Process Simulation and Optimization of Palm Oil Waste Combustion Using Aspen Plus

AbstractThe difficulty of controlling the required air during the incineration of the fiber and shell; the low heating value of solid fuel due to the excessive moisture content of fiber and shell and further formation of slagging or clinker in the reactor as a result of high ash content, are some problems raised when incinerating both fiber and shell. in this simulation work, the effects of air flow rate; moisture content of shell, moisture content of fiber, and moisture content of both shell and fiber; the ash content of fiber and shell and temperature were investigated and optimized on flue gas emissions and the combustion behavior using steady state simulation by ASPEN PLUS (Version 7.1). From the results obtained, the fiber-shell type solid fuel is preferable and the air flow rate should be controlled and suggested maintained at 30% excess air to regulate NOX emission. Besides, the moisture content and ash shows negative effect to the combustion efficiency and the moisture content is suggested in the range of 6%-19.5% for fiber and 5%-13% for shell. Last but not least, the operating temperature is suggested do not exceed 972oC to regulate the NOX emission.

IntroductionWorld demand for energy sources is increasing, and thus, renewable energy sources have become an alternative to the depleting fossil fuel. One type of renewable energy sources comes from combustion of biomass waste, which is also called solid fuel, to produce heat and energy. This is a promising technology to reduce waste and moreover provide a clean and renewable energy source by applying waste-to-wealth concept. Malaysia has become the largest exporter of oil palm product in the world (Foo & Hameed, 2010). While generating huge income from oil palm business, there are abundant of oil palm biomass waste (generall y fiber and shell) generated at the same time. This biomass waste has been utilized to generate energy and electricity to support the mill process. In addition, the fiber and shell are also burnt to generate steam for downstream processes that required steam such as sterilization. As such, a lot of savings can be done because this energy is considered free for the palm milling process. At the same time, using the fiber and shell as boiler fuel can help to dispose these bulky materials which can contribute to environmental pollution. The energy content varies depending on the moisture, residual oil contents and its high specific energy content. in 2003 a simulation work has been done by Mahlia et al., (Mahlia, Abdulmuin, Alamsyah, & Mukhlishien, 2003) to develop a steady-space dynamic model for a palm waste boiler. The solid fuel used was also fiber and shell from palm oil processing. However, in Mahlia et al. study, moisture content and calorific value of fuel, and air-fuel ratio are assumed to be constant, while temperature of the boiler is assumed to be proportional to fuel rate. This is not the case apply for the current simulation study because the ultimate aim in this simulation work is to study the effect of moisture content, ash content, air flow rate and temperature on combustion process and flue gas emission. Bignal et al. (Bignal, Langridge, & Zhou, 2008) has investigated the effect of moisture content of fuel and boiler operating conditions on pollutant concentrations, and it is suggested that solid fuel should have low moisture content in order to reduce air pollutants. In another study, Yang et al. (Yang, Sharifi, & Swithenbank, 2004) has carried out mathematical simulations and experiments to study the effect of primary air flow rate and moisture level in the fuel on the combustion process of wood chips and the incineration of simulated municipal solid wastes. From Yang et al. (Yang, et al., 2004) study, it is found that volatile release and char burning has been intensified with increasing in the primary air flow rate until a critical point is reached, and also increase in moisture level in the fuel produces a higher flame front temperature at low air flow rates. However, in the current simulation, the effect of moisture level in the fuel on calorific value or mass enthalpy is highlighted, but not on the flame front

temperature. While for air flow rate, its effect on flue gas emission is studied on current simulation work, which can be compared with study from Yang et al. (Yang, et al., 2004). The typical values for proximate and ultimate analysis on the dry basis of shell and fiber biomass feedstock are shown in Table 1.

Theory

Heating value is an important indication for solid fuel, and it can be reported as high heating value (HHV) or low heating value (LHV). The difference between HHV and LHV is equal to the heat of vaporization of water formed by combustion of the fuel. The potential energy from fiber and shell can be obtained by Dulong formula based on ultimate analysis:

From the steam generated amount, the amount ofsolid fuel consumed in the boiler can be known. With a known palm oil milling capacity, the amount of steam needed to generate electricity can be calculated based on the following equation:Steam required = Energy required to process 1 ton of FFB × Milling capacity × Amount of steam required to produce 1kWh electricTheoretically, the potential energy conversion from fiber and shell for palm oil mill can be represented by the following formula:

EP =MfLHVf +MSLHVS (3) where Mf and Ms is the mass composition of fiber

and shell in total solid fuel.

Material and Method

An Overview to Simulation

The software used in the simulation was ASPEN P LU S ( V e r s i on 7 . 1 ) . Th e l a y o u t o f s i mu l a t i o n method was shown in Figure 1, which was from reactor selection until gaining the final result of simulation.

Reactor Selection

The unit operation selected in ASPEN PLUS to run the simulation was shown in Figure 2. Reactor B1 was RYield which represented Reactor Yield, and it was used as non stoichiometric reactor based on known yield distribution. Reactor B2 was RGibbs which represented Reactor Gibbs was used for rigorous reaction and multiphase equilibrium based on Gibbs free energy minimization. As a result of the reaction occurred between the fuel and air, flue gas that composed of carbon, hydrogen, oxygen, carbon monoxide, carbon dioxide, water vapor, nitrogen, sulfur, nitrogen dioxide, nitrogen trioxide, and sulfur dioxide were produced. Unit operation B3 was the two streams heat exchanger which modeled co- current or counter current shell and tube heat exchanger. Flue gas with temperature at 800°C and 20.27bar was exchanged heat with water at 70°C to produce steam with temperature 260.69°C. This high pressure and temperature steam was the final product to generate electricity or for downstream processes usage.

Data Specification and Aspen Simulation

In the simulation work, data and specifications of input material such as mass flow rate, temperature, pressure, and component composition were entered. During the simulation, several conditions were studied by varying air flow rate, moisture content of fiber and also moisture content of shell, ash content of fiber and shell, and also temperature of combustion

reaction. Thus, data specifications and properties for these conditions were explained separately. Some assumptions were made in the simulation which may affect the accuracy of the final result. One of them was assuming coal properties for fiber and shell density and enthalpy because both were nonconventional material that the composition of the material was unknown. Coal would be the nearest similar material to both fiber and shell in the property method and models available in ASPEN PLUS as they were all from plant origin and with wood characteristic. In addition, throughout the simulation, the mill capacity was assumed to be 30 ton/hour. Moreover, as fiber and shell were nonconventional material, there were defined by the percentage value of ultimate and proximate analysis from literature in ASPEN PLUS. However, the sulfur analysis for both fiber and shell were assumed to have zero value due to its uncommonness and unavailability of data.

Air Flow Rate as Variable

When analyzing the effect of air flow rate on the flue gas composition produced, sensitivity function in the model analysis tool was activated by varying air flow rate with pre-set range of limit. Thus, a set of result within the pre-set limit range of air flow rate were generated and could be analyzed by plotting graph. Table 2 showed the input data for the condition when air flow rate was set as the variable.

Moisture Content of Fiber/Shell as Variable

Simulation was done to study the effect of varying moisture content of only shell, only fiber and both on the mass enthalpy produced in the process. When only moisture content of shell was considered, the stream for fiber was removed and vice versa for fiber case. Air flow rate in this case was held constant at 57,000 kg/hr. Sensitivity function was applied to flue gas stream in order to obtain the optimum moisture content value for both fiber and shell. Moisture content of fiber and shell was varied by changing the value in the proximate analysis. Suggested moisture content for simulation for fiber was 6 wt%, 19.5 wt%, and 33 wt% while for shell was 5 wt%, 9 wt% and 13 wt%. The result obtained from si mulation was analyzed by plotting graph of mass enthalpy against moisture content to observe the trend.

Ash Content of Fiber and Shell as Variable

The ash content of fiber and shell were varied to study its effect on gross calorific value or heating value of the fuel. The ash content value was varied in the ultimate analysis of both fiber and shell to change the composition of the raw material input while other parameters

held constant. Ash content effect on heating value was studied separately for fiber and shell, but not in combination of both fiber and shell condition. Input data was tabulated as shown in Table 3.

Temperature of Combustion Process as Variable

Temperature of the combustion process was altered in reactor B2 shown in Figure 5.2 to study its effect on the gas emission. A range of temperature from 600°C to 1200°C was chosen as conventional combustion process produced flue gas at temperature 800°C. Thus, simulation result of flue gas component flow rate with varying temperature at 600°C, 800°C, 1000°C, and 1200°C was obtained. A graph about mass flow rate of flue gas emission against the temperature of the combustion process was plotted, and the trend was analyzed in result and discussion section. During the simulation of studying temperature effect, other parameters were held constant. Input data for simulation run with different temperature value was tabulated in Table 4. It is noted that other input data entered in the proximate and ultimate analysis for fiber and shell, as well as the properties of water were assumed to be constant as listed in Table 1.

Results and Discussion

Summary of Simulation Results

For a palm oil mill with the capacity of 30 tons, FFB/hr, the minimum steam required for palm oil process, and electricity was 18 ton/hr. In this report; three types of solid fuel feed are used in the simulation, which is a mixture of 70% fiber and 30% shell, 100% fiber, and 100% shell. Besides, the air flow rate is first assumed to be 57000 kg/hr and the amount of air flow rate will be verified in the section 4.2. From the calculation, the input of solid fuel for

each set of simulation is shown as Table 5.

From the calculation, it shows that solid fuel consumed in simulation set 2 is higher than set 1 and set 3. This is because the fiber has lower heating value/calorific value compared to the shell, and this can be proven from the simulation results.

Figure 3 shows the calorific value of the solid fuel determined by computer simulation. From the results obtained, it is obviously shown that shell has the highest calori fic value, and the fiber has lowest heating value. Meanwhile, the simulation set 1, which comprised of solid fuel mixture, has moderate heating value. The high calorific value of shell can be due to the low moisture content. Besides, there are many similar researches, which were proven that moisture content in the shell is usually much lower than fiber and eventually has higher calorific value (Li, Yin, Zhang, Liu, & Yan, 2009; Olufayo, 1989; Werther, Saenger, Hartge,

Ogada, & Siagi, 2000). Experimentally, the calorific value for palm waste is in the range of 18MJ/hr to 20MJ/hr (Yusoff, 2004). The results obtained from the simulation are obviously higher than the experimental value. This may due to the limitation of software Aspen Plus used in this report. In the nutshell, it is suggested that the shell which has the highest calorific value is the most suitable solid fuel for combustion. However, the amount of shell produced for a typical palm oil mill is 60kg/hr for every 1ton FFB/hr. This amount is not enough to generate steam for whole plant. Thus, it is suggested that solid fuel with the fiber-shell mixture is the better solution. The composition of flue gas is analyzed in this report with the aid of computer simulation. Figure 4 shows the major gas constituent in the flue gas for different type of solid fuel. From the results obtained, the CO2 is the major component in the flue gas and followed by SO2, NOX. Meanwhile, the simulation results show that there is no CO gas emission. This is due to the complete combustion of solid fuel in the excess air, and all the carbon are converted to CO2. Besides, the results show that the amount of NO3 is extremely small and can be negligible and the NO is the majority of NOX. According to the research, the NOX are predominantly NO and NO2, in which NO is the 95% oftotalNOX (Ganapathy,2003).

When compare the gas emission among the solid fuel, solid fuel with fiber solely has the highest CO2 and SO2 emission. On the other hand, the solid fuel with shell solely has the highest NOX emission. Meanwhile, the solid fuel consists of fiber-shell mixture has moderate emission for all gases. This obviously showed that shell which has the highest calorific value is not suitable to become the only solid fuel in the combustion. In order to reduce the emission, the solid fuel should comprise of both fiber and shell. A table that summarized the results of gas emission for all types of solid fuel showed as below.

Effect of Air Flow Rate to the Composition of Flue Gas

In order to avoid emission of CO, the solid fuel must be combusted in the condition of excess air. In this report, the air flow rate is varying in the range from 20000 kg/hr to 40000 kg/hr to determine the minimum air required and investigate the effect of air flow rate to the emission. Theoretically, the minimum air required for complete combustion can be determined by observing the emission of CO and CO2. Figure 5 shows the amount of CO and CO2 released when burning of fiber-shell mixture at different air flow rate. From the results obtained, the amount of CO released is decreasing with the increase of air flow rate while the amount of CO2 is directly proportional to the air flow rate. The present of CO at the front part of the graph shows the insufficiency of oxygen in reaction. However, both graphs are become flat when the air flow rate increases to 24827 kg/hr. This is because complete combustion has occurred at the air flow rate of 24827 kg/hr. At this point, all carbons in the solid fuel are converted into CO2 and further increment in air flow rate will not increase the amount of CO2 released. Hence, the air flow rate required for fiber-shell solid fuel with the mill capacity 30 tons FFB/hr is 24827kg/hr. The effect of combustion in the excess air will be analyzed by observing the emission of SO2 and NOX.

From the results showed in Figure 6, the emission of SO2 has the similar trend with CO2. The amount of SO2 is increasing sharply until the point of minimum air flow rate, and it becomes constant under the condition of complete combustion. For the emission of NOX, the trend was slightly different with the gases previously and the amount of NOX released is relatively small, especially the emission of NO3. According to the results, the NOX starts to release when there is sufficient for oxygen present, and its amount is increasing in the condition of excess air.

The trend of NOX emission can be explained through the theory of NOX formation in literature. According to Ganapathy, (Ganapathy, 2003) NOX are produced during the combustion of solid fuel through the oxidation of atmospheric nitrogen and fuel-bound nitrogen. These sources produce three kinds of NOX, which is fuel NOX, prompt NOX, and thermal NOX. The fuel NOX is generated when nitrogen in fuel combines with oxygen in combustion. This type of NOX is insensitive to flame temperature but is influenced by oxygen. Thus, when there is sufficient oxygen present, the nitrogen bound in solid fuel will turn into NOX and the amount of fuel NOX eventually becomes constant at certain air flow rate. On the other hand, the thermal NOX produced will not become constant when the air flow rate increasing. This is because the thermal NOX is produced when atmospheric nitrogen combines with oxygen under intensive heat. When there is a rise in air flow rate, the amount of NOX generated will be increased exponentially. Hence, the mass flow rate of NO, NO2, and NO3 in the figure 7 increase exponentially when the air flow rate is increasing.According to the industry practice, the 30% excess air is applied to the solid fuel combustion. Thus, when 30% excess air is applied to the simulation, the composition of each gas is shown as table 7.

Comparing the reading in Table 7 and 8, the amount of CO2 and SO2 released are same. However, the amount of NOX released in 30% excess air is significantly reduced. In short, the solid fuel which consists of solid only has the best performance in 30% excess air and the solid fuel which comprises fiber-shell mixture has moderate gases emission. Hence, it is suggested that the solid fuel in palm oil mill should be comprised of fiber-shell mixture or shell only.

Effect of Moisture Content to the Heating Value and Heat Duty of Boiler

The proxi mate analysis of fi ber and shell is significantly different. The fiber has the typical moisture content in the range of 6% to 33% while it is 5% to 13% for the shell (Li, et al., 2009). As a result, the effect of moisture content to the heat of combustion is investigated in this report. Table 9 shows the summarized of each trial in the simulation.

From results in Figure 8, the mass enthalpy is directly proportional to the moisture content. As mention previously, the present of moisture content in solid fuel eventually decreases the heating value. The computer simulation failed to perform the effect of moisture content to heating value can be due to the limitation of Aspen Plus. The computer simulation is suggested alter other properties on solid fuel in order to get the constant heating value for the solid fuel. However, the effect of moisture content to low combustion efficient can be observed from the Figure 9 which shows the relationship between the moisture content and heat duty of reactor.

From the graph plotted, the heat duty of reactor is directly proportional to the moisture content. When the moisture content is high, the reactor has to work more to provide sufficient energy for combustion. Again, the results show that solid fuel with fiber only is not environmental and economically friendly. The solid fuel used in the boiler should be shell or fiber- shell mixture. Figure 10 shows the heat duty of reactor by using fiber-shell mixture as solid fuel.

From the graph plotted in Figure 10, it shows that the effect of fiber moisture is more significant compared to the shell. This may be due to the high moisture content of fiber which required extra energy to vaporized water in the solid fuel. In short, the fiber- shell mixtures are better than fiber only. Therefore, the fiber-shell solid fuel, which has the moisture content similar with trial 4 to 6 is suggested. The results may slightly deviate from the experimental value due to the assumption Aspen Plus.

Effect of Ash Content to the Heating Value

Ash content is referring to the mass fraction of incombustible material in solid fuel. It was found that the ash is the heat sink in the same way as moisture, lowering combustion efficiency (Ciolkosz, 2010). Besides, the similar literature showed that systems that are designed to combust wood can be overwhelmed by the volume of ash if other biofuels are used, which can reduce the combustion efficiency or clog the ash handling mechanisms (Pritchard, 2002). Thus, the relationship between heating value of solid fuel and ash content is the study to predict the effect to combustion efficiency. Table 10 shows the summarized of each trial in the simulation.

From the results present in Figure 11, the heating value of solid fuel is inversely proportional to the ash content. When there is high ash in the solid fuel, the heat released from the combustion will decrease. This is because the heat of combustion for a given fuel is mostly a function of the fuel’s chemical composition.

Thus, more incombustible ash contained in a specific amount of solid fuel is eventually decreasing the heat of combustion and yield low combustion efficiency. In terms of heat duty, the heat duty for the furnace will be lower at high ash content (Figure 12). This is because less work is done by the furnace to combust the solid fuel which has the high volume of incombustible material.

Effect of Temperature to Gas Emission

As discussed in section 4.2, the NOX emission will not be constant with the increasing of air flow rate due to the thermal NO . However, GanapathyX(Ganapathy, 2003) proposed that the formation ofthermal NOX increase exponentially with the increase in operating temperature because it is a function of flame temperature. For the natural gas in 15% excess air, it was found that each 37.78oC increase in combustion temperature will increase the flame temperature by 18.33oC. So, the operating temperature is the study in this section to investigate the effect of temperature to NOX and determine the optimal operating temperature. The temperature in the furnace is varying from 600oC to 1200oC and the NOX released is shown in Figure 13

and 14.

From the results in Figure 13 and 14, the amount of NOX released is exponentially increased to the increase of operating temperature. The emission is increasing sharply at higher temperature (1000oC). According to the information provided by Ciolkosz (Ciolkosz, 2010), the operating temperature of furnace should not over 972oC in order to regulate to NOX emission. Therefore, the assumption of 800oC operating temperature in this report is acceptable.CONCLUSIONFrom the results obtained from computer simulation, the mixture of fiber-shell solid fuel is better compared to the solid fuel with fiber only. The shell which has the highest heating value and lower emission cannot be used as the only fuel in combustion because the typical amount of shell produced is unable to sustain whole mill. Moreover, the air flow rate should be controlled and the emission at 30% excess air is acceptable. Furthermore, the moisture content is determined lower the combustion efficiency, and the suggested moisture content of fiber and shell is 6%-19.5% and 5%-13%, respecti vel y. Last but not least, the operating temperature should not exceed 972oC as it will promote the NOX emission.Due to the limitation of Aspen Plus, the effect of moisture content on heating value and the tolerable ash content is unable to identify. However, it is suggested that heating value is inversely proportional to moisture content and the ash content. In order to improve the

findings in the future, other simulation software, which is purposely designed for combustion or design of boiler, can replace the Aspen Plus in order to get an accurate result. For future research purpose, ASPEN PLUS is suggested to be replaced by other computer simulation software which is design for boiler.

References

Bignal, K. L., Langridge, S., & Zhou, J. L. (2008). Release of polycyclic aromatic hydrocarbons, carbon monoxide and particulate matter from biomass combustion in a wood-fired boiler under varying boiler conditions. Atmospheric Environment, 42(39), 8863-8871.

Ciolkosz, D. (2010). Characteristic of Biomass as Heating Fuel. In P. s. College of Agriculture Science (Ed.), Renewable and Alternative Energy Fact Sheet (pp. 4). Pennsylvania: Agricultural Research and Cooperative Extension.Foo, K. Y., & Hameed, B. H. (2010). Insight into the applications of palm oil mill effluent: A renewable utilization of the industrial agricultural waste. Renewable and Sustainable Energy Reviews, 14(5), 1445-1452.Ganapathy, V. (2003). Industrial boilers and heat recovery steam generators: design, applications, and calculations: Marcel Dekker.Li, J., Yin, Y., Zhang, X., Liu, J., & Yan, R. (2009). Hydrogen-rich gas production by steam gasification of palm oil wastes over supported tri-metallic catalyst. International Journal of Hydrogen Energy, 34(22), 9108-9115.Mahlia, T. M. I., Abdulmuin, M. Z., Alamsyah, T. M. I., & Mukhlishien, D. (2003). Dynamic modeling and simulation of a palm wastes boiler. Renewable Energy, 28(8), 1235-1256.Olufayo, A. A. (1989). Combustion-related characteristics of some by-products of agricultural processing industries. Biomass, 18(1), 69-72.Pritchard, D. (2002). Biomass Combustion Gas Turbine CHP: Talboltt’s Heating Ltd.Vassilev, S. V., Baxter, D., Andersen, L. K., & Vassileva, C. G. (2010). An overview of the chemical composition of biomass. Fuel, 89(5), 913-933.Werther, J., Saenger, M., Hartge, E. U., Ogada, T., & Siagi, Z. (2000). Combustion of agricultural residues. Progress in Energy and Combustion Science, 26(1), 1-27.Yang, Y. B., Sharifi, V. N., & Swithenbank, J. (2004). Effect of air flow rate and fuel moisture on the burning behaviours of biomass and simulated municipal solid wastes in packed beds. Fuel, 83(11– 12), 1553-1562.Yusoff, S. (2004). Renewable Energy from Palm Oil: Innovation on Effective Utilization of Waste. International Journal of Cleaner Production, 14(1), 87-93.