Investigation of Process Parameters Influence on Municipal Solid Waste Gasification with CO2 Capture via Process Simulation Approach

Integration of gasification with CO2 capture using CaO sorbent is proposed as an alternative treatment to convert municipal solid waste (MSW) into energy. Aspen Plus process simulator was employed to study the process. Two models were built to represent the non-sorbent and the sorbent-enabled MSW gasification. The model validation against available experimental data shows high accuracy of the simulation result. The effect of CO2 capture using CaO sorbent on the syngas composition and lower heating value (LHV) was observed by comparing the two models, and sensitivity analysis was performed on both models. Several process parameters affecting the syngas composition and LHV were investigated, including CaO/MSW ratio, temperature, equivalence ratio, and steam/MSW ratio. The addition of CaO sorbent for CO2 capture was found to successfully reduce the CO2 content in the syngas, increase the H2 composition, and improve the syngas LHV at the temperature below 750 oC. The maximum H2 composition of 56.67% was obtained from the sorbentenabled gasification. It was found that increasing equivalence ratio leads to a higher H2 concentration and syngas LHV. Raising steam/MSW ratio also increases the H2 production, but also reduces the LHV of the syngas. Observation of the temperature effect found the highest H2 production at 650 oC for both non-sorbent and sorbent-enabled gasification.


Introduction
As the fourth most populated country in the world, Indonesia produces a vast amount of municipal solid waste (MSW). In 2019, around 176,000 t of MSW was produced daily, making up a total of 64 Mt annual MSW generation. Due to the lack of infrastructure and human resources, waste processing is not yet available on large scale. Around 70% of the generated MSW end up in open landfills and the rest are burnt, buried, left unmanaged, or even dumped into the country's rivers (Khalil et al. 2019). A lot of environmental problems have arisen due to this poor waste management, including land pollution, water pollution, and emission of greenhouse gases (Korai et al. 2016). MSW in the landfills is the third largest contributor of methane emission, accounting for around 550 Tg of global methane emission per year (Zuberi and Ali 2015). This number raises serious concerns since methane is a powerful greenhouse gas with 21-23 times higher global warming potential than carbon dioxide (Eggleston et al. 2006).
On the other hand, the rapid growth of population and economy in Indonesia also leads to energy security problems. Indonesia's energy consumption has risen significantly from 893.76 million barrel of oil equivalent (BOE) in 2009 to 936.33 million BOE in 2018 (Sutijastoto et al. 2010, Adi et al. 2019 and is predicted to grow continuously in the upcoming years. Since 2004, Indonesia's local oil production has been unable to meet this high fossil fuel demand, forcing the country to meet its energy requirement by importing oil. The majority of Indonesia's energy share comes from fossil fuel-based energy while renewable energy sources contribute to less than 5% of the energy mix (Khalil et al. 2019). Indonesia Presidential Decree No. 79 of 2014 has stated a policy for national energy mix, requiring 23% total contribution of new and renewable fuels as the main energy source by 2025 (Putro et al. 2020). Rigorous and strategic effort for renewable energy development is urgently required to achieve this target.
Waste-to-energy (WTE) conversion is a promising solution to solve both the overproduced waste and energy security problems at once. MSW can be regarded as a potential source of energy due to its high calorific value (Zhao and Wang 2018), with HHV of around 19-20 MJ/kg depending on the composition (Jingxia 2018). A range of processing technology is available to convert MSW into energy, including several means of thermochemical and biochemical conversion (Mishra andMohanty 2018, Chen andWang 2016). Although biochemical conversion technology generally offers lower treatment cost, it is not preferred for large volume of waste due to the difficulty to control the growth of the bacteria (Sudibyo et al. 2017). Therefore, the option narrows down to thermochemical conversion technologies, including incineration, pyrolysis, and gasification.
Incineration is basically an oxidation reaction of a combustible material, during which energy is recovered in the form of heat (Autret et al. 2007). Although the technology is widely implemented around the world, it has a major drawback of producing hazardous by-products such as dioxins and furans, especially for feedstocks containing chlorine (Al-Salem et al. 2010). In addition, incineration can only produce heat and electricity without recovery of any value-added chemicals. In contrast, gasification produces valuable chemicals and fuels while also recovers energy in the form of heat and electricity (Consonni and Viganò 2012). Gasification also has other advantages regarding its cleaner emission and higher energy recovery efficiency (Dong et al. 2018), which makes it a more environmentally friendly option (Luo et al. 2012).
Gasification is a thermochemical conversion process where biomass is converted into gaseous products in the presence of a gasification agent (Molino et al. 2018). The process comprises several steps including feedstock drying, pyrolysis or decomposition, combustion, and gasification. This technology enhances the calorific value of the feedstock by reducing the C/H mass ratio of the biomass. Some of the commonly used gasification agents are air, steam, oxygen, and carbon dioxide (Sikarwar et al. 2016). The selection of the gasification agent determines the calorific value of the product. Other parameters affecting the biomass gasification process and product include the gasification temperature, type of feedstock, amount of the gasifying agent, and the presence of catalyst and sorbent (Parthasarathy and Narayanan 2014).
The main product of biomass gasification is syngas, which is a mixture of gases containing H2, CO, CO2, and CH4. The syngas can be utilized to generate electricity by combustion in gas turbines or by using fuel cell (Toonssen et al. 2011). Alternatively, it can be converted through Fischer-Tropsch synthesis to produce liquid transportation fuels (Dos Santos and Alencar 2020). Syngas is also utilized as basic raw material for production of chemicals such as methanol and a wide range of other products (Doranehgard et al. 2017). In addition, syngas contains hydrogen, which is often considered as its most important constituent (Moghadam et al. 2014).
Hydrogen is well known as a clean fuel with high energy density. It can be utilized for production of electricity or vehicles fuel without generating toxic emissions (Hosseini and Wahid 2016). Compared to conventional hydrocarbon fuels, hydrogen energy density is 2.75 times higher at around 122 kJ/g (Kapdan and Kargi 2006). Currently, the main pathway for commercial hydrogen production uses fossil fuels as the feedstock, which makes it a carbon-intensive process. Alternative routes for hydrogen production include water electrolysis, biological methods, and nuclear production (Doranehgard et al. 2017). However, hydrogen production from biomass gasification is highly regarded as the more promising alternative due to safety, economic, and environmental factors, therefore it has been extensively studied in the literature (Salkuyeh et al. 2018;Shayan et al. 2018;Peng et al. 2017).
According to Doranehgard et al. (2017), the production of CO2 emission remains one of the main barriers in the development of biomass gasification. The CO2 emission also affects hydrogen production by lowering its concentration in the syngas. However, the syngas hydrogen concentration can be enriched using CO2 capture technology which removes some of the CO2 from the mixture, therefore producing hydrogen-enriched syngas. Among several technological options, CO2 removal using CaO sorbent is currently gaining attention due to its relatively low cost and high effectivity (Manovic and Anthony 2010). This technology captures CO2 from gaseous mixture by CaO carbonation reaction where CaO reacts with CO2 to form CaCO3.
Several experimental works have been conducted on CO2 capture using CaO sorbent for biomass gasification. The previous studies used different biomass feedstocks such as corn stalk (Li et al. 2017), sugarcane leaves (Bunma and Kuchonthara 2018), palm kernel shell (Shahbaz et al. 2017), sewage sludge , and saw dust (Acharya et al. 2010). The results of all the previous studies agree that CaO sorbent works effectively to remove CO2 and enhance hydrogen concentration in the syngas. However, only a limited number of works has been reported on MSW gasification with CO2 capture using CaO sorbent (Hu et al. 2015).
A systematic understanding of the gasification characteristic is crucial in the process design and development. With the lack of experimental data, simulation-based study can be a feasible approach to provide additional insights regarding the process. Simulation modelling-based research has an advantage of reduced time, cost, and resources compared to experimental work, while still providing high accuracy results (Rupesh et al. 2016). Several simulation studies on biomass gasification with CO2 capture using CaO sorbent have been published. The reported studies mostly utilized Aspen Plus process simulation, which successfully simulated the process and generated results with high validity compared to available experimental data (Rupesh et al. 2016, Shahbaz et al. 2017, Zhou et al. 2019, Gao et al. 2018. However, none of the previous simulation studies used MSW as the feedstock. Aspen Plus is an extensive process modelling software which is capable to simulate complex industrial processes and generate accurate results. The software is equipped with unit operation blocks with built-in mathematical models. Aspen Plus contains a wide database of various chemical compounds and their thermodynamic, physical, and chemical properties. User can also select the suitable thermodynamic model according to the nature of the simulated process. In addition, Aspen Plus is equipped with powerful analysis tools such as sensitivity analysis and optimization (Begum et al. 2014).
In this study, process simulation is employed to understand the effect of CaO sorption on syngas production via MSW gasification. The simulation model is built in Aspen Plus software and validated against experimental data to ensure the model's accuracy. The composition and calorific value of the produced syngas from non-sorbent and sorbent-enabled gasification are compared. Furthermore, several parameters affecting the syngas composition and calorific value were also investigated, including CaO/MSW ratio, temperature, equivalence ratio, and steam/MSW ratio. The understanding of how these key parameters affect gasification is a substantial key in developing MSW gasification process integrated with CaO sorbent for CO2 capture.

Methods
The gasification of MSW with sorbent-enabled CO2 capture was modelled using Aspen Plus v8.6 process simulator. The simulation was built with several assumptions (Rupesh et al. 2016, Shahbaz et al. 2017 a. The process is in steady-state condition. b. The gasifier and CaO carbonation reactors are isothermal and operate at atmospheric pressure. c. Tar and higher hydrocarbons production is neglected. d. The catalytic activity and capacity reduction of CaO are neglected. e. Char is considered as graphite carbon. f. Sulfur is only converted to H2S and nitrogen is only converted to NH3. Peng-Robinson equation of state with Boston-Mathias alpha function was chosen for this simulation as this property package is suitable for high temperature processes, including gasification (Ramzan et al. 2011). The property package estimated thermophysical properties of the conventional components such as H2, CO, CO2, CH4, H2O, O2, N2, NH3, H2S, and other gases included in the simulation.   Biomass and ash were classified as non-conventional components, thus the enthalpy model HCOALGEN and density model DCOALIGT were selected for their thermophysical properties calculation. CaO, CaCO3, and C were described as solid components with available thermophysical properties stored in Aspen Plus data.
The sorbent-enabled gasification process block diagram is presented on Figure 1. Since gasifier is not included in the Aspen Plus default unit operations, the actual gasification process was divided into several stages in the simulation, each one represented by a unit operation available in Aspen Plus. The gasification stages comprise drying, decomposition, and gasification. The feedstock's proximate and ultimate analysis are listed on Table 1 and 2, while the main equipment used in the simulation are detailed on Table 3.
The feedstock enters the process through stream MSW into DRIER, where early heating and drying process take place to reduce the moisture content of the MSW. Since MSW is specified as a single non-conventional component, the drying process is modelled using RStoic reactor with a calculator block which specifies the percentage of water removal. The dried MSW enters the decomposition stage where it is decomposed into C, H, O, N, and S elements. This decomposition occurs in RYield reactor DECOMP, which is specified using FORTRAN statement. The decomposed elements are fed to the gasification reactor GASIF, modelled by RGibbs reactor in the Aspen Plus simulation. This type of reactor uses minimization of Gibbs free energy to calculate syngas composition, assuming complete chemical equilibrium. The syngas produced from gasifier is fed to CO2 capture reactor CO2CAPT, also modelled using RGibbs reactor. In this reactor, removal of CO2 through CaO carbonation into CaCO3 takes place, removing a portion of CO2 from the syngas mixture. The reactions inside the gasifier and CO2 capture reactor are displayed in Table 4.
The simulated model was validated by comparing the syngas composition obtained from the simulation with that of experimental data published by Mahishi and Goswami (Mahishi and Goswami 2007). The deviation of the model from the experimental result is calculated using root mean square error (RMSE) as expressed on equation 1. An accurate model is indicated by a small RMSE value. Where: • LHVgas is the lower heating value of syngas (MJ/Nm 3 ) • Y F 4 is the mole fraction of H2 • Y JK is the mole fraction of CO • Y JF O is the mole fraction of CH4 The validated model was used to investigate the effect of sorbent-enabled CO2 capture on syngas composition and lower heating value (LHV) at different operating conditions. The LHV of the syngas is calculated based on the composition of combustible gases including H2, CO, and CH4 (Sittisun et al. 2019). The calculation formula for LHV is shown in Equation (2) (Rupesh et al. 2016). The simulation was performed as two cases: the non-sorbent case which excludes the CO2 capture reactor system and the sorbent-enabled case. The syngas composition and LHV of the two cases were compared as the effect of parameters including CaO/MSW ratio, temperature, equivalence ratio, and steam/MSW ratio were studied. The base case was simulated at CaO/MSW ratio of unity, temperature of 600 o C, equivalence ratio of 0.5, and zero steam flow. Using sensitivity analysis feature in Aspen Plus, the parametric study was performed by varying CaO/MSW ratio from zero to unity, temperature from 500 to 1500 o C, equivalence ratio from 0.05 to 1, and steam/MSW ratio from 0.05 to unity. Each sensitivity analysis was run by changing only one parameter while keeping the other conditions at the base case value.

Model Validation
The simulated model (Figure 2) was validated against available experimental data of biomass gasification using CaO sorbent for CO2 capture (Mahishi and Goswami 2007). The composition of the syngas obtained from the simulation model was compared to the experimental data. The validation results are displayed on Figure 3 for nonsorbent gasification model and Figure 4 for sorbentenabled gasification model. The accuracy of the model results is quantified statistically by calculating the root mean square error (RMSE) as shown in Equation 1. Compared to previous publications (Rupesh et al. 2016;Al Amoodi et al. 2013), the result shows better agreement between simulation and experimental data, with acceptable RMSE of 4.07% for non-sorbent gasification and 5.63% for sorbent-enabled gasification. The sorbentenabled gasification model demonstrates higher RMSE compared to the non-sorbent case, as it includes more complex process model which increases the possibility of deviation. However, the validation result proved the ability of both models to predict the process output with a high accuracy. Therefore, the models can be used to further investigate the effect of process parameters on MSW gasification with CaO sorbent for CO2 capture.

Effect of CaO/MSW Ratio
One of the main purposes of this study is to observe the effect of CO2 capture using CaO sorbent on MSW gasification product. Therefore, the effect of CaO/MSW ratio on syngas composition was investigated. The addition of CaO sorbent is intended to remove CO2 content in the syngas and enhance H2 composition. Increasing CaO/MSW ratio means providing more sorbent for CO2 carbonation reaction (R9). Figure 5 demonstrates that CO2 removal performance is improved as the CaO/MSW ratio is varied from 0.05 to unity, indicated by the notable drop of CO2 percentage from 14.54% to 0.43%. As a result, H2 mole fraction increases from 20.00% to 23.31%, CO mole fraction from 9.41% to 10.97%, and CH4 mole fraction from 0.95% to 1.10%. Since the addition of CaO sorbent improved the concentration of H2, CO, and CH4, the lower heating value of the syngas increases accordingly as shown on Figure 6. The trends observed here agree with the results from previous study (Acharya et al. 2010).

Effect of Temperature
Gasification temperature is an important parameter affecting equilibrium and rate of the chemical reactions (R1-R9). The effect of temperature can be explained by basic law of chemical reactions: Lower temperature favours the exothermic reactions, and higher temperature favours the endothermic reactions. Meanwhile, the reaction rate gets higher as the temperature increases. Combination of these effects causes variation of syngas composition over temperature as observed in this simulation study. In this study, CaO sorbent is utilized to remove a portion of CO2 in the syngas through CaO carbonation reaction (R9), thus improving the syngas H2 composition. The optimal operating temperature of the carbonator is 580-700 o C, which is caused by the trade-off between the reaction equilibrium and kinetics. Above this temperature range, the efficiency of CO2 capture drops significantly and approaches zero at around 775 o C (Hanak et al. 2015). The addition of CaO sorbent can no longer affect syngas composition when the operating temperature is over this limit. The result of this study strongly agrees with theory. Comparison of the syngas composition for non-sorbent case (Figure 7) and sorbent-enabled case (Figure 8) suggested that CaO sorbent affects the syngas composition only at temperature range of 500-750 o C. This conclusion is also supported by Figure 9 which suggested that LHV of the syngas produced from non-sorbent and sorbent-enabled gasification only differs at the temperature below 750 o C. Observation of syngas composition and LHV at higher temperature shows no difference between the non-sorbent case and sorbent-enabled cases, indicating the absence of CO2 removal reaction. The effect of temperature on H2 composition is similar for non-sorbent and sorbent-enabled cases. At the temperature below 650 o C, both trends demonstrate that temperature raise increases H2 composition to a maximum value, which is observed at 650 o C for both cases. This is mainly attributed to the increasing rate of the water-gas (R5) and steam-methane reforming (R7, R8) reactions. In addition, at lower temperature the equilibrium of watergas shift reaction (R6) is favoured towards the right side to produce H2. However, above 650 o C, the temperature starts to favour the reverse direction of water-gas shift reaction (R6), causing H2 composition to decrease as the temperature increases. The maximum H2 composition for sorbent-enabled case is observed to be higher at 23.52% compared to 20.58% for non-sorbent case.
The production of CO is contributed mainly by the endothermic boudouard reaction (R4), water-gas reaction (R5), and steam-methane reforming reaction (R7). For both non-sorbent and sorbent-enabled cases, the graphs indicate a consistent increase of CO composition as temperature gets higher. This can be attributed to the increasing rate of the three reactions and the nature of endothermic reactions which is favoured by higher temperature. On the contrary, CH4 production for both cases consistently decreases as the temperature raises. This is attributed to the methane-consuming steammethane reforming reaction (R6, R7) which is also favoured by the higher temperature.
For non-sorbent case, the graph shows consistent reduction of CO2 composition as temperature increases. This can be explained by the endothermic boudouard reaction (R4) and exothermic water-gas shift reaction (R6). CO2 production from both reactions is more favourable at lower temperature, whereas higher temperature drives the reactions equilibrium to the other side which consumes CO2 instead. Composition of CO2 for the sorbent-enabled case is affected by gasification reactions (R1-R8) and CaO carbonation reaction (R9). The trend appears to be more complex as higher temperature inhibits CO2 production from gasification, but also promotes desorption of CaCO3 and releases additional CO2 to the syngas mixture. At 500-600 o C, temperature increase lowers CO2 composition down to its lowest point. This is the combined effect of the reduced CO2 production and the increasing CaO sorption activity at this temperature range. However, when the temperature is raised above 600 o C, the notable drop in CaO sorption activity causes CO2 composition to increase up to its maximum point at 750 o C. The CO2 composition around 600-750 o C is nonetheless still lower compared to nonsorbent case, indicating that CO2 removal reaction still takes place. The CO2 capture activity is no longer detected above 750 o C, as the CO2 composition obtained beyond this temperature is identical with that of the non-sorbent case. Figure 9 compares the lower heating value of syngas produced from non-sorbent gasification and sorbentenabled gasification. The syngas LHV for the non-sorbent case gets higher along with temperature raise, which is mainly contributed by the increase of CO content in the syngas. However, a different trend is observed for the sorbent-enabled case. Raising the temperature between 500-600 o C causes significant rise on H2 and CO content in the syngas, therefore a steep ascent of the lower heating value is observed. The lower heating value reaches its peak at 4.26 MJ/Nm 3 at the temperature of 600 o C. Between 600-750 o C, temperature rise affects the LHV negatively, as the H2 and CO composition are decreasing at this temperature range. Beyond 750 o C, no CaO sorption activity is detected as indicated by the identical value of syngas LHV from both models. It can be concluded from the graph that CaO sorption activity can significantly increase the LHV of the syngas.
Similar trends are observed in other works. Experimental study of Acharya et al. (2010) found that maximum H2 composition was obtained at 670 o C, while Hu et al. (2015) reported highest composition of H2 at 750 o C. Both studies also observed similar trends for CO and CH4 composition. In addition, a simulation study by Rupesh et al. (2016) generated similar trend result for H2, CO, CO2, and CH4 composition for both non-sorbent and sorbent-enabled cases.

Effect of Equivalence Ratio
The equivalence ratio (ER) represents the ratio of oxygen to biomass. A higher ER value means more oxidizing agent is present, promoting more oxidation reactions (R1-R3). This will lead to an increase of CO2 and H2O compositions, which will consequently reduce the amount of CO and H2 (Niu et al. 2013). The result on both Figure 10 for nonsorbent case and Figure 11 for sorbent-enabled case agree with this theory. On both cases, compositions of CO and H2 consistently decrease as ER and CO2 composition increase. The main difference between the two graphs is that the sorbent-enabled case has significantly lower amount of CO2, which is attributed to the CO2 capture. The reduced CO and H2 contents also affect the syngas lower heating value negatively. As seen on Figure 12, syngas LHV decreases significantly as the ER value is raised. It is also observed that the LHV of the syngas generated from the sorbent-enabled case is generally higher than that from the non-sorbent gasification, which is mainly attributed to the lower CO2 content and higher concentration of H2 and CO in the sorbent-enabled gasification. The trends obtained here are highly consistent with previously published results (Rupesh et al. 2016, Chen et al. 2013.

Effect of Steam/MSW Ratio
Addition of steam affects gasification by promoting watergas reaction (R5), water-gas shift reaction (R6), and steam-methane reforming reactions (R7, R8). Higher steam/MSW ratio drives these reactions equilibrium to the right side, increasing H2 and CO2 production while simultaneously reducing CO and CH4 composition. This is consistent with the result presented on Figure 13 for the non-sorbent case. Composition of CO2 rises from 14.12% to 19.37% and H2 mole fraction increases from 18.19% to 25.85% as the steam/MSW ratio is varied from zero to unity. Meanwhile, the mole fraction of CO decreases from 10.71% to 4.07%, and CH4 composition is also reduced from 1.18% to 0.15%. Similar result was obtained by Niu et al. (2013). However, for the sorbent-enabled case (Figure 14), the addition of CO2 capture system inhibits the raise of CO2 composition. Consequently, the highest CO2 composition is limited to 5.75% at steam/biomass ratio of unity, while the hydrogen composition is higher at 30.21%. This result agrees with the previous work published by Rupesh et al. (2016). The effect of steam addition on the syngas LHV is presented on Figure 15. For the observed range, it can be seen that syngas LHV from sorbent-enabled gasification is always higher than that from non-sorbent gasification. For both cases, the increase of steam/MSW ratio causes a decrease of the syngas LHV. This is attributed to the declining CO composition as previously discussed. Although H2 composition increases with steam addition, it cannot compensate for the reduced calorific value due to the decrease of CO production.

Conclusion
Gasification of municipal solid waste integrated with CO2 capture using CaO sorbent was successfully simulated using Aspen Plus process simulator. Two models were developed to compare the non-sorbent and the sorbentenabled gasification. Comparison of both model's results with available experimental data showed good agreement with acceptable root mean square error, indicating reliability of the model for predicting the actual process. The CaO addition was found to significantly reduce CO2 content and improve H2 concentration and LHV of the syngas. The sorbent activity is limited below the temperature of 750 o C, with optimum hydrogen concentration at 650 o C for both non-sorbent and sorbentenabled cases. The increase of equivalence ratio and steam/MSW ratio positively affect the hydrogen production. A maximum hydrogen composition of 56.67% was obtained using sorbent-enabled gasification. Overall, this study provides a significant insight regarding the influence of key parameters on the investigated process. However, the assumption of complete chemical equilibrium does not always apply to the actual process. Therefore, future work should focus on representing the actual condition more accurately by taking the reaction kinetics into consideration.