2.1. A New Process Design
This work proposes a first-of-its-kind integrated process that allows an efficient use of these two technologies through a combined reactor that is able to maximize the performance of both SCWG and SCWO. The SCW-GcO flowsheet is illustrated in
Figure 1, where the main input and output streams are reported.
In principle, the primary feedstock could be any organic matter in solid or liquid state. In the case of municipal solid waste (MSW) the pre-treatment could be a pyrolysis unit that transforms the waste into an oily stream and a carbonaceous stream. Since for a given gasifier the admissible range of feedstock properties is narrow [
19], feedstock at the gasifier could be preferably an organic waste in liquid state such as a mixture of solvents, chemicals, and oil from the chemical and processing industries. On the contrary, oxidation can support much more change to feedstock composition and can also accept solids suspended in water.
Waste in liquid state mixed with water (H-OIL-IN) is first sent to the gasification chamber of the combined reactor, where it is partially converted to a gasification reactor effluent (GAS-OUT). The conversion is almost completed in a catalytic post-gasification reactor; after a cooling section (HEX4–5) and gas–liquid separation (DEMIST2), a liquid water and oil mixture (17 L; “oil” being partially gasified organic matter whose amount depends on the gasification yield) and a hydrogen-rich syngas stream (17G) are obtained. The latter is first treated in an H2S trap and then in a methanation reactor (METHANAT) to improve CH4 yield. A second flash separation stage (DEMIST3) allows acid water (30 L) to be removed from the syngas, which is then treated in a membrane separator (STAGE1–2) to remove CO2. The liquid stream 17 L is heated in a high-pressure heat exchanger (HEX4, AUXHEX5) and is continuously fed to the oxidation chamber (WTR-RCY) together with compressed air and a secondary charge of waste (3; solid or liquid organic feedstock, or both) that has the role of producing heat. Indeed, almost all organic carbon is completely converted to CO2 (conversion yield >99.9%), producing the heat necessary to sustain gasification. The output from the oxidation reactor is cooled (HEX1–2) and separated (DEMIST1) to remove carbon dioxide, nitrogen, and other gas (23) from the liquid acid water (11L). Two cooling systems (AIRCOOL1–2) connected to heat exchangers allow heat to recover from processing the hot stream and finally are used to generate electricity in a thermoelectric generator (TEG1–2–3).
This process arrangement allows two drawbacks of gasification to be overcome:
Since supercritical water gasification does not reach 100% efficiency (typical efficiency is between 60 and 90% depending on the feedstock), the liquid residue of gasification that is a harmful waste can be destroyed in the integrated oxidation section.
The SCWO generates the heat that is necessary to sustain the endothermic gasification with an improvement in the heat balance of the process.
The main feature of the combined plant is that the two reactors (gasification and oxidation) are fully integrated from a chemical and thermal point of view as shown in
Figure 2.
The reactor behaves as a countercurrent tube in-tube heat exchanger, where hot oxidation products preheat the cold gasification input feed through the wall of thin titanium shields. The outside wall of the reactor is made of stainless steel resistant to high operating pressures. The internal volume is separated in three coaxial chambers by means of septa made in titanium to stand up to corrosion. The differential pressure between the chambers is regulated at few bars in order to use thin septa. The central chamber is used for oxidation, with air supplied by a compressor. To ensure that air reacts exclusively in the central part of the reactor and to avoid the formation of hot spots, it is introduced directly into the reaction zone through a tube with lateral holes. The adjoining area is where gasification occurs, without oxygen. The third chamber, in contact with the reactor wall, is fluxed by carbon dioxide or another gas, which acts as a cooling fluid and an inert fluid that protect the wall from corrosion. This arrangement allows for the use of stainless steel as a pressure standing shell, and not expensive special alloys, instead of Inconel.
2.2. Model and Simulation
In the following paragraph a base process scheme of SCW-GcO with a nominal capacity of 100 kg/h is presented and discussed. The process was simulated using the Aspen Plus™ package. The conceptual process design is that described in
Figure 1.
As an input of the simulation, we selected heavy oil for gasification and fine carbon black for oxidation, the properties of which are reported in
Table 1.
The main method selected to calculate the thermodynamic properties and run the simulation was the ELECNRTL (Not-Random Two-Liquids for Electrolytes), which is suitable for aqueous systems in which salt solubility and precipitation phenomena are of interest. A number of selected ionic species (e.g., sulfides, chlorides, sulfates, nitrates, carbonates) were added to the simulation through the electrolyte wizard procedure, allowing for the calculation of the pH of the reactor effluents and the simulation of the neutralization section.
A second property method, the PSRK (Predictive Soave–Redlich–Kwong), was applied to some model blocks (e.g., gas–liquid separators) to better predict the solubility of low molecular weight gases in water systems under very high pressures.
Most of the components selected from the Aspen database to be part of the simulation were defined as “conventional” (e.g., common light gases, water, hydrocarbons, ionic species), i.e., they are handled according to the selected thermodynamic methods and participates to equilibria and reactions. Some compounds were defined as “conventional inert solid” (e.g., common oxides and their hydrated variants) in order to simulate the formation of solid inert ashes in the oxidation reactor whenever metal atoms enter the system under any form or species. Metal oxides, if formed inside the reactor, were simulated to be easily separated and purged as solids from the supercritical effluent.
The heavy oil inlet stream was characterized molecularly through an arbitrary selection of 36 compounds representative of few component families (linear alkanes, cycloalkanes, branched alkanes, aromatics, polyaromatics, sulfur-containing compounds). The selection of components was guided by information retrieved from literature about the typical species and families found in low-sulfur heavy fuel oil [
20,
21]. The complete composition is reported in
Table A1.
Similarly, the carbon black inlet stream, which may be rich in sulfur as well, was characterized assuming the arbitrary composition reported in
Table 1 [
22,
23], where high heating values (HHV) are also reported.
In both cases, the great number of model compounds selected, even if arbitrary, gives the simulator much flexibility when solving and closing material balances in reaction blocks.
The SCW-GcO reactor was assumed to be a hierarchical block that simulated the operation of the integrated oxidation/gasification reactor in supercritical water. The internal blocks were all solved with chemistry and Henry components disabled. Within the hierarchical block, the thermal exchanges were simulated with heater/cooler blocks in which the set specification was the duty thermal or the outlet temperature as calculated by the physical–mathematical model of the reactor developed with MATLAB code. The setup of the innovative reactor is described in
Figure A1 in
Appendix A.
H2STRAP was a block that simulated the H2S trap with selective adsorption at high pressure (~250 atm) on iron oxides (FeO). The operation was simulated with an adiabatic and isobaric stoichiometric reactor, in which the chemical reaction H2S + FeO → FeS + H2O was specified with a conversion of hydrogen sulfide equal to 99.9999%.
STAGE-1 and STAGE-2 were blocks that simulated a membrane separation system set up with split fractions obtained from performance simulations of third-party membrane modules and lamination valves to take account of in/out pressure drops.
Although simulations were performed at various temperature and feed concentrations, we report in
Table 2 data corresponding to only one set of operating conditions, with the gasification temperature set at 600 °C and oxidation temperature set at 800 °C. The input of the plant was made up of four streams: pure water at 16 kg/h (WATER-1), pure water at 14.17 kg/h (WATER-2), carbon black at 4.0 kg/h (C-BLACK), pyrolysis oil at 4.72 kg/h (HEAVYOIL), and air at 60 kg/h (AIR). The simulated output steams were gaseous products of oxidation after separation from the liquid phase (23), aqueous products of oxidation and gasification containing sulfuric and hydrochloric acid (ACIDWTR), ash formed in the oxidation and gasification sections, methane as the main product stream (CH4-OUT), and residue CO
2 and H
2S obtained after gas cleaning (CO2-OUT and FES).
2.2.1. Gasifier and Oxidizer Model
The gasification section was modeled by applying a RGIBBS block, which was able to predict the final product composition based on the principle of minimizing the total Gibbs free energy. The expected species specified in the Gibbs block consisted of major gas constituents (CO, H2, CO2, and CH4), light hydrocarbons (C2H4 and C2H6), inorganic species (HCl, H2S, N2, NH3, COS, and HCN), and tar components (C6H6, C7H8 and C10H8, and higher hydrocarbons).
The choice of a RGIBBS reactor to model the gasification reactions is supported by other previous papers by the same authors [
24], where it was stated that the calculated outlet composition of the gas phase is in good agreement with experimental data obtained at a residence time in the reactor of 120 s and a mean reaction temperature of 600 °C.
The gasification efficiency, defined as
expresses the mass of gas produced in the gasification chamber with respect to the amount of organic matter fed to the reactor. It was calculated at 43.3%. The flow rate and gas compositions of produced gas are reported in
Table 3. These values refer to stream 17G at the exit of the separator.
The liquid stream 17 L was mainly composed of water (98 wt%), carbon dioxide (1.55 wt%), methane (0.18 wt%), hydrogen sulfide (0.18 wt%), and traces of hydrogen, carbon monoxide, and ethane.
As for the oxidation section, the efficiency was expressed as % of TOC removal from the inlet stream of organic matter, where TOC is the total organic carbon (mg/L):
The feeds were stream SLRY-IN composed of carbon black, water, and polluted water from gasification and stoichiometric air (AIR). The effluent, after cooling in two heat exchangers, was separated in a separator (DEMIST) where liquid stream 11 L and gas stream 23 were obtained. In our case study, the TOC removal was calculated as 99.3%, which is a value coherent with that measured experimentally, which typically is above 99%. Stream 23 was composed of N2 (76 wt%), CO2 (22.62 wt%), Ar (1.14 wt%), H2O (0.18 wt%), and O2 (0.01 wt%), which were discharged into the atmosphere without further post-treatments.
The oxidation section was simulated through a separate RGibbs reactor in which the expected outlet species were exclusively H
2O and CO
2 as the main oxidation products; inorganic acids such as HCl, H
2SO
4, and HNO
3 (involved in aqueous salt equilibria in downstream equipment); and inert solid metal oxides that were easily separated. Sulfur and nitrogen oxides were not included among the outflow species, as they were not expected in the dry gaseous fraction of the effluent, as confirmed by the literature [
10].
2.2.2. Life Cycle Assessment
The goals of this work were the evaluation of the energy–environmental performance and the identification of the hot spots in the SCW-GcO process, compared to conventional incineration processes of dangerous waste, solvent mixtures, and exhaust oils. Analysis was carried out according to a “cradle to gate” approach: from the extraction of raw materials to the production of methane derived from the treatment of waste. The supply chain of the waste to be treated was neglected, such as leachate, carbon black, heavy oil, and the plant start-up phase. Looking at the entire useful life, considered to be 10 years, in terms of environmental impact the start-up phase was negligible. The end-of-life and disposal phases of the plant were also not taken into consideration. In fact, since the object of study was a pilot system in the construction phase, the results obtained referring to this phase would be characterized by a high uncertainty and would therefore be unreliable. Two cases were examined:
By avoided impact we mean that the methane produced decreased the consumption of natural gas from fossil sources used for feeding the end-user distribution network. This benefit associated with the production of methane was appreciated exclusively during use. During production this had no influence on energy consumption or the environmental impacts assessed. The functional unit (FU) of reference for the LCA was 1 ton of treated waste: carbon black, heavy oil, and leachate. System boundaries (SB) determined the process units to be included within the evaluated system. The system boundaries set for the LCA of SCW-GcO are shown in
Figure 3. In Case 2, the boundaries were extended to a power plant for electricity production using methane.
The mass and type of materials required for assembly was quantified for each component. Construction processes of the innovative reactors were analyzed in detail, considering the amount of energy used for assembly and processing of materials. For the working phase we referred to a useful life of 10 years and it was assumed that a working year consisted of 8000 h. Mass and energy flows (in and out) from this period of time were then quantified.
Auxiliary inputs for catalyzer components (catalytic post-gasifier and methanation reactor), the H2S trap, and acid water treatment were not considered because they were negligible compared to the main in/out currents. The energy consumption of pumps, compressors, and inverters was assumed considering rated powers and corrected with coefficients related to the working phase.