciner, the rates were 93.27%, 92.98%, 91.58% and 88.50%, respectively. The blending proportion of rice husk reduced the emission of CO₂ and NO gases, significantly narrowing the high-concentration areas of these gases and notably decreasing the local maximum concentration of NO. Furthermore, a higher substitution ratio led to a more pronounced emission reduction effect. Taking into account production demands and pollutant emission reduction, a 10% or less rice husk blending ratio was deemed more suitable for the ILC precalciner, while a 40% blending ratio was more appropriate for the SLC precalciner.

5.3. Model Limitations

The present study has several limitations that should be considered when interpreting the results. First, the CFD model employs several simplifications, which may affect the accuracy of local temperature and species predictions at high rice husk fractions. Second, the boundary conditions were derived from a single thermal calibration campaign at one cement plant; therefore, the quantitative optimal blending ratios are strictly applicable to similar precalciner geometries and operating conditions. Third, this work did not incorporate a comprehensive ash chemistry model, so slagging, fouling, and clinker quality impacts are only discussed qualitatively. Finally, an economic analysis of rice husk handling and potential grinding requirements was beyond the scope of this simulation study.

The qualitative trends of temperature reduction, lower decomposition, and emission reduction are expected to hold for other precalciner geometries when using high-volatile, low-nitrogen biomass fuels similar to rice husk. However, the exact optimal blending ratios are specific to the studied 8000 t/d line and its operating window. Plants with longer solid residence times or active temperature control may be able to accommodate slightly higher substitution rates. Practical implementation requires adjustments to fuel storage and transport, recalibration of fuel feeding and excess air, and close monitoring of clinker quality (particularly free lime and silica modulus) during the fuel transition.