Optimization of Atmospheric Fluidized Bed Combustion Boiler by Computational Fluid Dynamics

. Viscose rayon is a natural fabric from regenerated cellulose derived from wood pulp, bamboo, or cotton linters. The viscose rayon industry relies on electricity to generate steam for textile production. This study aims to optimize the 45-ton-per-hour Atmospheric Fluidized Bed Combustion (AFBC) boiler. Its efficiency is 78.16%, lower than the industry benchmark of 85.78%. Boiler efficiency decreases as the flue gas temperature increases, reaching 220 degrees Celsius, indicating suboptimal coal fuel utilization for water heating. The study explores two controllable variables: coal calorific value and combustion air mass flow rate for optimum efficiency. ANSYS Fluent 2021, a Computational Fluid Dynamics (CFD) software, is used for analysis. Boiler efficiency is calculated using the indirect method following ASME PTC 4.0. The research shows that adjusting the combustion air mass flow rate and coal calorific value increases efficiency to 81.04%.


Introduction
Viscose rayon, also known as viscose, is a regenerated cellulose fiber made from natural sources such as wood pulp, bamboo, or cotton linters.However, its production process requires a significant amount of electric power to run the production equipment and generate the necessary gas steam.Typically, multiple production lines are employed simultaneously in the viscose industry.Figure 1 shows an example of an industry utilizing five production plants simultaneously, with each plant requiring different amounts of steam: 6.7 tons per hour (tph), 7.8 tph, 10.0 tph, 14.9 tph, and 9.9 tph, respectively.The required steam can vary depending on the production volume, with a linear relationship observed between steam demand and production.In the industry observed, the steam demand for viscose fiber production has been steadily increasing over Corresponding author: fajar.ramadhan002@binus.ac.id the past three months, in line with the increased steam production from the Energy and Utilities Department, which is responsible for producing steam to support the main viscose rayon fiber production process.This department consists of Energy Production 1, Energy Production 2, and Utilities units.In addition to the steam required for the main process, the produced steam is also used to drive turbines, generating electrical energy.The energy requirements for the rayon industry are supplied from various departments, with each having a capacity of 9900 MW from National Electric Power Company (Perusahaan Listrik Negara, PLN), 13100 MW from Energy Production 2, and 9500 MW from Energy Production 1. Notably, the power supplied by PLN is the lowest according to the cooperation agreement and regulations governing the use of electrical energy for the industry.Energy Production 2 plays a significant role in electrical and steam production for viscose production processes.
The power plant comprises one Circulating Fluidized Bed Combustion (CFBC) boiler with a capacity of 130 tph and a maximum power generation of 20.5 MW.It has been operational since 2008 and has maintained an impressive efficiency of 88%, aligning with the company's efficiency standards.Our research is conducted at Energy Production 1, which consists of four Atmospheric Fluidized Bed Combustion (AFBC) boiler units.Three units have an individual capacity of 22.5 tph, and the fourth unit has a capacity of 45 tph.Initially, the power plant used coal and sludge as fuel.However, as per the Minister of Environment Regulation No. 21 of 2008, the emission standard of sulfur dioxide (SO 2 ) in steam power plants was set at 550 mg/Nm 3 , resulting in the discontinuation of sludge as fuel.Sludge is a hazardous industrial waste that contains high levels of sulfur.The production cost includes three cost components: raw material cost, employee cost, and additional expenses.
A significant portion of the operating cost for any industrial process is related to fuel consumption [1].The largest production cost component for Energy Production 1 is the raw material cost, specifically the cost of purchasing coal.The cost of purchasing coal accounts for 30% of the total production cost incurred in producing viscose fiber.The high cost of coal is due to the rising price of coal.According to the Indonesia Coal Index (ICI 4) in July 2022, the price of coal with a calorific value of 5100 kcal/kg is Rp2,000,000 per ton.Potential energy savings from coal can minimize production costs more efficiently.
Boilers are pressure vessels used to heat water or produce steam for heating and electricity generation in industries [2].The heat required to vaporize the water is obtained from the combustion of coal in the combustion chamber.The boiler under investigation is an AFBC type with a capacity of 45 tph, which has a problem in its operating process, specifically a decrease in efficiency from 85.78% to 78.61%.
Heat can be lost from the boiler in various ways, such as flue gas loss, radiation, blowdown, and others.The boiler faces several problems, including a high flue gas absorption between combustion gas and feedwater in the furnace.Reduction of flue gas temperature can result in potential energy savings by reducing coal consumption and saving production costs.The boiler also experiences a quick decrease in bed temperature, which leads to frequent sand regeneration.Poor coal quality creates new problems such as relative frequency imbalance in the Induced Draft Fan (IDF) and clogging of coal supply from the bunker to the screw feeder due to high coal moisture content.
This research builds on the findings of several previous studies.Firstly, in Ref. [3], the efficiency of a boiler in a 660 MW power plant was improved by optimally adjusting the excess air coefficient.Secondly, in another study [4], it was suggested that the AFBC boiler efficiency could be enhanced by using feedwater extracted from a turbine steam and then heating it with a high-pressure heater.The third study demonstrated that increasing oxygen (O2) concentration in the inlet channel can raise the temperature inside the furnace, lower carbon dioxide (CO2) emissions, and effectively reduce carbon monoxide (CO) emissions in the exhaust gases.The fourth study emphasized the importance of using the appropriate coal particle size to achieve optimal combustion efficiency [5].Lastly, the fifth study found that optimizing the coal and combustion air mixing ratio prior to entering the combustion chamber could boost boiler efficiency by 4%, resulting in reduced operating costs and air emissions [1].
This research aims to enhance the efficiency of the AFBC boiler by optimizing the Gross Calorific Value (GCV) of coal thorough determining the optimal coal mixing ratio, analyzing the calorific value and content in the coal mixture, adjusting combustion air to achieve efficient combustion, and reducing the flue gas heat loss.
The manuscript is organized into three parts.Firstly, we explore the impact of increasing coal calorific value and mass flow rate of combustion air on boiler efficiency using CFD simulations in ANSYS Fluent 2021 R2 software.Secondly, we determine the boiler efficiency based on the ASME PTC 4.0 standards.Lastly, we evaluate the cost savings resulting from the enhanced boiler efficiency.

Research stages
The problem-solving stages in this research can be illustrated through Figure 2  Determining the most optimal coal blending ratio, calculating the mass of the blend, the calorific value of the blend, and calculating the content of each coal element.
Calculating the optimal air requirement for combustion.
Performing Computational Fluid Dynamics (CFD) simulation of the post-improvement conditions using ANSYS Fluent 2021 R2 to determine the temperature, velocity, and composition of the exhaust gas, as well as observe the combustion process occurring in the combustion chamber.
Calculating the efficiency of the AFBC boiler under the conditions after coal and combustion air optimization.
Is the efficiency of the boiler after optimization greater than the actual boiler efficiency, and is the exhaust gas temperature after optimization lower than the actual Calculating cost savings in production.

CFD Simulation Method in ANSYS Fluent 2021 R2
The present research aims to develop a computational fluid dynamics (CFD) simulation model for industrial boilers, specifically focusing on the analysis of fluid flow and heat transfer, such as temperature, velocity, and density distribution, utilizing ANSYS Fluent software [6].
Moreover, this study employs ANSYS Fluent 2021 R2 to analyze the flue gas fluid, heat transfer process in the combustion chamber, flue gas velocity, and gas composition.CFD is a reliable method for analyzing the temperature characteristics of fluid flow, which eliminates the need for direct testing.The CFD simulations in this research follow three stages: Pre-Processing, Processing, and Post-Processing.The study focuses on a 45 tph AFBC boiler and aims to analyze the processes that occur within the boiler involving fluid flow from combustion air, coal, and flue gas. Figure 2 illustrates the CFD simulation model of the boiler.The AFBC boiler model shown in Figure 2 was created using Inventor software and subsequently exported to the ANSYS Fluent 2021 R2 compatible format for simulation.The model was designed from the combustion chamber to the flue gas channel to achieve optimal simulation results and high accuracy levels.Once the model was created, the continuous solid domain was transformed into a discrete computational domain using a finite number of elements, allowing the structural equations to be solved using numerical CFD methods.
The model depicted in Figure 3 has undergone hexahedral discretization, which was selected for its superior regularity, alignment, expressiveness, and accuracy with fewer elements, thereby making the simulation more efficient [7].The subsequent step involves defining boundary conditions to establish parameters as the basis for numerical computation.This process is crucial as it affects the outcomes of numerical analysis.To account for varying rates of heat transfer, the model is divided into three zones: economizer, steam drum, and superheater.The heat transfer rates in these zones, as calculated using actual data, are 4094 kJ/s, 24628 kJ/s, and 5930 kJ/s, respectively.The process parameter data for all variations used to define the CFD simulation boundary conditions are presented in Table 1.The presented process parameter data in Table 1 defines the boundary conditions for CFD simulation.The table also shows the three simulated conditions: Variation 1 represents the actual condition, Variation 2 represents improvement condition 1, and Variation 3 represents improvement condition 2. The variations differ in the parameters of coal mass flow rate, coal calorific value, and combustion air requirement.

Boiler Efficiency Method
The performance and emissions testing of industrial coalfired boilers can be carried out using the energy balance method, which is standardized by the American Society of Mechanical Engineer Power Test Code (ASME PTC 4.0-2013) [8].Boiler efficiency can be calculated by either the direct method or the indirect method [9].For analyzing efficiency, energy and exergy analyses, as well as exhaust gas flow conditions in coal-fired boilers, the indirect method and calculation of boiler losses are commonly used [10].The indirect method is generally considered more effective and accurate than the direct method, and the largest heat losses are often associated with dry flue gas and moisture content in the fuel [11].In this research, the indirect method, or the calculation of heat losses in the boiler is employed.The losses in liquid, gas, and solid fuel-fired boilers are shown below.We define the following parameters: : Loss due to dry flue gas : Loss due to hydrogen in fuel (H2) : Loss due to moisture in fuel (H2O) : Loss due to moisture in air (H2O) : Loss due to carbon monoxide (CO) : Loss due to radiation, convection, and others : Loss due to unburned carbon in fly ash : Loss due to unburned carbon in bottom ash The boiler efficiency in percent is computed by: The amount of the theoretical air required for combustion is obtained by: where C represents the percentage of carbon, H is the percentage of hydrogen, O is the percentage of oxygen, and S is the percentage of sulfur.
The excess air in percentage is estimated by: where (CO ) is the theoretical CO in percent and (CO ) is the actual measured CO in the flue gas analyzer in mol C, and the (CO ) is obtained by: The term actual mass of air supplied per kilogram of fuel (AAS) refers to the amount of air supplied to the combustion process in relation to the amount of fuel used: x theoretical air (4) The heat loss due to the dry flue gas: where is the mass of dry flue gas in kg/kg fuel, which consists of CO mass, N mass in the fuel, N mass in the supplied combustion air, O mass in the exhaust gass.The H O or water vapor in the exhaust gas cannot be considered, is the specific heat of the flue gass in J/kg K, is the flue gas temperature in K and is the ambient temperature in K.
The heat loss due to evaporation of the water formed by in the fuel is calculated by: where H is the mass of percentage of hydrogen in the fuel and is the specific heat of steam in J/kg K.
The heat loss due to moisture in the fuel is: The heat loss due to moisture in the air is: where is the actual mass of air supplied per kg fuel, and is the moisture factor in kg water/kg dry air.
The heat loss due to incomplete combustion is computed by: where CO is the volume of CO in flue gas analyzer in percentage, CO2 is the actual volume of CO2 in flue gas in percentage and C is the carbon content in kg/kg fuel.
The heat loss due to radiation and convection is: 196.85 + 68.0 68.5 100 (10) where is the velocity of flue gas in m/s.
The heat loss due to unburned carbon in fly ash is: = ash collected × burned fuel/ GCV fly ash GCV flue gas 100 (11) The heat loss due to unburned carbon in bottom ash is: = collected ash × burned fuel/GCV bottom ash GCV flue gas 100 (12)

Result and Discussion
The simulation results are presented in three variations: the existing condition and two improvements (variation 1 and variation 2).These variations are differentiated by their input parameter values, such as coal calorific value, proximate and ultimate analysis, and combustion air mass flow rate.The coal used in the existing condition simulation is based on actual field data.For variations 1 and 2, coal mixing is applied to improve the calorific value, which initially was below the production standard of 5100 kcal/kg.The actual value of 4549 kcal/kg was improved to 5160.5 kcal/kg and 5217.5 kcal/kg, respectively, through coal mixing.These two variations are then used as inputs in the CFD simulation.The combustion air used in the simulation is calculated based on the ultimate analysis of coal variations 1, 2, and 3. Excess air is adjusted by secondary air to obtain an amount of oxygen in the exhaust gas of 5%, which is in accordance with production standards.The CFD simulation provides various parameter values such as exhaust gas temperature, velocity, and mass fraction of gases including CO2, CO, O2, SO2, and NOX.Table 2 shows the simulation results for all variations, namely variation 1 in the actual condition, variation 2 in improvement 1 condition, and variation 3 in improvement 2 condition.The exhaust gas temperature shows an increase with an increase in coal calorific value and combustion air, which is due to the combustion reaction of coal with higher calorific value and optimal combustion air, resulting in high combustion heat as well.The exhaust gas velocity increases with an increase in coal calorific value and combustion air, which is because the hot gas produced from combustion in the combustion chamber becomes higher.The mass fraction of CO2 and CO gases shows opposite results, where an increase in coal calorific value and combustion air can increase CO2 gas and reduce CO gas in the exhaust gas.This is due to more optimal combustion in the combustion chamber.The mass fraction of O2 gas increases due to an increase in the amount of combustion air.The mass fraction of SO2 gas increases because the coal used has an increasing calorific value, and coal with high calorific value tends to have a high sulfur content as well.The mass fraction of NOx gas increases due to the addition of combustion air and the high nitrogen content of the coal.In Figure 4, the temperature distribution in variations 2 and 3 is quite similar due to the similar coal calorific values, whereas in variation 1, the temperature distribution is very different due to the significantly different coal calorific value compared to variations 2 and 3. Fig. 6.The correlation between the calorific value of coal and the exhaust gas temperature Above Figure 5 explains that the in the actual variation is quite low compared to improvement variations 1 and 2, while the standard temperature parameter is in the range of 176three variations fall within the standard range.
After comparing the temperature parameters of the three variations above, additional calculations are carried out for improvement variations 1 and 2 by reducing the mass flow rate of coal so that the exhaust gas temperature does not exceed the actual exhaust gas temperature of temperature in the standard range of 176improving the boiler efficiency.
The next CFD simulation on the boiler is to reduce the mass flow rate of coal entering the combustion chamber with the note that the heat transfer rate that occurs in each zone in the boiler, such as the economizer, steam drum, and superheater, is the same as the heat transfer rate in the actual condition so that the results of improving the boiler efficiency will be accurate and the reduction in coal can be calculated in the simulation.Therefore, it can be said that this CFD simulation aims to determine the lowest exhaust gas temperature that is within the standard range for the minimum coal supply to achieve the most optimal efficiency and savings in the boiler.
Table 3 explains that the reduction of coal consumption supplied to the combustion chamber affects parameters such as the amount of combustion air, exhaust gas temperature, and exhaust gas velocity.As a reference to obtain the most optimal coal saving results, it is necessary to compare the simulation exhaust gas with the standard exhaust gas temperature range of the company.
Table 3 shows that improvement 1 produces the most optimal value when the coal supply to the combustion chamber is 1.05 kg/s with an exhaust gas temperature of optimal value when the coal supply to the system is 0.95 Therefore, it can be concluded from the above simulation results that the most optimal value is found in improvement 2 with the lowest coal supply of 0.95 kg/s.The above Figure 6 shows the temperature distribution in the boiler combustion chamber area, and it can be seen that the temperature is evenly distributed.Compared to the previous Figure 3, the temperature color contour in the superheater, steam drum, and economizer zones appears to be evenly distributed, while in the flue gas duct, the temperature color contour is lower.This result is the most optimal from the aspect of coal supply, combustion air, and resulting flue gas temperature.
Table 4 shows the data of the most optimal improvement results consisting of several parameters, exhaust gas velocity of 2.257 m/s, CO2 mass fraction of 0.021, CO mass fraction of 5.86 x 10-6, O2 mass fraction of 0.00313, SO2 mass fraction of 0.000710, and NOX mass fraction of 1.19 x 10-8.Table 5 explains that the improvement results show a positive outcome as it can save coal consumption in the boiler.The boiler consumption decreased from 8220 kg/h to 3420 kg/h, which is equivalent to a saving of 4800 kg/h.This is a good result because it optimizes the coal consumption for the combustion process in the boiler.The boiler efficiency increased from its initial condition before the improvement.From the table above, the boiler efficiency increased by 2.43%.This is in line with previous studies which stated that optimizing the air-tofuel ratio along with the pre-mixing of air and fuel before entering the combustion chamber can increase boiler efficiency by 4%, resulting in significant energy savings, lower operating costs, and greenhouse gas emission reduction [1].
To calculate the total cost savings of production, the difference between the existing coal consumption and the consumption after improvement is multiplied by the total operating time of the boiler in one year, which is 8424 hours, reduced by the routine overhaul time of 336 hours per year.Then, it is multiplied by the price of mediumrank coal of Rp2,000,000 per ton, resulting in cost savings of Rp80,870,400,000.This is highly recommended to be fixed immediately as it is beneficial for the company.
Improved boiler efficiency can have a profound impact on reducing greenhouse gas emissions, including carbon dioxide (CO2), nitrogen oxides (NOx), and sulfur dioxide (SO2).By decreasing these emissions, the optimization of the boiler contributes significantly to both climate change mitigation and air pollution reduction.Moreover, enhancing the efficiency of coal utilization not only leads to a decrease in the consumption of fossil fuels but also supports the effective management of more sustainable energy resources.Through the utilization of the optimization process and advanced simulation techniques using Computational Fluid Dynamics (CFD) software, it becomes possible to identify and rectify inefficient airflow patterns and combustion processes.This optimization helps in minimizing waste and reducing energy losses, ultimately improving the overall environmental performance of the boiler system.
Implementing boiler optimization solutions using CFD simulation software enables researchers and engineers to conduct analysis and improvements without disrupting actual operations of the boiler.The use of simplified software and calculation methods, such as Excel, facilitates efficient calculations and performance evaluations of the boiler.Data obtained through the Plant Information Management System (PIMS) and other data sources provide reliability and accuracy in conducting analysis and optimization.
The boiler optimization paper makes a significant contribution to the development of more efficient and sustainable practices in the power generation industry or industries utilizing boilers.By improving boiler efficiency, fuel usage can be optimized, and operational costs can be reduced, thereby enhancing the effectiveness and sustainability of the industry.The research findings and proposed solutions in the paper can serve as a reference and guide for other industries in their efforts to improve efficiency and reduce the negative environmental impacts of combustion processes.

Concluding Remarks
The increase in the calorific value of coal and the theoretical amount of air for combustion in the boiler's combustion chamber can increase the temperature and velocity of the flue gas.The mass fractions of O2, CO2, CO, SO2, and NOx decrease as the combustion process in the boiler becomes more optimal.Mixing coal to increase its calorific value and determining the appropriate theoretical amount of air for combustion in the boiler's combustion chamber can increase efficiency from 78.61% to 81.04% and reduce coal consumption from 8220 kg/h to 3420 kg/h.This increase in efficiency results in a total cost savings of Rp80,870,400,000 per year.
The first author would like to express gratitude to the advisor, all lecturers, and colleagues at the BINUS Graduate Program -Master of Industrial Engineering, Bina Nusantara University.The authors utilized ChatGPT to proofread the manuscript.

Fig. 1 .
Fig. 1.The steam demand in the viscose rayon factory Conferences 426, 02074 (2023) https://doi.org/10.1051/e3sconf/202342602074ICOBAR 2023 below.Start Actual process parameter data: -Temperature, pressure, mass flow rate of coal, coal content analysis (proximate, ultimate, ADB) -Temperature, pressure, air flow rate (FD Fan, SA Fan, ID Fan) -Temperature, pressure, and flow rate of feedwater -Temperature, pressure, and flow rate of steam -Baseline boiler efficiency -Bed height in the combustion chamber -Boiler specifications and dimensions Computational Fluid Dynamics (CFD) simulation of the actual boiler conditions using ANSYS Fluent 2021 R2 was performed to investigate the combustion process in the combustion chamber.The efficiency of the AFBC boiler under actual conditions was calculated using the indirect method following the ASME PTC 4.0-2013 standard.Is the efficiency of the boiler in the existing condition lower than the baseline boiler efficiency ( Conferences 426, 02074 (2023) https://doi.org/10.1051/e3sconf/202342602074ICOBAR 2023

Fig. 3 .
Fig. 3.The Computational Fluid Dynamic Model of the AFBC Boiler.

Fig. 4 .
Fig. 4. The Finite Element Model Mesh of the Boiler Model Depicted in Fig. 2

Fig. 5 .
Fig. 5. Temperature distribution for all Figure 4 shows the distribution of flue gas temperature for all variations.Variation 1 in the actual condition has a value of 4549 kcal/kg, variation 2 in improvement 1

Table 1 .
The model process parameter data

Table 2 .
Simulation results for all variations of parameters

Table 3 .
Simulation results of reducing coal consumption

Table 4 .
Simulation Results of Improvement

Table 5 .
Calculation of Improvement Results