Gasification of Refuse-Derived Fuel (RDF)

Abstract In this work, the gasification of a fraction of municipal solid waste, MSW, generally separated from inorganic materials and biodegradable components, the so-called refuse-derived fuel (RDF), was studied using material characterisation methods, and the modelling of an industrial scale process was presented. The composition of RDF was determined by the separation of a representative sample into its basic components (paper, foils, hard plastics, textiles). All RDF components as well as a representative mixed sample of the RDF were studied using a thermogravimetric analysis (TGA), elemental analysis and bomb calorimetry to determine their proximate and elemental compositions, and a higher heating value. An industrial scale gasification process was studied by mathematical modelling and computer simulations. All techniques, gasification with air, gasification with oxygen, and gasification with both oxygen and steam were investigated under different conditions. The RDF conversion of 100 % was achieved by the gasification with air at the air to RDF mass ratio of 3.2. The gas heating value was 4.4 MJ/Nm3. The gasification of RDF using oxygen enables the production of gas with the heating value of around 10 MJ/Nm3 at the oxygen to RDF mass ratio of 0.65. By increasing the steam to the RDF mass ratio, the contents of H2 and CO2 increased, while the content of CO, reactor temperature and the gas heating value decreased.


INTRODUCTION
Many countries, including Slovakia, still landfill the majority of their municipal solid waste (MSW). Such landfilling results not only in significant environmental challenges, but also in wasting valuable resources. MSW represents a valuable source of material and energy. However, the heterogeneity and composition variations are important challenges for mechanical recycling of MSW. On the other hand, feedstock recycling methods such as pyrolysis and gasification have a big potential to be used for recovery of valuable chemicals and energy from MSW. The fraction of MSW generally separated from inorganic materials and biodegradable components is called refuse-derived fuel (RDF). The gasification of RDF results in the production of a combustible gas containing mainly H 2 , CO, CO 2 , and N 2 if air is used as the gasification agent. Some methane, light hydrocarbons, H 2 S, NH 3, and tar are also present in the gas [1].
A number of power generation plants based on gasification technology are in operation worldwide. A list of operating and planed gasification plants has been provided by Higman and Van der Burgt [2]. The modelling of gasification enables predicting optimal process conditions and reducing the number of experiments in the processes of design and operation. A simple equilibrium mathematical model based on the minimisation of Gibbs free energy was used by some authors to model the gasification of RDF. Materazzi et al. [3] used the equilibrium model to evaluate a two-stage waste gasification system. It results from their work that except for H 2 the predicted composition of gas is in good coherence with the experimental data. The equilibrium model was also applied by Barba et al. for RDF gasification [4], and Jarungthammachote and Dutta [5] for the gasification of charcoal and coconut shell. The authors found that the model data is in good agreement with the experimental one. However, Li et al. [6] applied an equilibrium model for coal gasification and declared that this model has some limitations in predicting gas composition. Generally, the equilibrium model provides good results at temperatures above 1000 °C, and it fails at the reactor temperatures below 800 °C [7][8].
In this work, the equilibrium model was combined with some empirical correlations based on laboratory experiments to predict the carbon conversion, reactor temperature, gas composition, gas heating value and the gas tar content under different types and amounts of gasification agents. The input data for the model was determined by the characterisation of single components of RDF such as paper, polyethylene foils, hard plastics and textiles.

RAW MATERIAL
The RDF samples used in this study were provided by a recycling company based in Slovakia. The composition of the RDF was estimated by hand separation of a representative sample of RDF into its basic material categories. The contents of individual components of RDF are shown in Table 1.
The sample of RDF with the composition shown in Table 1 was homogenised by grounding in a cryogenic mill. In addition, homogenised samples of individual material categories of RDF were prepared. Using simultaneous thermogravimetric (TG)/differential scanning calorimetric (DSC) measurements (Netzsch STA 409 PC Luxx, selb, Germany), a proximate analysis of individual components of RDF was done. The experimental conditions were met at the linear heating rate of 5 K·min -1 in the nitrogen flow of 60 ml·h -1 . The samples were heated from 293 K to 1073 K; at this temperature, they were maintained for around 30 min and then combusted by entering the oxygen to the system. Samples of individual components of RDF, of around 20 mg, were used in the TG/DSC measurements.
The dry basis elemental composition of individual components of RDF was estimated by a Vario Macro Cube ELEMENTAR elemental analyzer. A CHNS (Carbon, Hydrogen, Nitrogen, Sulphur) module with the combustion tube temperature of 1423 K, and the reduction tube temperature of 1073 K was used. The module did not enable the determination of the Cl content. The mass of the sample was around 1 g. The results of both proximate and elemental analyses are given in Table 2. and e are given in Table 3. Then the global material balance of RDF gasification can be written as: where, x 1 and x 2 are the numbers of moles of oxygen and steam used for each mole of the RDF and, x 3 , x 4 , x 5 , x 6 , x 7 , x 8 , x 9 , and x 10 are the numbers of moles of each corresponding species, obtained from each mole of RDF based on a single atom of carbon.
CH O N S is an empiric formula of tar that was estimated in the same way as that of the RDF using the results of the GC/MS analysis of biomass tar [9].
The following linear dependence between the tar content and the temperature based on biomass gasification experiments in [9] was assumed: where y represents the amount of tar in mg of tar per g of RDF and T is the temperature in K.
The mass balance of individual elements provides five algebraic equations. Together with the total mass balance, it makes six equations; however, to estimate all ten unknown stoichiometric coefficients, at least ten equations are required. The equilibrium constant of four gasification reactions should be added to the model. The following reactions were selected for this study: The relation between the equilibrium constant and the equilibrium composition of the products of individual reactions is given by: where K a is the equilibrium constant, P 0 the pressure at which K a is defined [Pa], P the system pressure where Δ r H is the reaction enthalpy at temperature T [J·mol -1 ], Δ r S the reaction entropy at temperature T In the gasification reactor, the reaction temperature is obtained by partial oxidation of the feed. The global enthalpy balance of the reactor is: Substituting the equations (10) and (9) into the equation (8)  The used heating values for products and components are shown in Table 4. The lower heating values of RDF components were measured experimentally [10] and the heat of combustion of the gas components was found in chemical engineering literature.

RESULTS AND DISCUSSION
The model described above was solved using MS EXCEL in the first step, then it was implemented into the ASPEN Plus simulation environment. The gasification with the following three different gasification agents was carried out:  gasification of RDF using air  gasification of RDF using pure oxygen  gasification of RDF using pure oxygen and steam The variable was the mass ratio of the gasification agent (air, oxygen or steam) and the observed output parameters were: conversion of RDF, reactor temperature, gas composition, and gas higher heating value (HHV).  (RDF) , the adiabatic reactor temperature is around 850 K and the conversion of RDF is only 40-50%. A 100 % conversion was achieved at m (Air) /m (RDF) =3.2 at the reactor temperature of around 950 K. The gas heating value was relatively low, 4.5 MJ·N -1 ·m -3 , because of a high nitrogen content, and it remained practically constant for m (Air) /m (RDF) ranging from 2 to 3.2, when it rapidly decreased because of the starting CO oxidation to CO 2 and a rapid increase of nitrogen content as a result of the decreasing contents of H 2 and CO. A low reactor temperature can cause a high content of tar in the produced gas. But increasing the reactor temperature requires the oxidation of H 2 and CO and also a rapid increase of nitrogen content in the produced gas which results in losing the energy content of the gas. Based on this observation, the optimal value of m (Air) /m (RDF) for the gasification of RDF studied in this work was 3.2.  The tar content of the produced gas decreased with the increasing reactor temperature and the air to RDF mass ratio. At optimum m (Air) /m (RDF) =3.2, the tar content was 280 mg/Nm 3 , which is not very high but still above the limit value of 50 mg/Nm 3 [2] for gas turbines.

Figure 3: Content of tar in the produced gas at different m (Air) /m (RDF) values using air as the gasification agent
The contents of H 2 S and NH 3 decreased with increasing m (Air) /m (RDF) (Figure 4); the content of H 2 S at m (Air) /m (RDF) =3.2 was 0.015 mol %. The combustion of such gas results in the SO 2 content in the flue gases of around 50 mg/Nm 3 , which is below the EU limit value for the combustion of solid fuels (200 mg/Nm 3 ).  The oxygen blown gasification of RDF enables the production of gas with a 2-2.5 fold higher heating value. However, the gasification plant requires an air separation unit. The complete conversion of RDF was achieved at m (O2) /m (RDF) =0.65; the adiabatic reactor temperature under these conditions was around 1100 K and the gas heating value was approx. 10 MJ/Nm 3 .

Figure 4: Content of H 2 S and NH 3 of the produced gas at different m (Air) /m (RDF) values using air as the gasification agent
The content of H 2 showed its maximum at m (O2) /m (RDF) =0.38 (50 mol.%), but the conversion at this point was only 50 %. The content of CO showed its maximum at m (O2) /m (RDF) =1.05, (55 mol.%), but the H 2 content at this point was 25 mol.% and that of CO 2 was 18 mol.%, while its minimum value at m (O2) /m (RDF) =0.75 was 10 mol.%. Based on these observations ( Figures 5 and 6), the optimal value of m (O2) /m (RDF) for the gasification of the RDF studied in this work was 0.67. The equilibrium model based calculated gas compositions published by other authors showed a similar trend. However, the experimentally measured data showed lower H 2 and higher CO 2 contents of the gas. These differences are significant particularly at temperatures below 800 °C. The addition of steam to the gasification reactor enables taking care of the water-gas shift reaction and the steam gasification reaction of carbon. As a result of these reactions, a higher content of H 2 and CO 2 and a lower content of CO in the produced gas were obtained. At constant m (O2) /m (RDF) =0.67, the gas composition, reactor temperature and the gas heating value at different m (steam) /m (RDF) values were determined. Preheated steam with the temperature of 600 K and the pressure of 12 bar was considered. It was found that by increasing the mass ratio of steam to RDF from zero to 0.55, the content of H 2 increased from 44 mol.% to 51 mol.% and that of CO 2 increased from 10 mol. % to 22 mol.%. However, the content of CO decreased from 41 mol.% to 23 mol.%, the reactor temperature decreased by 40 K and the gas heating value decreased from 10.1 MJ/Nm 3 to 8.7 MJ/Nm 3 . The effect of steam addition to a gasification reactor at m (O2) /m (RDF) =0.67 is shown in Figure 7.

CONCLUSION
The 100% RDF conversion was achieved in the gasification with air at m (air) /m (RDF) = 3.2. However, the gas heating value was only 4.4 MJ/Nm 3 , and the reactor temperature was 950 K. The produced gas contained 19 mol.% of H 2 , 16.5 mol.% of CO, 7.6 mol.% of CO 2 , N 2 , and a minor amount of other components.
The contents of H 2 S and NH 3 did not exceed the EU limits for SO 2 and NOx caused by the combustion of these components.
The gasification of RDF using oxygen enables the production of gas with the heating value of around 10 MJ/Nm 3 at m (O2)/ m (RDF) =0.67. Under these conditions, the content of H 2 was 44 mol·%, that of CO was 41 mol·%, and that of CO 2 was around 10 mol.%; the content of other components including methane was 5 mol.%.
By increasing the m (steam) /m (RDF) ratio, the contents of H 2 and CO 2 increased. However, the content of CO, the reactor temperature and the gas heating value decreased.