POLITECNICO DI MILANO School of Industrial Engineering Department of Energy SUPERCRITICAL WATER GASIFICATION OF BIOMASS: KINETIC APPROACH & PROCESS SIMULATION Relatore: Professor Emanuele Martelli Co-relatore: Professor Wiebren de Jong (Delft University of Technology) Master of Science Thesis of: Guchan Yapar - 797694 Academic Year 2014-2015 To the coal miners of my country and the world; I hope one day you will not have to risk your lives to give us a light' ∗to my beloved Mother and nephew Aslan. ACKNOWLEDGEMENTS I would like to first give my thanks to Prof. Emanule Martelli, supervisor of my thesis work in Politecnico di Milano, for his great help even from an other country. I am very grateful to my supervisor in Delft University of Technology, Prof. Wiebren de Jong, for all his support and effort he spent for the success of this work. Special thanks to my daily supervisor Onursal Yakaboylu who is a great science person and a friend. Wishing you the best of luck. I would like to say thank you, to my friends in Italy, Turkey and new ones in Netherlands; you have made this journey more memorable. ABSTRACT Thermochemical biomass conversion is a method to transform stored energy to higher energy density solid, liquid or gas fuels by chemically decomposing the high molecular weight biopolymers. Biomass derived green gas has been worked extensively with the aim of taking over the place of fossil based combustible gas, for people's self-sufficiency in energy consumption along the earth and preserving a sustainable environment. Biomass gasification at high temperature - low pressure conditions targets to produce a combustible gas, abundant of H2 (high calorific gas) and CO (reactant of further shift reactions). Stated technology is able to scale up to 600 MW energy production capacities for various biomass or residues but with different thermal efficiencies due to the external heat requirement. Among all other physical and chemical properties, moisture content of the biomass may result in thermal efficiency issues and parallel to this power production decreases. Supercritical Water Gasification (SCWG) allows the thermochemical conversion of wet biomass in hydrothermal media. High-pressure biomass slurry can be kept in aqueous media until the critical point of water. Right above nearcritical temperature level (∼330 ℃) hydrogen bond number decreases and water molecule becomes more apolar, as a consequence water solubility increases allowing one-phase homogenous reactions; furthermore, ion concentration in the hydrothermal media increases, which favors the acid or base catalyzed reactions. Thermo dynamical benefits have aroused the curiosity of several science groups for the last 20 years, providing modeling and experiment studies on supercritical water gasification of biomass (SCWG). However, detailed kinetic modeling of the process for lignocellulosic biomass is missing, strongly needed for further technology improvements. In this work, review and evaluation of the biomass kinetic approach has been performed with the aim of proposing reaction network and rate parameters for three main biomass structural compounds; cellulose, hemicellulose and lignin. AspenPlus 8.3 engineering software package has been aided for the simulation of the tubular reactors, integrated with the compiled kinetic data from literature and proposed network. Validation results show that reactor simulation has a good accuracy of predicting gasification efficiency, carbon efficiency and gas yield for a range of residence time (15-47 s), pressure (20-30 MPa) and biochemical composition (mostly agricultural residues). Reactor simulation validations showed the possibility of inserting consecutive reactors to a simplified SCWG process simulation. Process scheme including 3 tubular flow reactors and auxiliary units has been designed; according to which, carbonmonoxide low product gas and high gasification efficiency (4,45 % , 92,24%, respectively) have been achieved. The results of this work show that kinetic model developed for lignocellulosic biomass, based on the literature data is able to predict gasification efficiency, intermediate compound composition, gas composition and thermal energy output for different operating conditions and reactor lengths. TABLE OF CONTENTS 1 INTRODUCTION ......................................................................................................................... 1 1.1 BACKGROUND ........................................................................................................................................ 1 1.2 BIOMASS CONVERSION TECHNOLOGIES-PLACEMENT OF HYDROTHERMAL BIOMASS GASIFICATION ................................................................................................................................................... 2 1.3 SUPERCRITICAL WATER ...................................................................................................................... 5 1.4 SCWG OF BIOMASS .............................................................................................................................. 7 1.4.1 Main Research Groups and Experiments on SCWG ..........................................................9 1.5 REACTOR SYSTEMS .............................................................................................................................12 1.6 SCOPE AND OBJECTIVES OF THE STUDY ..........................................................................................14 2 LITERATURE REVIEW OF BIOMASS DEGRADATION IN HYDROTHERMAL MEDIA: REACTION MECHANISM AND KINETICS ..................................................................16 2.1 CELLULOSE DECOMPOSITION ...........................................................................................................17 2.1.1 Glucose Decomposition .............................................................................................................. 20 2.2 HEMICELLULOSE DECOMPOSITION ..................................................................................................23 2.3 LIGNIN DECOMPOSITION ...................................................................................................................25 3 PROPOSED REACTION PATHWAY .....................................................................................31 3.1 Cellulose Pathway: ........................................................................................................................... 32 3.1.1 Decomposition of Cellobiose .................................................................................................... 32 3.1.2 Reactions of Glucose-Fructose decomposition products ............................................ 33 3.2 Hemicellulose Pathway.................................................................................................................. 35 3.3 LIGNIN PATHWAY:..............................................................................................................................37 3.4 ACID DECOMPOSITION PATHWAY: ..................................................................................................39 3.5 KINETICS DATA AND ARRHENIUS PARAMETERS...........................................................................41 4 INTEGRATION OF KINETIC DATA SET TO THE REACTOR SIMULATION- TEST RUN & VALIDATION .......................................................................................................................52 4.1 INTEGRATION OF KINETIC DATA SET TO THE ASPENPLUSTM SIMULATION .............................52 4.2 RESULTS OF THE REACTOR TEST-RUN ...........................................................................................54 4.3 VALIDATION OF THE SIMULATION ...................................................................................................59 4.3.1 Effect of Residence Time ............................................................................................................ 61 4.3.2 Effect of Pressure .......................................................................................................................... 64 4.3.3 Effect of Feedstock........................................................................................................................ 66 5 SENSITIVITY ANALYSIS .........................................................................................................69 5.1 SENSITIVITY ANALYSIS FOR REACTOR HEATING RATE EFFECT: ...............................................69 5.2 SCWG PROCESS SCHEME BUILD-UP & SENSITIVITY ANALYSIS FOR PARAMETER VARIATIONS ON SCWG PLANT SIMULATION ............................................................................................77 5.2.1 Design Parameters and Unit Selections ............................................................................. 77 5.2.2 SCWG OF CELLULOSIC BIOMASS PLANT PROCESS SIMULATION ...........................................80 5.2.3 Effect of Maximum Temperature .......................................................................................... 81 5.2.4 Effect of Biomass Load ............................................................................................................... 82 5.2.5 Effect of Biomass Type................................................................................................................ 84 CONCLUSIONS AND FURTHER DEVELOPMENTS ..................................................................87 REFERENCES:....................................................................................................................................89 ANNEXES ............................................................................................................................................95 A- SUBCRITICAL REGION REACTIONS ....................................................................................................95 B- SUPERCRITICAL REGION REACTIONS ................................................................................................96 C- RGIBSS REACTION STOICHIOMETRIES (FOR LIGNIN PATHWAY) ..................................................98 D- COMPONENT LIST ................................................................................................................................98 FIGURE 1.1 APPLICATION OPTION SCHEME OF SCWG PROCESS OUTLET [7] .................................................................. 4 FIGURE 1.2 PHASE DIAGRAM OF WATER [8].......................................................................................................................... 5 FIGURE 1.3 PROPERTIES OF SUPERCRITICAL WATER [10].................................................................................................. 6 FIGURE 1.4 SIMPLIFIED PROCESS SCHEME OF THE PDU FOR SCWG OF BIOMASS OPERATED AT UNIVERSITY OF TWENTE [7]...................................................................................................................................................................13 FIGURE 2.1 CELLOBIOSE DECOMPOSITION PATHWAY: 300 ℃ < T < 400 ℃, 25 < P < 40 MPA [51] ..........17 FIGURE 2.2 -LN(1 - X) VERSUS RESIDENCE TIME FOR CELLOBIOSE DECOMPOSITION [51] .........................................18 FIGURE 2.3 ARRHENIUS PLOT OF THE RATE CONSTANT OF CONVERSION OF MICROCRYSTALLINE CELLULOSE IN WATER (290 C – 400 OC, 25 MPA), [50]...............................................................................................................19 FIGURE 2.4 MECHANISM OF GLUCOSE DECOMPOSITION IN SUB- AND SUPERCRITICAL WATER: 300 ℃ < T < 400 ℃, < P < 30 MPA [53]............................................................................................................................................21 FIGURE 2.5 GLUCOSE DECOMPOSITION PATHWAY IN SUB- AND SUPERCRITICAL WATER: 300 ℃ < T < 460 ℃ , P = 25 MPA [3]. .......................................................................................................................................................22 FIGURE 2.6 D-XYLOSE DECOMPOSITION IN HOT COMPRESSED WATER: 180℃ < T < 220℃, P = 10 MPA [56] ..........................................................................................................................................................................................23 FIGURE 2.7 KINETIC MODEL REACTION MECHANISM FOR DECOMPOSITION OF XYLOSE IN SUPERCRITICAL WATER 450℃ < T < 650℃, P = 25 MPA [57] ............................................................................................................24 FIGURE 2.8 D-XYLOSE DECOMPOSITION AND GASIFICATION PATH IN SUPERCRITICAL WATER: KINETIC MODEL REACTION MECHANISM FOR DECOMPOSITION OF XYLOSE IN SUPERCRITICAL WATER 450℃ < T < 650℃ , P = 25 MPA [57] .....................................................................................................................................................24 FIGURE 2.9 GC-MS CHROMATOGRAM FOR SOLUBLE PRODUCTS OF GUAIACOL DECOMPOSITION IN SUPERCRITICAL WATER [59] ...................................................................................................................................................................25 FIGURE 2.10 SIMPLE REACTION PATHWAY FOR GUAIACOL (LIGNIN MODEL COMPONENT) IN SUPERCRITICAL WATER [59] ...................................................................................................................................................................26 FIGURE 2.11 SIMPLE REACTION PATHWAY OF CATECHOL DECOMPOSITION IN NEAR- AND SUPERCRITICAL WATER [60] .................................................................................................................................................................................27 FIGURE 2.12 CATECHOL FORMATION FROM GUAIACOL IN SUB- AND SUPERCRITICAL WATER [61] .........................28 FIGURE 2.13 PHENOL FORMATION FROM GUAIACOL AND CATECHOL IN SUB- AND SUPERCRITICAL WATER [61]..29 FIGURE 2.14 REACTION PATHWAY OF GUAIACOL UNDER HYDROTHERMAL CONDITIONS: 300℃ < T < 450℃ , P = 25 MPA [61] .....................................................................................................................................................30 FIGURE 2.15 REACTION SCHEME OF PHENOL BENZENE DECOMPOSITION IN SUPERCRITICAL WATER: 300℃ < T < 450℃, P = 25 MPA [39] ...............................................................................................................................30 FIGURE 3.1 REACTION SCHEME OF CELLOBIOSE IN NEAR CRITICAL WATER: K1, K2, K3, KGE.G, KGG.G FROM [51], KG.E, KG.A, KG.GLY, KG.F, KG.A, KF.E, KF.GLY FROM [53], KGLY.DIH, KDIH.GLY, KGLY.P, KDIH.P FROM [22], KG.5, KF.5, KF.FU , KFU.CH FROM [3]....................................................................................................................................................................................33 FIGURE 3.2 INTERMEDIATE DECOMPOSITION SCHEME KF.ACID, KP.ACID, KA.ACID, KE.ACID [54]; K5.LF, K5.FF [62] ..........................................................................................................................................................................................34 FIGURE 3.3 DECOMPOSITION SCHEME OF D-XYLOSE IN SUBCRITICAL WATER [56] .....................................................35 FIGURE 3.4 D-XYLOSE DECOMPOSITION AND GASIFICATION SCHEME IN SUPERCRITICAL WATER KXY.FU,KXY.WSHS,KFU.WSHS,KAA.GA FROM [57] ......................................................................................................................36 FIGURE 3.5 GUAIACOL DECOMPOSITION SCHEME IN SUBCRITICAL WATER [61] ...........................................................37 FIGURE 3.6 GUAIACOL DECOMPOSITION SCHEME IN SUPERCRITICAL WATER KGU.CH, KHU.GA, KGU.OC, KGU.C, KGU.T, KC.OC, KT.CH FROM [61]; KC.T, KP.C, KP.T, KP,GA, KP.CH, KT.B, KB.T, KB.P, KB.GA, KB.NA, KNA.CH, KB.CH FROM [39] ..........................................................................................................................................................38 FIGURE 3.7 ORGANIC ACID GASIFICATION SCHEME IN SUPERCRITICAL WATER. KFA.GA1, KFA.GA2 [63]; KAA.GA [64]; KWSHS.GA, KPA.GA, KMF.AA [57]; LACTIC ACID REACTIONS [65] ............................................................39 FIGURE 3.8 ARRHENIUS RELATION OF GLUCOSE TO GLYCERALDEHYDE DECOMPOSITION REACTION ........................43 FIGURE 4.1 SIMULATION RESULTS OF CELLOBIOSE DECOMPOSITION PRODUCTS IN ISOTHERMAL SUBCRITICAL REACTOR AT 370 ℃.....................................................................................................................................................55 FIGURE 4.2 SIMULATION RESULTS OF GLUCOSE DECOMPOSITION PRODUCTS IN ISOTHERMAL SUBCRITICAL REACTOR AT 370 ℃.....................................................................................................................................................56 FIGURE 4.3 SIMULATION RESULTS OF XYLOSE DECOMPOSITION PRODUCTS IN ISOTHERMAL SUBCRITICAL REACTOR AT 370 ℃ ......................................................................................................................................................................56 FIGURE 4.4 SIMULATION RESULTS OF GUAIACOL DECOMPOSITION PRODUCTS IN ISOTHERMAL SUBCRITICAL REACTOR AT 370 ℃.....................................................................................................................................................57 FIGURE 4.5 SIMULATION RESULTS OF ORGANIC ACID DECOMPOSITION PRODUCTS IN ISOTHERMAL SUBCRITICAL REACTOR AT 370 ℃.....................................................................................................................................................57 FIGURE 4.6 SIMULATION RESULTS OF XYLOSE DECOMPOSITION PRODUCTS IN ISOTHERMAL SUPERCRITICAL REACTOR AT 650 ℃.....................................................................................................................................................58 TABLE OF FIGURES FIGURE 4.7 TEMPERATURE PROFILE IN SUBCRITICAL REACTOR ......................................................................................59 FIGURE 4.8 TEMPERATURE PROFILE IN SUPERCRITICAL REACTOR ..................................................................................60 FIGURE 4.9 SCHEME OF THE EXPERIMENTAL SETUP USED IN LU’S WORK [9] ...............................................................61 FIGURE 4.10 EFFECT OF RESIDENCE TIME ON PRODUCT GAS MOLAR FLOW...................................................................62 FIGURE 4.11 COMPARISON BETWEEN SIMULATION RESULTS GE, CE AND EXPERIMENTAL RESULTS GE*, CE* FOR RESIDENCE TIME BETWEEN 15 AND 47 SECONDS ...................................................................................................63 FIGURE 4.12 COMPARISON BETWEEN PRODUCT YIELD RESULTS OF THE SIMULATION AND EXPERIMENTAL RESULTS (*), FOR RESIDENCE TIME BETWEEN 15 AND 47 SECONDS ...................................................................................64 FIGURE 4.13 COMPARISON BETWEEN SIMULATION RESULTS GE, CE AND EXPERIMENTAL RESULTS GE*, CE* FOR RESIDENCE REACTOR PRESSURE BETWEEN 200 AND 300 BARS ..........................................................................65 FIGURE 4.14 COMPARISON BETWEEN PRODUCT YIELD RESULTS OF THE SIMULATION AND EXPERIMENTAL RESULTS (*), FOR REACTOR PRESSURE BETWEEN 200 AND 300 BARS ...............................................................................66 FIGURE 4.15 COMPARISON BETWEEN SIMULATION RESULTS GE, CE AND EXPERIMENTAL RESULTS GE*, CE* FOR DIFFERENT FEEDSTOCK; RICE STRAW, PEANUT SHELL, CORN STALK, CORN COB AND WOOD SAWDUST. ........67 FIGURE 5.1 TEMPERATURE PROFILE IN SUBCRITICAL REACTOR FOR THE 1ST SCENARIO .............................................71 FIGURE 5.2 TEMPERATURE PROFILE IN SUPERCRITICAL REACTOR FOR THE 1ST SCENARIO ........................................71 FIGURE 5.3 TEMPERATURE PROFILE IN SUBCRITICAL REACTOR FOR THE 2ND SCENARIO ............................................72 FIGURE 5.4 TEMPERATURE PROFILE IN SUPERCRITICAL REACTOR FOR THE 2ND SCENARIO ........................................72 FIGURE 5.5 TEMPERATURE PROFILE IN SUBCRITICAL REACTOR FOR THE 3RD SCENARIO .............................................73 FIGURE 5.6 TEMPERATURE PROFILE IN SUPERCRITICAL REACTOR FOR THE 3RD SCENARIO ........................................73 FIGURE 5.7 GE AND CE VALUES OF 1ST 2ND AND 3RD SCENARIOS ......................................................................................74 FIGURE 5.8 CARBON MOLE RATIO OF UNCONVERTED LIQUID IN THE EFFLUENT, CARBON (UNC.LIQ)% AND MOLAR FLOWRATE OF CARBON CONTAINED IN CHAR FOR 1ST 2ND AND 3RD SCENARIOS...................................................75 FIGURE 5.9 PRODUCT GAS COMPOSITION VALUES FOR FIRST, SECOND AND THIRD SCENARIOS ..................................76 FIGURE 5.10 SIMPLIFIED HEAT EXCHANGE UNIT SCHEME.................................................................................................78 FIGURE 5.11 GE AND CE RESULTS OF THE PROCESS SIMULATION FOR TMAX= 650℃, 600 ℃, 550 ℃ ..................81 FIGURE 5.12 BIOMASS LOAD EFFECT ON GE AND CE ........................................................................................................82 FIGURE 5.13 BIOMASS LOAD EFFECT ON THERMAL ENERGY SUPPLIED AND NET THERMAL ENERGY PRODUCTION RATE ................................................................................................................................................................................83 FIGURE 5.14 BIOMASS LOAD EFFECT ON SPECIFIC THERMAL ENERGY SUPPLIED AND NET THERMAL ENERGY PRODUCTION RATE ........................................................................................................................................................84 FIGURE 5.15 BIOMASS TYPE EFFECT ON GE, CE AND CGE ..............................................................................................85 FIGURE 5.16 BIOMASS TYPE EFFECT ON PRODUCT GAS COMPOSITIONS..........................................................................86 TABLE 1.1 PROXIMATE ANALYSIS AND CALORIFIC VALUES OF PIG MANURE, ALGAE, COW MANURE AND RICE STRAW SAMPLES [5]..................................................................................................................................................................... 3 TABLE 1.2 SELECTED SCWG RESEARCH GROUPS [25] ....................................................................................................... 9 TABLE 3.1 CELLOBIOSE DECOMPOSITION RATE CONSTANTS ............................................................................................41 TABLE 3.2 GLUCOSE DECOMPOSITION RATE CONSTANTS ..................................................................................................42 TABLE 3.3ARRHENIUS PARAMETERS OF CELLOBIOSE AND GLUCOSE DECOMPOSITION ...............................................43 TABLE 3.4 GLUCOSE DECOMPOSITION RATE CONSTANTS CONTINUED [3] .....................................................................44 TABLE 3.5 ARRHENIUS PARAMETERS OF GLUCOSE DECOMPOSITION CONTINUED........................................................44 TABLE 3.6 XYLOSE DECOMPOSITION RATE CONSTANTS IN SUBCRITICAL WATER ..........................................................45 TABLE 3.7 ARRHENIUS PARAMETERS OF D-XYLOSE DECOMPOSITION SUBCRITICAL WATER.......................................45 TABLE 3.8 ARRHENIUS PARAMETER OF D-XYLOSE DECOMPOSITION SUPERCRITICAL WATER ....................................45 TABLE 3.9 LIGNIN DECOMPOSITION RATE CONSTANTS SUB- AND SUPERCRITICAL WATER [61] ...............................46 TABLE 3.10 ARRHENIUS PARAMETERS OF GUAIACOL DECOMPOSITION SUBCRITICAL WATER ...................................47 TABLE 3.11 ARRHENIUS PARAMETERS OF GUAIACOL DECOMPOSITION SUPERCRITICAL WATER ..............................47 TABLE 3.12 PHENOL- BENZENE DECOMPOSITION REACTION RATE CONSTANTS SUPERCRITICAL [39] ....................48 TABLE 3.13 ARRHENIUS PARAMETER OF PHENOL-BENZENE DECOMPOSITION IN SUPERCRITICAL WATER .............48 TABLE 3.14 ARRHENIUS PARAMETERS OF FORMIC ACID DECOMPOSITION [63]...........................................................49 TABLE 3.15 ARRHENIUS PARAMETERS OF ACETIC AND PROPIONIC ACID DECOMPOSITION ........................................49 TABLE 3.16 LACTIC ACID DECOMPOSITION REACTION RATE CONSTANTS [65] ............................................................50 TABLE 3.17 ARRHENIUS PARAMETERS OF LACTIC ACID DECOMPOSITION SUPERCRITICAL WATER ...........................51 TABLE 4.1SIMULATION SET PARAMETERS FOR PRELIMINARY RUN .................................................................................54 TABLE 4.2 SIMULATION OPERATING PARAMETERS; COMPARISONS FOR VARYING RESIDENCE TIME .........................61 TABLE 4.3 SIMULATION OPERATING PARAMETERS; COMPARISONS FOR VARYING PRESSURE .....................................64 TABLE 4.4 BIOCHEMICAL AND ASH MASS COMPOSITIONS OF DRY BIOMASS SAMPLES USED IN THE SIMULATION ....66 TABLE 5.1 SIMULATION SET PARAMETERS ..........................................................................................................................70 TABLE 5.2 TUBULAR REACTOR SIZING PARAMETERS .........................................................................................................79 TABLE 5.3 BASE CASE SIMULATION RESULTS ......................................................................................................................80 NOMENCLATURE Roman M 𝑤𝑎𝑡𝑒𝑟 ∆𝐻𝑣𝑎𝑝 Pc Tc V S t k ks A Ea Kw r R T0 Ci a, b c 𝑚𝑔,𝑜 𝑚𝑏𝑖𝑜,𝑑𝑟𝑦 𝑛𝑐,𝑔 𝑛𝑐,𝑏𝑖𝑜 𝑛𝑖 𝑌𝑖 ∆𝑇𝐿𝑀 𝑈 ℎ𝑖 𝑘𝑤𝑎𝑙𝑙 𝑟 𝑄 𝐿𝑡𝑢𝑏𝑒 𝐷𝑡𝑢𝑏𝑒 𝑄𝑠𝑢𝑝𝑝,𝑡ℎ 𝑄𝑡ℎ,𝑛𝑒𝑡 𝑄 𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝑇𝑟 Units Moisture content Heat of vaporization of water Critical pressure Critical temperature Volume Surface area Time Reaction rate constant Surface reaction rate constant Pre-exponential factor Activation energy Water ion product Reaction rate Gas constant Reference temperature Molar concentration of i Volume parameter of SRK Mathias-Copeman constants Mass flow rate of gas outlet Mass flow rate of dry biomass inlet Molar flow rate of carbon in gas outlet Molar flow-rate of carbon in biomass inlet Molar flow rate of gas i Molar yield of gas i Log-mean temperature difference Overall heat transfer coefficient Convective heat transfer coefficient of fluid i Conductive heat transfer coefficient of wall Radius Heat transfer rate Tube length Diameter of tube Externally supplied heat Net thermal energy production rate Net specific thermal energy production Reduced temperature kgwater/kgbio MJ.kg-1 MPa 0C m3 m2 s s-1 m.s-1 s-1 kJ.mol-1 mol2.dm-6 mol.m-3s-1 J.mol-1K-1 K mol.m-1 kg/hr kg/hr mol/hr mol/hr mol/hr mol/kg K W.m-2K-1 W.m-2K-1 W.m-1K-1 m MJ.s-1 m m MJ.s-1 MJ.s-1 MJ.kg-1 Greek 𝜀 𝛼𝑖 𝑣̃ 𝜔 Dielectric constant Stoichiometric coefficient of i Specific volume Accentric factor ∏ Multiplication m3.kg-1 Abbreviations PFR CSTR TOC DP WSHS SCW SCWG SCWO HHV LHV SOFC PEM CMC GE CE CGE Plug flow reactor Continuous stirred tank reactor Water soluble organic compound Decomposition product Water soluble humic solid Supercritical water Supercritical water Gasification Supercritical water Oxidation Higher (Gross) heating Value Lower (Net) heating value Solid oxide fuel cell Proton exchange membrane Carboxy methyl cellulose Gasification efficiency Carbon gas efficiency Cold gas efficiency MJ.kg-1 MJ.kg-1 % % % 1 Introduction 1.1 Background Pursue of continual global economic growth obliges mankind to guarantee a way to store, convert and amplify the energy. Huge difference between human energy exploitation rate and fossil fuel recovery (millions of years) do not leave any doubt about necessity of finding efficient and wide range applicable renewable energy production ways. Biomass is a dispersed, energy source contained in plants and animal wastes [1]. Energy from biomass is carbon neutral and environmentally benign thanks to the continuous loop of carbon between plants and atmosphere. Its exploitation is mainly done through biological or thermal conversion processes. Biological technologies mainly use microorganisms as energy converting agents from biomass to fuel. The oldest thermal biomass conversion, combustion have been used by generations and in developing countries still in general use. Biomass accounts for 14% of the world energy consumption while even though world annual biomass potential is estimated to be around 146 billion metric tons (almost half of the worlds annual energy demand) [2]. Fully exploitation of biomass is not realistic and reasonable, however it is clear that effective and widespread technologies are capable of decreasing the consumption of other fossil fuels. Higher technology processes, including biological technologies, have been worked in order to achieve high energy production capacity and wide scope of biomass use. Standard gasification has the biggest energy production share from biomass sources due to highenergy content gas product and technology maturity. Performance of the gasification process depend many operating conditions and biomass properties, including moisture. Thermal or electrical energy dissipated for the evaporation of water contained in wet biomass (grass, rice straw, sugar cane, algae, manure etc.) is still one of the biggest challenges of the current technology. However, degradation of biomass with the presence of water is also possible according to the early thermodynamic findings. Activity of water in the process can be multiplied by increasing the slurry temperature near the critical point under critical pressure (Tc, Pc). Physical changes in hot and compressed water are favorable for rapid hydrolysis, isomerization, dehydration and some other decomposition reactions [3]. Low carbonmonoxide (CO) compositions is a result of favored forward water gas shift reaction (WGSR) in SCWG. In this chapter; firstly, the biomass conversion technologies are evaluated by their strengths and weaknesses in order to propose possible specific advantages of the study subject SCWG. Property behavior of water in near- and superciritical conditions is reviewed for a better understanding of the effects of water on reaction system. Further analysis was done for the studies held on the subject in terms of experiments, modelling studies and reactor designs. 1 1.2 Biomass Conversion technologies-Placement of Hydrothermal Biomass Gasification Biomass is a general term for various biological sources, byproducts and wastes. Feasibility and capability of a process is strongly dependent on physical and chemical properties of the feedstock supplied. The main division for biomass processing technologies can be made as biological and thermal. The most studied and well-known processes are classified as follows. Thermo-chemical Processes Combustion: Aims to produce thermal energy from combustion reaction enthalpy Supercritical water oxidation (SCWO) Conventional (High Temperature) Gasification: Aims to produce syngas (H2, CO) in oxygen deprived media Hydrothermal Liquefaction Hydrothermal Gasification (Supercritical Water Gasification): Aims to produce H2 and CH4 rich product gas in hot compressed water. Pyrolysis: Aims to produce high energy value product as solid char, liquid biofuel and/or combustible gas. Biochemical Processes Fermentation Anaerobic digestion: Aims to produce biogas as a digestion products of bacterial growth High temperature gasification is the most comparable technology to the hydrothermal biomass gasification due to the aimed product similarities (H2, CO and CH4) and scale of operation. Tradeoff between the energy recovered in the product and energy input to run the thermochemical process mainly determines energy efficiency and feasibility of a system. Many types of biomass, such as wet forestry, rice straw, animal manure, algae and sewage sludge contain high amount of moisture as received (up to 95%) [4]. Proximate analysis and calorific values of several wet biomass types are shown in Table 1.1. Lower Heating Value (LHV) of biomass source is estimated by subtracting the heat of vaporization of water contained from Higher Heating Value (HHV) as shown in 𝑤𝑎𝑡𝑒𝑟 Equation 1-1, where M refers to moisture content of the fuel and ∆𝐻𝑣𝑎𝑝 means heat of vaporization of water. Use of wet biomass for classical gas- phase gasification and liquefaction processes therefore may result in high drying cost and low efficiency in energetic and economical ways. 𝑤𝑎𝑡𝑒𝑟 𝐿𝐻𝑉 = 𝐻𝐻𝑉 ∙ (1 − 𝑀) + ∆𝐻𝑣𝑎𝑝 ∙𝑀 Equation 1-1 2 Table 1.1 Proximate analysis and calorific values of pig manure, algae, cow manure and rice straw samples [5] Pig Manure Type Algae Cow Manure Rice Straw wt% wt% 92.10 2.80 71.87 22.61 13.88 13.67 7.01 15.64 wt% wt% 4.05 1.05 1.67 3.85 60.52 11.94 61.95 15.40 -1.24 4.86 13.61 12.50 1.09 7.06 14.95 13.67 Proximate Analysis Moisture content Ash content Volatile matter Fixed carbon Calorific Values Net calorific value (LHV) MJ/kg Gross calorific value (HHV) MJ/kg At this very point, hydrothermal conversion of biomass in supercritical water is a promising alternative. Hydrothermal terminology refers to aqueous media at high temperature and pressure close to the critical point of water (221 bar and 374 ℃). Hydrothermal gasification (or SCWG) of biomass does not require evaporation of the water from biomass, which helps to save significant amount of external energy depending on the moisture level of the feedstock. High temperature and pressure water near and beyond its critical point changes phase properties sharply from liquid to supercritical water. Biomass constituents become miscible in hot and compressed water, resulting in lower mass transfer resistance and higher rate homogenous phase hydrolysis reactions. Physical changes in near- and supercritical water (in density, viscosity, number of hydrogen bonds, dielectric constant and ion product) enables water to be not only a solvent but also a natural catalyst and reactant [6] Renewable combustible gas production via SCWG is advantageous at some aspects: Firstly product gas is at high pressure (25-30 MPa), convenient for the utilization of the stream. Next, very low CO concentration in the product gas dismisses the need of WGS reactor. High CO2 concentration on the other hand eases the possible CO2 sequestration. SCWG product gas composed mainly of H2 and CH4 can be mixed with the natural gas distribution network, according to the short-medium strategy of Netherlands. In addition to that product gas can be used for SOFC or PEM fuel cell applications, in the latter case purified H2 is needed. Except still being a novel process, mentioned advantages create operational options for the process in short-medium term plans (see Figure 1.1) 3 Figure 1.1 Application option scheme of SCWG process outlet [7] Other process advantages of SCWG can be listed as follows: [7] Low selectivity towards char and tar forming reactions (compared to standard thermochemical biomass conversion processes) High solubility of CO2 in high pressure water to be separated easily in a low pressure flash column Low solubility of inorganics (near-critical and supercritical region) which are separated in a solid-liquid separator Compact process equipment due to high stream density (comparing to standard gasification process) Complete biomass conversion without use of catalyst (T>650℃) Besides the exceptional advantages of the SCWG process, char and tar plugging problems and energy intensive operating conditions are the main drawbacks to be worked on. 4 1.3 Supercritical Water The properties of water show variance in the wide region of temperature and pressure from room conditions up to critical point and beyond. Critical point is defined where temperature is critical temperature (TC=374,4 ℃) and pressure is critical pressure (PC=22,1 MPa), for water. As depicted in Figure 1.2, the critical point is the end point of liquid to vapor transition line in the vapor-liquid region, along which evaporation temperature rises with pressure. Beyond the critical point, water does not change phase simply from liquid to vapor or vice versa but becomes single-phase supercritical fluid. In the region before and after that point water changes its properties rapidly. For example, its dipole moment decreases to a point of a conventional solvent, it becomes as polar as acetone and its pH value decreases three units compared to ambient water, which results in a high concentration of H3O+. Figure 1.2 Phase diagram of water [8] As can be viewed in Figure 1.3, water at its critical point acts in between liquid water and steam, beyond the critical point the conductivity, density and viscosity of water becomes smaller (gas like) with temperature. Specific heat capacity of water shows a peak at which water is at critical condition. Rapid property changes as such can be used as an advantage in an SCWG process. For example, density change of water in limited temperature range decreases its dielectric constant which allows the water to be a very good solvent not only for organic compounds but also product gases which allows reactions to proceed in one phase, resulting in higher reaction rates. The high ion product constant (KW) of water shows catalytic effect for ionic type reactions especially in the sub- and near-critical region where KW is the largest. SCWG of wet biomass, therefore, is a 5 challenging but attractive technology mainly due to the tunable properties of water in the region of sub- and supercritical and its non-monotonic effects on overall reaction kinetics [9]. Figure 1.3 Properties of supercritical water [10] 6 1.4 SCWG of biomass The idea of combustible gas production from feed biomass with the presence of water was initiated by the method of steam reforming of biomass to produce hydrogen as a competitive to fossil fuel product hydrogen [11], [12]. 𝐶6 𝐻10 𝑂5 + 7𝐻2 𝑂 → 6𝐶𝑂2 + 12𝐻2 The ideal stoichiometric equation was developed to interpret the reaction of cellulose (C6H10O5) with steam for hydrogen production. In 1978 a thermochemical equilibrium study was made by Antal [12] in order to predict the hydrogen production from the reaction of 1 mole of cellulose with 7 moles of steam. According to his equilibrium prediction at 600 ℃ and 1 bar no solid carbon remained unconverted. Kinetic studies, on the other hand, showed that steam does not take part in the fast pyrolysis of biomass at atmospheric conditions [13],[14]. Experimental studies confirmed the kinetic studies, resulting in char and tar formation plus the presence of high molecular weight hydrocarbons in product gas[15]. Corella [16] reported correlated results by observing a char yield of 10-20 % from pyrolysis of wood dust; char and tar gasification in order to increase gas efficiency of the steam reforming at atmospheric pressure was also investigated. It was found that tar can be to a large extent be gasified at 800 – 875 ℃ at atmospheric pressure with calcined dolomite as bed material in a secondary fluidized bed, but char remained partly unconverted [16]. These findings show that the mentioned typical thermochemical process conditions of the system are not adequate to produce low molecular weight product gases from biomass by steam reforming at atmospheric pressure. Biomass conversion in sub- and supercritical water (SCW) was studied by Modell, showing tar and gas formation without char production at pressures exceeding the critical pressure of 22,1 MPa [17]. Experiments showed that cellulose decomposes in subcritical water rapidly, liquid water at 190 ℃ and 22,1 MPa is able to hydrothermolyse lignin and hemicellulose in a short duration (few minutes) [17]–[19]. Solvolysis products of biomass compounds further decompose by isomerization, condensation, fragmentation and dehydration reactions [18], [20]–[23], producing low molecular weight gas (composed mainly of hydrogen, carbondioxide and methane) and low amount of tars [24]. The SCWG process, according to the studies conducted is capable of achieving complete biomass conversion to combustible gases. Maximum process temperature is an important parameter to point out the desired product and decision on catalysis use. Biomass gasification in hydrothermal conditions can be divided into three main categories [25]: low temperature catalytic gasification (𝑇 ≪ 𝑇𝑐 ) moderate temperature catalytic or non-catalytic gasification (near critical temperatures to 500 ℃) high temperature catalytic or non-catalytic gasification (𝑇 > 500 ℃) 7 In this work our scope of investigation is high temperature non-catalytic gasification at 500 – 800 ℃ to produce combustible gases, which will be expressed simply as SCWG. 8 1.4.1 Main Research Groups and Experiments on SCWG There are five main research groups that have conducted different experiments on SCWG of biomass. Table 1.2 displays the operating conditions, feedstock and reactor systems of the researches [25]. Table 1.2 Selected SCWG research groups [25] T( ℃) Research Feedstocks Group Antal, U. of <22 wt.% wet biomass (corn ~600 Hawaii starch, potato starch and wastes, wood sawdust, water hyacinth, cellulose, macadamia nut shell, sugar cane bagasse, glucose and other model compounds) P (MPa) Reactor System Ref. 22-34.5 Hastelloy & Inconel tubular reactors, capillary tube reactors, packed bed systems; catalysts: activated carbon and charcoal [26]– [30] Kruse, Dinjus and Boukis, FZK, Karlsruhe 1–5 wt.% glucose, vanillin, 400–700 25–50 glycine, sawdust, straw, cellulose, plants, meats, corn silage with ethanol and pyroligneous acid, pyrocatechol and phytomass, corn starch, clover grass, sewage sludge and lignin Pilot plant, 2–4 min; batch, CSTR, tubular reactors; 60 s–1 h; alkali catalysts: KOH, K2CO3 and KHCO3 [31]–[33] van Swaaij, U. of Twente, Netherland s Matsumura , U. of Hiroshima, Japan Model compounds of 1–18 460–800 24–30 wt.% formic acid, glucose, glycerol and pinewood [34]– [36] State Key Lab (Xi’an Jiaotong University, China) Sawdust, rice straw, rice shell, 600–800 25 wheat stalk, peanut shell, corn stalk, corn cob, sorghum stalk, CMC (carboxymethylcellulose), cellulose and glucose 10–90 s, novel screening technique using fusedquartz capillary tubes (id ¼ 1 mm); catalysts: alkali metal and Ru/TiO2 Pilot plant, hydrothermal pretreatment, partial oxidation (H2O2), batch SS316 tubes, <20 min, fluidized bed reactors; catalysts: alkali metal, nickel and metal oxides Miniature plant and bench scale tubular reactors Sawdust, rice straw, cabbage; <600 model compounds: cellulose, xylan, lignin reagents and glucose 25 9 [37]– [39] [9], [40], [41] Research group of Antal and co-workers from University of Hawaii have investigated the wide range of biomass gasification in hydrothermal media in 1990. According to the experiments, promising conversion results as high as >85% and also suppression of tar and char forming reactions were obtained. Catalytic effects of metallic reactor and material and the use of activated carbon for catalyst are other topics investigated by this group [26]–[30]. Kruse and Dinjus at the Institut für Technische Chemie , FZK in Germany worked on wide range of biomass types by using batch and continuous tubular and mixed reactors. The starting investigation of the study points out the process conditions that maximize the hydrogen yield in product gas. Complete biomass conversion to a high hydrogen yield product gas was experimented with addition of K2CO3 and KOH alkali salts at supercritical conditions (600 ℃ and 25 MPa). Researchers have constructed a SCWG pilot plant “ Verena” in 2003, with a flow capacity of 100 kg h-1 and maximum operating temperature and pressure of 660 oC and 35 MPa, respectively [31]–[33]. The chemical engineering research group of University of Twente, Netherlands have contributed to the subject by introducing quartz capillary flow tubular reactors that can work at well beyond critical temperature of water (600 ℃). Use of a quartz reactor allowed the group to clearly identify the reaction kinetics with/without catalytic effect since it is known that metal reactor surface is prone to catalytic activity at high temperature and pressure. SCWG of glycerol, glucose and pinewood was investigated in a quartz capillary reactor, it was found that non-catalytic complete gasification of biomass is possible but for very low feed loads [34]–[36]. Matsumura has been investigating the SCWG of biomass at Hiroshima University in Japan with the collaboration of various researchers. His group’s main interests are the pretreatment of biomass to enable high biomass load feed, and lower temperature catalytic biomass conversion in hydrothermal conditions. Kato and Matsumura achieved a high concentration cabbage slurry feed by feed pretreatment processing taking place between 147–197 ℃ .The researchers have constructed a power plant which is located in the university campus of Hiroshima [37]–[39]. The research group of State Key Lab, China has been conducting studies on gasification of biomass feedstock in hydrothermal media since 1977. Their research covers a wide scope of biomass, including glucose, cellulose, sawdust, rice straw, rice shell, wheat stalk, peanut shell, corn stalk, corn cob, and sorghum stalk. Experiments run at or close to 25 MPa and 650 ℃ resulted in high biomass gasification efficiencies (> 85%) as well as high hydrogen and low CO yield [9], [40], [41]. Process simulation of the SCWG of biomass has been conducted in literature with chemical equilibrium considerations for the reacting system. Even though it is a hypothetical assumption and provides output data without time and dimension inside the reactor, such modeling work is vital for analyzing thermodynamics of the system. Fiori et al. [42] simulated the SCWG of biomass process for various feedstock, such as glycerol, microalgae, sewage sludge, grape marc and phenol. An AspenPlusTM process flow sheet was built in order to analyze the product gas 10 composition and yield at different operating conditions as well as to determine the thermal efficiency and feasibility of the process. An SCWG reactor was simulated using the ‘RGibbs’ module of the software, which applies Gibbs free energy minimization to quantify product gas composition. Process units of the study were feed pump, heat exchanger, Gibbs reactor, gas-liquid separator and gas burner. It was found, according to the simulation, that thermal sustainability of the process can be achieved if 15-25% of feed concentration is applied. The maximum hydrogen yield was found as 8,5 kg H2/100 kgfeed. The net power achieved for 1000 kgfeed/hr was 150 kWe, by utilizing hydrogen in the fuel cell and expanding gas product of the reactor. 11 1.5 Reactor Systems Biomass gasification under hydrothermal conditions have been studied since 1970’s. It should be noted that except continual efforts in researches and findings it is still defined as a novel process [7]. Optimal reactor type and reaction parameters have been investigated which will favor the gas production efficiency, overcome reactor plugging and shut down problems and enable high throughput. The largest plant in operation was built in 2003 in Germany (FZK) in order to process wet wine distillery residues with a design capacity of 100 L h-1 [25]. European subsidies and grant awarded by Japanese NEDO supported the construction of a process development unit (PDU) in Enschede, Netherlands with a design capacity of 30 L h-1, which is devoted to the Ph.D. research studies held in University of Twente. There are other smaller scale units for laboratory work of University of Hawaii, Osaka Gas, Pacific Northwest National Laboratory, Hiroshima University and TNO and BTG (in Netherlands) [7]. Batch Reactor Applications Batch autoclave reactors are used for analyzing product yield and composition, with simpler operation at supercritical temperature and pressure conditions. Main drawbacks of batch autoclaves can be stated as high heat transfer resistance (i.e. long heat up period), resulting difficulties in understanding kinetic response of the system to the temperature and uncontrollable catalytic effect due to the reactor material [7]. University of Twente has been investigating batch operations in quartz capillary reactor which shows great advantage in operation simplicity and observation abilities but difficulties in pressure measurements [7]. Continuous Reactor Applications Continuous flow reactors are of great importance in order to scale up the technology to the commercial level. Continuous stirred tank reactors (CSTR’ s) are less favored than the tubular reactors owing to mechanical challenges in mixing ability and heat insulation as well as low concentration feed requirement due to the rapid dilution to the effluent concentration (obliges bigger reactor volume than tubular reactor for the same throughput). Tubular reactors on the other hand may suffer from plugging issues due to the char and tar deposition [7], [43] Fluidized bed reactors are capable of solving char and tar related plugging issues and also high heat and mass transfer capabilities. Nature of SCWG requires more work for fluidized bed technology; currently research group in University of Twente conducted investigation on micro fluidized bed reactor as a part of NEDO Grant Research[7]. A PDU tubular reactor system was built for SCWG of biomass operation, which comprised of a continuous flow reactor, which is the most studied reactor system with straightforward operation and control. Figure 1.4 shows the general process scheme of the PDU. There are four liquid containers for continuous feed 12 sufficient for at least 2 hours of operation. A liquid feed pump pressurizes the stream and pumps it into a heat exchanger, where heat is exchanged between outlet and inlet streams. Efficient heat exchange between the fluids is vital for feasibility of the process; in this example the cold stream flows through the outer tube and the hot stream through the inner tube. Near-critical fluid is fed to the tubular reactor, which operates at 650 ℃ and 300 bar, and is heated up externally by natural gas burners. Product gas is separated from water in high and low pressure flash separators. High solubility of CO2 in supercritical water is advantageous for separation operation simplicity. Figure 1.4 Simplified process scheme of the PDU for SCWG of biomass operated at University of Twente [7] It is noteworthy that, for a successful reactor design with high gasification and thermal efficiency and low char/tar formation, reaction kinetic data of hydrothermal gasification of biomass (in the whole range of operation) needs to be collected; by which favored operation conditions can be realized for desired reaction pathway and products. 13 1.6 Scope and Objectives of the Study Biochemical compounds in biomass undergo various decomposition reactions in hydrothermal media. There have been several experimental and thermodynamic modeling works on SCWG ranging between simple model compounds to real biomass samples. It has been found that SCWG is a promising technology for producing high calorific value gas from wet biomass, with less tar and char formation owing to the high solubility of solid biomass in hot and compressed water and rapid hydrolysis reactions [7]. Furthermore, catalytic effects of water involving ionic mechanisms have been reported especially in near-critical region at which water ion product boosts. For product distribution, design and optimization of a reactor, kinetic approach to the process is necessary by which rate laws and kinetic parameters of the biomass decomposition pathway can be quantified. There are few kinetic studies for SCWG [44]–[47], mostly focused on ultimate results such as gasification efficiency and gas yields. Reaction paths including the intermediates (source of gas products) and direct pyrolytic gasification reactions have not been studied together; hence, limited knowledge is present on rates of reactions constituting the biomass degradation reaction network. Kinetic and process data in the literature are available for specific model compounds and for low flowrates due to the time and space limitations of the experimental set-ups. Process realization, on the other hand, requires the analysis of real biomass slurries with higher feed flow-rate and in a broad range of operating parameters, combined with possible variations in the reactor design specifics. Computer aided simulation is an efficient tool for multipurpose evaluation of the process parameters and design specifications (e.g. operating pressure and temperature, residence time, biomass type, feed composition etc.). Kinetic approach and simulation require careful review and collection of the biomass hydrothermal decomposition data in the whole relevant operation range, enabling the instantaneous composition analysis inside the reactor. In this set of work, a reaction kinetics approach to the “SCWG of biomass” process was applied, with the objective of predicting distribution of key compounds from the initial heating up of the feed stream till the end product. Main goal of the work is to elucidate the main decomposition and gasification reaction pathways of complete biomass hydrocarbons and to analyze the influence of operational condition changes on gasification and thermal efficiency of the process. Possible solutions for the ongoing process challenges are aimed at by analyzing the instantaneous kinetic response of the reacting system to the tuned variables. 14 Research questions to be answered: Are the reported experimental kinetic data for individual model biomass compounds capable of predicting the real lignocellulosic biomass (mainly agricultural residues) behavior? Is AspenPlusTM software capable of simulating the SCWG of model compound combinations and deliver valid results based on the reaction kinetics approach? What can be the possible process conditions to favor high gas yields and low tar/char formation yields? Is biomass gasification based on the SCWG process feasible with reasonable yields and thermal self-sufficiency? 15 2 Literature Review of Biomass Degradation in Hydrothermal Media: Reaction Mechanism and Kinetics Early studies were performed by Antal, on standard gasification of lignocellulosic biomass. A detailed reaction kinetics pathway approach to the dry gasification of biomass is challenging, due to the complexity of the reacting conditions[4]. Dry gasification of biomass takes place heterogeneously and mostly surface enhanced, adding to that it has complex structure consisting of cellulose, hemicellulose and lignin [4]. Nature of the classical gasification prevents to define the degradation of biomass reaction pathway with precise single reaction steps. Biomass conversion in hot and compressed water is based on a significantly different reaction mechanism than air blown gasification, by enabling water to be a reaction medium and a solvent. Crystal cellulose (initially) degrades to sugar rapidly in hot and compressed water via a hydrolysis pathway instead of pyrolysis[4]. Hydrolysis products generally show high solubility in hot-compressed water, makes the medium a homogenous, one phase solution. Biomass gasification in sub- and supercritical water, from solid biomass to the product contains very high number of grifted reactions. Pathways through key decomposition compounds would enable the kinetic modeling of the complex reaction network. Main intermediate compounds of the biomass conversion in sub- and supercritical condition are aldehydes, ketones, furfurals and organic acids[4]. In this chapter, kinetic studies and reaction pathway proposals of various science groups are reviewed based on the three main biopolymer compounds. Decomposition reactions and mechanisms, main products, compositional data and proposed reaction networks for cellulose, hemicellulose and lignin are described in Chapters 2.1, 2.2, 2.3, respectively. 16 2.1 Cellulose Decomposition Cellulose is the most abundant constituent of the lignocellulosic biomass sources. Initially, cellulose decomposition with acid catalyst or enzymes was studied and hydrolysis to simple saccharides namely glucose and fructose was achieved[48][49]. The use of catalysts is found useful for increasing the solubility and reactivity of cellulose by breaking the inter-intramolecular hydrogen bonds [50]. Hot and pressurized water changes its physical properties significantly, namely ionic product, density and viscosity which enables the cellulose decomposition reaction to take place without a catalyst. Cellulose (and glucose) decomposition in hydrothermal media follows several organic reaction pathways, such as hydrolysis, retro-aldol condensation, keto - anol tautomerism and dehydration [22], [23], [51], [52]. Kabyemela et al. studied the reaction kinetics and primary products of cellobiose decomposition. A biomass-water mixture with a 16 g/min flow rate was heated up and pressurized rapidly to the near- and supercritical condition (25-30 MPa, 300 – 400 ℃ and fed to a stainless steel reactor. Residence time was kept as low as 0,04 to 2s in order to analyze the primary products of cellobiose decomposition. Product concentrations obtained revealed that cellobiose conversion reaches 80% in 1,4 s at 350 ℃; and the same degree of conversion is achieved in 0,8 s at 400 ℃ . Glycosylerythrose (GE) and glycosylglycolaldehyde (GG) compositions are nearly constant; glucose and fructose compositions reach a maximum before converting to other product. Erythrose composition reaches the maximum with residence time. Cellobiose undergoes conversion by hydrolysis of glycosic bond and pyrolysis of the reducing end resulting in GE+GG and glucose + fructose, respectively [51]. Kabyemela proposes a reaction network for cellobiose decomposition thanks to the product distribution (see Figure 2.1). Notation kH refers to the hydrolysis pathway, k1 and k2 stand for pyrolysis reactions. Figure 2.1 Cellobiose decomposition pathway: 𝟑𝟎𝟎 ℃ < 𝑻 < 𝟒𝟎𝟎 ℃, 𝟐𝟓 < 𝑷 < 𝟒𝟎 𝑴𝑷𝒂 [51] 17 The linear relation of –ln(1-X) versus residence time, with X being the cellobiose conversion, shows that reaction behavior follows a first order decomposition reaction rate (see Figure 2.2). Figure 2.2 -ln(1 - X) versus residence time for cellobiose decomposition [51] The reaction type of cellobiose depolymerization changes from heterogeneous to homogenous when near critical conditions are reached via dissolution of cellulose in water [50]. Figure 2.3 shows the temperature relation of the overall cellulose decomposition rate constant in between 290 and 400 ℃ and 25 MPa. 18 Figure 2.3 Arrhenius plot of the rate constant of conversion of microcrystalline cellulose in water (290 C – 400 oC, 25 MPa), [50]. A surface enhanced reaction model (see Equation 2-1) was proposed by Sasaki et al. in order to explain the reactivity of microcrystalline cellulose in suband supercritical water[50]. 𝑑𝑉(𝑋) = −𝑘𝑠 . 𝑆(𝑋) 𝑑𝑡 Equation 2-1 where 𝑘𝑠 is surface reaction-rate constant (m.s-1), 𝑉 and 𝑆 are volume (m3) and surface area (m2) of the particle, respectively. Reaction rate calculations showed that conversion of cellulose in supercritical water is faster than in sub-critical water. The main reason of the change in kinetic parameters can be given as, from sub- to supercritical water surface driven reaction switches to homogenous reaction due to the enhanced swelling and dissolution of biomass. Hydrolysis is the favored mechanism in subcritical water while in supercritical water radical pyrolytic decomposition reaction is favored with high temperature and the mixture being homogenous. Kinetic parameters of overall cellobiose decomposition are Ea=145,9 and A=1011,9 s-1; Ea= 547,9 kJ.mol-1 and A=1044,6 in sub- and supercritical water, respectively. 19 2.1.1 Glucose Decomposition Early glucose decomposition kinetic studies in hot compressed water show that, glucose isomerization to fructose, glucose decomposition and fructose decomposition are the main reactions. Glucose epimerizes to the fructose, dehydrates to the anhydroglucose or decomposes to the intermediates by breaking C-C bonds under hydrothermal conditions. 1,6 anhydroglucose, dihydroxyacetone, pyruvaldehyde and small amount of 5-HMF are the decomposition products, detected in the liquid effluent [23] Furthermore, kinetic studies have been conducted by researches on glucose decomposition in sub- and supercritical water in order to obtain a complete reaction scheme. According to the Kabyemala et al., glucose and fructose reacts to form several organic compounds such as aldehydes, ketones and organic acids in hydrothermal conditions (573-673K and 25-40 Mpa) [53]. Glucose also isomerizes with fructose. Decomposition occurs mainly via Lobry de Bruyn Alberda van Ekenstein (LBAE) transformation, retro-aldol condensation, bond cleavage by double bond rule and dehydration reactions [53]. In Figure 2.4 it is shown that glucose ring structure opens up and forms intermediate aldehyde, ketone or enediol by LBAE transformation. Retro-aldol condensation products of glucose are glyceraldehyde and erythrose. Bond cleavage with double bond rule forms glycolaldehyde and also glyceraldehyde. Glucose also forms anhydroglucose by dehydration reaction. Glyceraldehyde undergoes reversible isomerization reaction with dihydroxyacetone, both of which dehydrates to form pyruvaldehyde [22]. The following step for the intermediate compounds formed is the hydrolysis to the organic acids [54]. 20 Figure 2.4 Mechanism of glucose decomposition in sub- and supercritical water: 𝟑𝟎𝟎 ℃ < 𝑻 < 𝟒𝟎𝟎 ℃, < 𝑷 < 𝟑𝟎 𝑴𝑷𝒂 [53] With the knowledge of main liquid products of hydrothermal glucose decomposition, possible char, furfural and 5-HMF formation reactions were investigated by Promdej and Matsumura, in between 300 ℃ – 500 ℃ and 25 MPa [3]. According to the lumped approach to the reaction network, main decomposition products of glucose and fructose were classified as 5-HMV, furfural, char, gas and the rest of water-soluble organic compounds (TOC). The effect of temperature on glucose conversion in sub- and supercritical water was analyzed with the aim of finding out not only the reaction network but also the mechanisms. Tar and char formation, among the main challenges of SCWG of biomass technology can be suppressed by tuning the water properties with the knowledge of compound favoring conditions. It was found that there are both ionic and radical reactions taking place during the glucose and fructose decomposition. Ionic type reactions were identified according to the nonArrhenius behavior in supercritical water due to the rapid decrease in density and ion product of water. Water below the supercritical region more efficiently dissociates into H3O+ and OH- ions, resulting in ion product increase 103-4 orders of magnitude (compared with normalized condition). Ionic reaction rates are favored by (increasing) temperature until the supercritical region, beyond which temperature has inverse effect on rate due to decrease in density (gas like 21 behavior). For radical reactions, obedience to the Arrhenius behavior is observed in sub- and supercritical water. Promdej and Matsumura suggested an ionic mechanism for 5-HMF, furfural and char formation and also fructose to TOC decomposition; on the other hand radical mechanism reactions pointed out were glucose, 5-HMF and furfural to TOC decomposition reactions in addition to the radical gasification reactions [3]. Char formation from glucose decomposition was also linked to the ionic reactions, unlike proposed radical reactions of lignin to char, which is discussed specifically in Chapter 2.3. According to their set of experiments, Promdej and Matsumura suggested order reaction for all glucose decomposition pathway reactions. Figure 2.5 displays the pathway of glucose conversion in hydrothermal conditions. 1st Figure 2.5 Glucose decomposition pathway in sub- and supercritical water: 𝟑𝟎𝟎 ℃ < 𝑻 < 𝟒𝟔𝟎 ℃, 𝑷 = 𝟐𝟓 𝑴𝑷𝒂 [3]. 22 2.2 Hemicellulose Decomposition Being one of the key components of the biochemical structure of biomass, hemicellulose and its degradation products in sub- and supercritical conditions is of great importance. D-Xylose, 5-carbon sugar is a main hydrolysis product of hemicellulose. D-xylose, as a monomer unit of hemicellulose has been investigated widely in order to understand the reaction kinetics of hemicellulose in sub- and supercritical water. According to the Sasaki’ s experiments with a flow type reactor, D-xylose decomposes via reverse aldol condensation in suband supercritical water (633 K-693 K, 25-40 MPa) [55]. Kinetics of xylose and its primary decomposition product furfural under hydrothermal conditions were investigated by Qi et al.[56] for temperatures between 180 ℃ and 220 ℃ and pressures below 10 MPa. Product composition distribution was reported in order to elucidate the decomposition network. It was found that xylose primarily undergoes decomposition to furfural (k1) and hydrolysis to undefined decomposition products, DP1 (k2). Furfural reacts in hydrothermal media to produce other products, DP2 (k3) (see Figure 2.6). Figure 2.6 D-xylose decomposition in hot compressed water: 𝟏𝟖𝟎℃ < 𝑻 < 𝟐𝟐𝟎℃, 𝑷 = 𝟏𝟎 𝑴𝑷𝒂 [56] For the overall reaction network elucidation, definition of DP1 and DP2 were required which would enable the modeling of the complete system. In literature, decomposition product definitions were not found for xylose in subcritical water. It was accepted that in supercritical water xylose and furfural decompose to form similar products but with possible different rate parameters. Goodwin et al. [57] studied the kinetic model of xylose decomposition for both analysis of intermediate DP’ s and simplified gasification pathway, for temperatures between 450 ℃ and 650 ℃ under 250 bars. According to the kinetic study of concern, xylose decomposition takes place in two directions; to furfural and to glyceraldehyde, methyl formate mixture. Furfural may react in two ways, to produce either water soluble humic solution (WSHS) or to maple lactone (C6H8O2), as shown in Figure 2.7. In other kinetic study of Goodwin and colleagues, WSHS as a decomposition product of furfural was analyzed to be a mixture of acetic acid and acrylic acid [58]. 23 Figure 2.7 Kinetic model reaction mechanism for decomposition of xylose in supercritical water 𝟒𝟓𝟎℃ < 𝑻 < 𝟔𝟓𝟎℃, 𝑷 = 𝟐𝟓 𝑴𝑷𝒂 [57] According to the kinetic study of the Goodwin, for high temperatures (450 ℃- 650 ℃) radical gasification reactions are favored and more precise product compositions were achieved with a simplified gasification pathway due to the . According to the proposed network, xylose either dehydrates to furfural or decomposes to WSHS. Furfural may undergo further decomposition to WSHS, which is the intermediate between liquid organics and gasification products, CO and H2 (see Figure 2.8). Figure 2.8 D-xylose decomposition and gasification path in supercritical water: Kinetic model reaction mechanism for decomposition of xylose in supercritical water 𝟒𝟓𝟎℃ < 𝑻 < 𝟔𝟓𝟎℃, 𝑷 = 𝟐𝟓 𝑴𝑷𝒂 [57] 24 2.3 Lignin Decomposition Lignin, one of the most abundant biochemical units of biomass with polymeric aromatic organic substance with oxygen-based functional substituents (such as hydroxyl and methyl groups), has been investigated for its kinetic behavior in sub- and supercritical water. Kanetake et al. [59] used guaiacol as a model compound of lignin, owing to its similar polymeric structure and functional units. Kinetic studies were performed for temperatures between 380 ℃ and 400 ℃ under various pressures for long residence times up to 90 minutes. GC-MS chromatogram peaks of the liquid effluent revealed that main decomposition products of lignin are catechol, phenol and o-cresol; additionally, some gaseous products were detected but not quantified (see Figure 2.9) Figure 2.9 GC-MS chromatogram for soluble products of guaiacol decomposition in supercritical water [59] Kinetic modeling was carried out by accepting 1st order decomposition reactions even though product compositions did not follow precisely, which were reasoned with direct gasification of guaiacol under supercritical conditions [49]. A reaction mechanism was suggested that dissolved compounds containing ether linkages form single ring phenolic compounds (catechol, phenol and ocresol) which may further re-polymerize to form high molecular weight compounds. Char residues after the supercritical water treatment were analyzed was and these were explained to be the dehydration mechanism product of low25 molecular phenolic compounds, monomers and oligomers [49]. The resulting proposed reaction mechanism of Kanetake and colleagues is displayed in Figure 2.10. Figure 2.10 Simple reaction pathway for guaiacol (lignin model component) in supercritical water [59] Catechol reaction kinetics and mechanism in sub- and supercritical water was further investigated by Wahyudiono et al. [60] for temperatures between 370 ℃ to 420 ℃ and residence times up to 240 minutes. Decomposition reaction rate was found to be extremely slow due to the strongly bound heteroatoms and carbon atoms even in the presence of supercritical water. Phenol formation from catechol mechanism was explained as follows; initiation of the reaction by scission of one of the O-H bonds and formation of phenoxy radical, further attack of the released H to the radical, resulting in phenol and OH- ion (see Figure 2.11). 26 Figure 2.11 Simple reaction pathway of catechol decomposition in near- and supercritical water [60] Yong et al. [61] investigated the reaction network and kinetic parameters of the guaiacol depolymerization in sub- and supercritical water for between 300 and 450 ℃ at a pressure of 25 MPa and for a very short residence time in between 0.5 – 40 s. According to the analysis of the flow reactor outlet stream the main decomposition products were found to be the single ring phenolic (phenol, catechol, cresol) and benzene structured aromatics in addition to gas and char[61]. Kinetic parameter calculations showed that overall guaiacol decomposition obeys the Arrhenius law, while in the network deviations from Arrhenius law was observed for some reactions (especially under supercritical conditions), which suggested the presence of different mechanisms in the network[61]. According to the yield analysis, formation reactions of some guaiacol degradation compounds were favored in subcritical condition while the rest was rather dominant in the supercritical water [61]. Water properties, such as dielectric constant (𝜀) and ion product (𝐾𝑤 ) deviate significantly from the ambient water when sub and near-critical conditions are achieved; 𝜀 decrease (≈ 10) and 𝐾𝑤 increase (≈10-10) favor ionic reactions and result in high rate constants. Beyond critical point at elevated temperatures, suppression of those mechanisms due to gas phase behavior of water can be observed. In their set of experiments, Yong et al. observed that guaiacol decomposition increases with temperature, especially in the supercritical region and it goes to near complete conversion within 40 s, indicating the radical type depolymerization of guaiacol, by chemical degradation of the aromatic ring substituent[61]. Gaseous product detection for short residence times especially in near- and supercritical conditions suggested the direct gasification of guaiacol and liquid products via radical reactions[61]. 27 Figure 2.12 Catechol formation from guaiacol in sub- and supercritical water [61] Complex reaction mechanisms of the phenolic and benzene products were studied [61], according to which catechol was found to be formed from guaiacol via both radical and ionic mechanisms as shown in Figure 2.12. Benzene formation reaction from guaiacol was dominant in the near-critical region, while it was not formed under supercritical conditions, pointing out the likely mechanism to be ionic. The high yield of phenol in the subcritical region indicated the presence of ionic reactions even though mechanism was found to be through the formation of phenolic radicals (see Figure 2.13). Hence, further kinetic parameters for the aromatic compound reactions were quantified but for some reactions (including the formation of phenolic compounds) mechanism propositions were left for future studies [61]. 28 Figure 2.13 Phenol formation from guaiacol and catechol in sub- and supercritical water [61] Formation of compounds composed of multiple benzene rings, abundantly found in the supercritical region was a fruitful information for understanding the indication of cross-linking between the single ring compounds [61]. Especially in the supercritical water high molecular weight fragments were formed as a result of the cross-linking reactions between the degradation fragments (radicals) [61], which are prone to combine further to produce char particles. An overall reaction pathway was elucidated from guaiacol to the aromatic degradation products, inter-conversion of them and formation of gas and char products[61]. The complex reaction network resulting (see Figure 2.14) suggests that guaiacol depolymerizes to catechol, cresol, phenol and benzene; catechol is only formed from guaiacol and converts to cresol and phenol; cresol is a hydrothermal degradation product of guaiacol and catechol, which may further convert to TOC; benzene is produced from guaiacol and TOC. In addition to that, char formation is sourced by guaiacol, phenol, TOC and benzene while gaseous products are formed from either guaiacol or TOC. 29 Figure 2.14 Reaction pathway of guaiacol under hydrothermal conditions: 𝟑𝟎𝟎℃ < 𝑻 < 𝟒𝟓𝟎℃, 𝑷 = 𝟐𝟓 𝑴𝑷𝒂 [61] Benzene and phenol are the essential decomposition products of lignin under hydrothermal conditions. Reaction network and kinetics of the benzene and phenol decomposition in SCW was investigated in the kinetic work of Yong et al. [39] for temperatures between 370 – 450 ℃ under 250 bar pressure within a short residence time between 0.5-100 s. According to the proposed network, gas production was sourced by phenol, benzene and TOC and char is formed by phenol, benzene and naphthalene polymerization reactions (see Figure 2.15). Figure 2.15 Reaction scheme of phenol benzene decomposition in supercritical water: 𝟑𝟎𝟎℃ < 𝑻 < 𝟒𝟓𝟎℃, 𝑷 = 𝟐𝟓 𝑴𝑷𝒂 [39] 30 3 Proposed Reaction Pathway Kinetics of the hydrothermal decomposition of biomass constituents has been studied based on experiments by several research groups as mentioned in Chapter 2. Most of these works; on the other hand, were targeting to clarify kinetics of a specific constituent (cellulose, hemicellulose or lignin) in a limited temperature range; according to their own scientific interest and missing works in the literature. For the application of the SCW gasification process simulation, with real biomass (composed of cellulose, hemicellulose and lignin) feed, integration of the most reasonable pathways with available kinetic parameters was the scope of the current work. By proposing an integrated decomposition pathway and deriving kinetic parameters from published reaction rates, this set of work enables the compositional analysis of the degradation products of a biomass with the flexibility of different feedstock and thermal condition. Integration of the kinetic works was established by choosing the sequential study of the same research groups for given model compounds, if available, otherwise pathways were linked only through the end products in order to preserve the liability of the kinetics data achieved according to the compositional distribution over a set network. Additionally, experimental conditions under which the data gathered were worked carefully by grouping the data according to the severity of the operating conditions. For reaction pathways, reported to show difference between sub – and supercritical water, grouping was made in order to identify better the decomposition products in specified operating conditions. In addition to that kinetic parameters of several reactions, with deviations from Arrhenius parameters in supercritical water were handled separately by considering the possibility of application to the simulation and extent of influence on the outlet compositions. Rapid decomposition reactions of the inlet compounds, reported to achieve complete conversion in a matter of seconds in subcritical water during the average residence time spent for the heating up the stream. On the other hand, for rather slowly reacting compounds, such as lignin derivatives, separate pathways were proposed for sub- and supercritical water in order to minimize the errors due to the influence of reaction mechanism. In spite of all, content of the study done, relies on the reported kinetic findings, which in some cases obliged to work with assumptions; details and reasoning of which are given in the following sections. Separate studies were carried out for three of the most abundant biochemical compounds in lignocellulosic biomass, cellulose, hemicellulose and lignin. Model compounds were chosen as cellobiose, xylose and guaiacol for cellulose, hemicellulose and lignin, respectively; all of which are the main decomposition products of the associated biopolymers with similar chemical structure. 31 3.1 Cellulose Pathway: 3.1.1 Decomposition of Cellobiose In the proposed reaction network, cellobiose as a model compound of cellulose undergoes to series and parallel reactions to form simpler organic compounds in hot and compressed water. Cellobiose initially depolymerizes in three essential ways; pyrolysis to glucosyl-erythrose + glycolaldehyde, glucosylglycolaldehyde + erythrose and hydrolysis to the glucose [51]. Glucose isomerizes to the fructose but fructose to glucose isomerization is not significant. Glucose under hydrothermal conditions, decomposes to erythrose, glyceraldehyde, 5-HMF and also dehydrates to 1,6 anhydroglucose. Fructose decomposition reactions produce glyceraldehyde, 5-HMF and furfural. Dihydroxyacetone takes place in 2-way isomerization reaction with glyceraldehyde. Pyruvaldehyde is a dehydration product of the isomer compounds. Even though the char formation from 5-HMF polymerization was observed in the subcritical region [3], kinetic data reported was not sufficient for the use (less than three rates were reported for different temperatures) in the model, hence excluded. The proposed pathway (see Figure 3.1) was mainly derived from the kinetic work performed for temperatures between 300 and 400 ℃, which may result in differences between the actual case and simulation for elevated temperatures, although simulations showed that cellulose to organic acids or to aromatics (i.e. furfural, 5-hmf) is completed rapidly, time required for complete conversion is shorter than the time spent in the subcritical region during the heat up period in the subcritical reactor [9], [23]. 32 Figure 3.1 Reaction scheme of cellobiose in near critical water: k1, k2, k3, kge.g, kgg.g from [51], kg.e, kg.a, kg.gly, kg.f, kg.a, kf.e, kf.gly from [53], kgly.dih, kdih.gly, kgly.p, kdih.p from [22], kg.5, kf.5, kf.fu , kfu.ch from [3] 3.1.2 Reactions of Glucose-Fructose decomposition products The proposed pathway of intermediates formed from cellobiose decomposition is mainly directed towards formation of organic acids. Decomposition from fructose, aldehydes, ketones to the organic acids was reported in the literature [53], [54], without indication of the reaction equations. Organic acid formation assumption was made according to the number of carbon atoms (that would result in 1st order decomposition reactions) and possible acid types detected in the SCWG effluent. Glycolaldehyde to acetic acid reaction was added to the network with the assumption of similar mechanism with the other aldehyde types in order to prevent substance’s accumulation in the system. Rate constants gathered for the intermediate decomposition reactions are in the same range of glucose decomposition reactions, which results in the same assumptions and error margins, except for 5-HMF decomposition. The complete reaction network is displayed in Figure 3.2 33 Figure 3.2 Intermediate decomposition scheme kf.acid, kp.acid, ka.acid, ke.acid [54]; k5.lf, k5.ff [62] 34 3.2 Hemicellulose Pathway D-Xylose was assumed to be the model compound, which is one of the main decomposition product of hemicellulose (e.g. arabinose, glucose) and widely studied compound for hydrothermal processes. D-xylose decomposition reaction pathway is displayed in Figure 3.3 and Figure 3.4 for sub – and supercritical region, respectively. The subcritical reaction pathway shows that D-xylose primarily decomposes to glyceraldehyde and methyl formate or forms furfural. Furfural further decomposes to the mixture of acetic acid and acrylic acid [56]. The Dxylose subcritical water decomposition reaction network is obtained from the research performed in the rather mild temperature and pressure condition (180 C- 220 oC, 10 MPa)[56]. For conditions near-critical region extrapolation of the kinetic data according to the Arrhenius law may result in errors. The extent of deviation due to the data collected is commented in paragraph 4.3. Figure 3.3 Decomposition scheme of D-xylose in subcritical water [56] In the supercritical water, D-xylose decomposes to the furfural or to an acetic acid and acrylic acid mixture; furfural undergoes a parallel reaction to the products of char (polymerization) and acetic acid-acrylic acid solution (WSHS). Acetic acid-acrylic acid mixture gasifies to CO and H2 beyond the supercritical point of water [57] Figure 3.4 shows the proposed pathway. 35 Figure 3.4 D-xylose decomposition and gasification scheme in supercritical water kxy.fu,kxy.wshs,kfu.wshs,kaa.ga from [57] 36 3.3 Lignin Pathway: Guaiacol was selected as model compound for the kinetic investigations of lignin due to its chemical structure and similar attached groups [61]. Guaiacol decomposition was studied separately in sub- and supercritical water since there exists sufficient literature data for detailed analysis in both regions [39], [61]. In the first reaction network, guaiacol decomposes to the single ring structures such as o-cresol and benzene or it directly gasifies in the subcritical water. Crosslinking between decomposition products produces heavier compounds, such as diphenyl (which is a model compound for TOC). Diphenyl and benzene further crosslinks the active sites of the ring structure that results in char formation. Diphenyl undergoes to a direct gasification via a parallel reaction. The reaction network of guaiacol in subcritical water is depicted in Figure 3.5. Figure 3.5 Guaiacol decomposition scheme in subcritical water [61] In supercritical water decomposition network is generated as shown in Figure 3.6. In this region, the research findings of Yong et al. in separate works [39], [61] were integrated in order to obtain a detailed reaction network for the lignin depolymerization in supercritical water. Guaiacol decomposition reactions are either ionic or radical. Radical reactions show good accordance with the Arrhenius equation even though, some kinetic data could not be fitted to the Arrhenius relation. The physical property change of water from the sub- to supercritical region is the main reason of the occurrence non-Arrhenius behavior of the ionic reactions. 37 Figure 3.6 Guaiacol decomposition scheme in supercritical water kgu.ch, khu.ga, kgu.oc, kgu.c, kgu.t, kc.oc, kt.ch from [61]; kc.t, kp.c, kp.t, kp,ga, kp.ch, kt.b, kb.t, kb.p, kb.ga, kb.na, kna.ch, kb.ch from [39] 38 3.4 Acid Decomposition Pathway: In the liquid effluent of the biomass conversion studies organic acids are detected as the secondary hydrolysis products, which are accepted to be the main gaseous product source. Acid decomposition to the gas phase mostly takes place at elevated temperatures higher than 370 ℃, below which they are almost stable. Literature was missing decomposition data of glycolic acid and levulinic acid. 2 carbon containing glycolic acid was proposed to hydrolyze to form carbondioxide and hydrogen and 5 carbon containing levulinic acid decomposes to form tertiary intermediates (lactic acid and acetaldehyde). Rest of the pathway was built according to the findings in the literature, shown in Figure 3.7 Figure 3.7 Organic acid gasification scheme in supercritical water. kfa.ga1, kfa.ga2 [63]; kaa.ga [64]; kwshs.ga, kpa.ga, kmf.aa [57]; lactic acid reactions [65] 39 Water gas shift (WGS) and methanation reactions given in Equation 3-1 and Equation 3-2 are the main gas phase reactions that take place between the gas species after they are formed from either intermediates or pyrolytic gasification of guaiacol and derivatives [47]. WGS 𝐶𝑂 + 𝐻2 𝑂 ↔ 𝐶𝑂2 + 𝐻2 Equation 3-1 Methanation 𝐶𝑂 + 3𝐻2 ↔ 𝐶𝐻4 + 𝐻2 𝑂 Equation 3-2 40 3.5 Kinetics Data and Arrhenius Parameters Cellobiose decomposition reaction rates were predicted in the experemintal study of Kabyemela [41], by assuming 1st order decompsition type rate equations and kinetic data were calculated accordingly. Kinetic data listed in Table 3.1 was in a good accordance with the Arrhenius relation (Equation 3-3) in the given temperature range. Pressure difference in the last set of parameters was accepted to have negligible significance for the reactions that take place in near-critical water; the resulting parameters are shown in Table 3.3. Table 3.1 Cellobiose decomposition rate constants Rate constant, s-1 300 ℃ 25 MPa Cellobiose 0,21 (overall) k1 0,03 k2 0,04 k3 0,14 kge.g 0,15 kgg.g 0,17 conditions 350 ℃ 25 MPa 400 ℃ 30 MPa Ref. 1,06 4,9 [51] 0,05 0,11 0,9 0,9 1,1 0,25 0,35 4,3 3,7 3,5 [51] [51] [51] [51] [51] Glucose and fructose decomposition reaction rate constants were derived by Kabyemela et al. [43] from his group’s own experimental findings as listed in Table 3.2. In the given temperature range reactions are in a good accordance with Arrhenius law. However, the ionic mechanism of fructose decomposition reactions and glucose to fructose isomerization reaction may result in deviation in supercritical water at elevated temperature, product composition results of the preliminary simulation runs show that, almost complete conversion of glucose and fructose take place in a duration less than the residence time spent in supercritical water. Hence, for the kinetic parameter calculation of this set of rate constants were done with the assumption of Arrhenius behavior, results are listed in Table 3.3. Furfural to char reaction rate data were not available, even though possible char formation was reported [3]. Kinetic rates of furfural to char reaction was assumed to be the same as phenol to char reaction (in supercritical water only), owing to the similar structure of the compounds. 41 Table 3.2 Glucose decomposition rate constants Rate constant, s-1 kg.a ke.acid ka.acid kf.gly kf.acid kf.e kg.e kg.gly kg.f kgly.dih kgly.p kdih.gly kdih.p conditions 350 ℃, 25 MPa 0,02 0,55 0,04 0,6 0,7 0,8 0,95 0,2 0,64 1,38 0,94 0,2 0,02 300 ℃, 25 MPa 0,01 0,1 0,01 0,1 0,18 0,1 0,21 0,05 0,2 0,4 0,19 0,03 0,01 400 ℃, 30 MPa 0,08 5 0,31 6,5 10,4 8 18,1 1 7 7,15 4,6 1,04 0,08 Ref. [53] [53] [53] [53] [53] [53] [3] [3] [3] [3] [3] [3] [3] Temperature dependence of the reaction rates was expressed with the well-known Arrhenius equation unless it was found to deviate from Arrhenius behavior; for simulating purposes Activation energy (Ea) and pre-exponential factor (A) kinetic parameters were derived. −𝐸𝑎 Equation 3-3 −𝐸𝑎 1 ∙ + ln(𝐴) 𝑅 𝑇 Equation 3-4 𝑘 = 𝐴 ∙ 𝑒 (𝑅𝑇) ln(𝑘) = Reaction rate constant dependence on temperature was tested with checking the linearity between the natural logarithm of reaction rate constant (ln k) and reciprocal temperature (1/T). The calculation method is illustrated in Figure 3.8, and resulting parameters including the one shown below are listed in Table 3.3. Cellobiose and glucose decomposition kinetic parameters were also investigated by Cantero et al. [54] by using the same rate constants as found in the literature [51], [53], [3]. Calculated values and Cantero’s values were in complete agreement, which pointed out the precision of the calculations. Glycolaldehyde to acid decomposition rate parameters were not found in the literature. Reported parameters were assumed to be the same as erythrose to acid decomposition reaction due to the similar bonding types. Arrhenius parameters of acid formation reactions were directly obtained from the literature [54] data as presented in Table 3.3. 42 1000/T(K) vs. ln(k) 0 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8 -0.5 ln (k) -1 -1.5 y = -11.51x + 17.016 R² = 0.9921 -2 -2.5 -3 -3.5 1000/T(K) Figure 3.8 Arrhenius relation of glucose to glyceraldehyde decomposition reaction Table 3.3Arrhenius Parameters of cellobiose and glucose decomposition Symbol 1 2 3 ge.g gg.g g.a f.gly f.e g.e g.gly g.f gly.dih gly.p dih.gly dih.p fu.ch e.acid a.acid p.acid f.acid glyo.acid Ea (kJ/mol) 66,89 69,30 108,60 106,10 110,50 65,93 133,04 140,44 141,34 95,69 112,69 154,36 82,56 77,32 88,65 87,90 124,72 109,29 94,00 128,63 124,72 A, s-1 3,1E04 7,8E04 1,4E09 3,6E44 1,4E08 9,0E03 1,2E11 5,3E11 1,2E12 2,5E07 3,0E09 1,5E13 7,6E06 4,6E05 1,8E07 1,4E04 2,1E10 8,0E07 6,59E07 7,4E10 2,1E10 Ref. [54] [54] [54] [54] 43 5.lf 5.ff 95,60 114,8 8,0E07 7,1E09 Promdej and Matsumura [3] studied the temperature dependence of glucose decomposition reactions in sub- and supercritical water, computed rate constants according to the compositional data of the experiment for temperatures between 300 ℃ and 460 ℃. In the kinetic simulation of the current work, 5-HMF formation reaction rate constants were used as displayed in Table 3.4. Table 3.4 Glucose decomposition rate constants continued [3] Rate constant, s-1 kg.5 kf.5 300 ℃, 25 MPa 350 ℃, 25 MPa conditions 400 ℃, 30 MPa 425 ℃, 450℃, 25 MPa 25 MPa 460 ℃, 30 MPa 6,0E-03 1,620E01 3,27E-02 2,15E-01 0,3 0,132 0,15 3,30E-01 3,82E-01 0,118 0,0932 0,087 In supercritical water, reaction rate constants of glucose and fructose dehydration reactions to form 5-HMF do not obey Arrhenius law due to the change in water properties as can be followed from Table 3.4. Kinetic parameters of 5-HMF formation from glucose and fructose were calculated for sub- and nearcritical region only, by expecting the rapid complete conversion in that region, as reasoned in paragraph 2.1.1. Resulting Arrhenius parameters applied to the model as summarized in Table 3.5 Table 3.5 Arrhenius Parameters of glucose decomposition continued Symbol g.5 f.5 Ea, kJ/mol 114,4 42,3 A, s-1 1,5E08 1,2E03 Xylose decomposition pathway and resulting rate constants were handled separately in sub- and supercritical water, owing to the availability of the literature data. In sub-critical water the main decomposition products of xylose are parallel reaction products, furfural and methyl formate-glyceraldehyde solution, for temperatures between 180 and 220 ℃ and under pressure of 10 MPa [56]. 44 Table 3.6 Xylose decomposition rate constants in subcritical water Rate constant, s-1 koverall kxy.fu kxy.gm kfu.aa 180 ℃ , 190 ℃, 10 MPa 10 MPa conditions 200 ℃ , 210 ℃ , 220 ℃ , 10MPa 10 MPa 10MPa Ref. 0,00450 0,00322 0,00128 3,28E-4 0,02420 0,0119 0,0123 6,47E-4 0,00960 0,00583 0,00377 4,57E-4 0,03420 0,0192 0,0150 7,93E-4 0,0653 0,0356 0,0297 1,22E-3 [56] [56] [56] [56] Kinetic parameters of the xylose decomposition reactions in subcritical water are tabulated in Table 3.7, according to which, D-xylose stays almost stable under the subcritical conditions. Table 3.7 Arrhenius parameters of d-xylose decomposition subcritical water Reaction Type kxy.fu kxy.gm kfu.aa Ea, kJ/mol 76,6 153,8 24,2 A, s-1 1230 7,38E08 2,95E-14 In supercritical water for elevated temperatures, the simplified gasification pathway of xylose was found to be more precise than detailed decomposition pathway in terms of product compositions, according to which rate constants and Arrhenius parameters were derived by Goodwin and colleagues [57] (see Table 3.8). Table 3.8 Arrhenius parameter of d-xylose decomposition supercritical water Symbol Overall xy.wshs fur.wshs wshs.ga Ea, kj/mol 147,5 154,7 100,5 142,7 A, s-1 1,3E13 6,6E14 1,7E6 3,5E8 Ref. [57] [57] [57] [57] Guaiacol depolymerization pathway was elucidated and 1st order reaction rate constants (see 45 Table 3.9) were fitted to the composition results of the experiment, performed by Yong et al. [61] for temperatures between 300 and 450 ℃ at a pressure of 25 MPa. Table 3.9 Lignin decomposition rate constants sub- and supercritical water [61] Rate Constant, s-1 kgu.t kgu.ga kgu.ch kgu.b kgu.c kgu.oc kc.t kt.p kp.ch koc.t kt.ch kt.b kt.ga kb.ch 300 ℃ 2 MPa conditions 350 ℃ , 370 ℃ , 390 ℃ , 420 ℃ , 450 ℃ , 250 MPa 250 MPa 250 MPa 250 MPa 250 MPa 9,54E-02 2,45E-05 0 1,92E-04 2,06E-04 8,52E-05 3,49E-02 6,76E-04 0,00E+00 1,21E-04 2,36E-05 5,09E-05 1,84E-03 1,34E-05 9,58E-02 7,74E-04 0 3,30E-04 9,01E-04 9,32E-02 4,38E-02 1,39E-03 0,00E+00 3,41E-03 1,91E-04 4,45E-04 2,50E-03 7,28E-03 1,28E-01 9,19E-04 0 1,85E-03 1,31E-03 1,37E-04 4,40E-02 1,54E-03 0,00E+00 3,99E-03 1,61E-04 2,31E-04 4,73E-05 7,81E-03 1,35E-01 3,62E-03 2,01E-04 0,00E+00 2,60E-03 1,01E-02 0,00E+00 1,26E-03 8,41E-04 1,38E-02 1,36E-03 0,00E+00 3,40E-03 0,00E+00 1,85E-01 8,51E-03 8,64E-03 0,00E+00 9,07E-02 1,94E-02 0,00E+00 9,41E-04 8,99E-04 2,60E-03 5,66E-03 0,00E+00 4,25E-03 0,00E+00 3,15E-01 1,55E-02 1,08E-02 0,00E+00 2,53E-02 3,54E-02 0,00E+00 8,84E-04 1,00E-03 2,50E-03 7,05E-03 0,00E+00 6,02E-03 0,00E+00 Guaiacol to o-cresol and guaiacol to catechol reaction rates show a decrease with temperature beyond the critical point. Slow reacting lignin derivatives, were expected to be present in the supercritical water and presence of these reactions in the set is considered to be important; therefore, average value of the reaction rates (that deviate from Arrhenius) were calculated and integrated to the model with negligible activation energy. Hence, neither a favoring nor a suppressing temperature effect on these specific reactions was assumed in supercritical water, with the purpose of favoring radical type of reactions over the ionic ones in the elevated temperatures. The extent of the error margin due to the non-suppression of the temperature increase for ionic reactions will be evaluated in the discussion section. Resulting kinetic parameters can be viewed in Table 3.10 and 46 Table 3.11. Table 3.10 Arrhenius Parameters of guaiacol decomposition subcritical water Symbol gu.t gu.ga gu.b gu.c gu.oc c.t c.oc t.p oc.t t.ch t.b t.ga b.ch Ea, kJ/mol 48,28 168,75 84,64 82,43 0 10,87 8,48 37,51 162,86 92,77 79,79 18,20 299,45 A, s-1 1067,1 6,7E+10 8284,2 6833,6 3,11E-02 0,3 0,005 1,8 9,5E=10 7508,6 1129,9 0,08 3,4E+22 Table 3.11 Arrhenius Parameters of guaiacol decomposition supercritical water Symbols gu.t gu.ga gu.ch gu.c gu.oc c.oc t.p p.ch oc.t t.ch Ea, kJ/mol 56,1 96,8 267,9 0 83,3 18,5 0 11,5 0 110,4 A, s-1 3377,9 158102,7 4,4E+17 3,9E-02 37049,1 0,06 1,0E-03 0,007 6,3E-03 815046,1 47 t.ga 37,8 3,3 Main guaiacol derivatives phenol and benzene were further analyzed for decomposition reaction pathway in supercritical water for temperatures between 370 and 450 ℃ under pressure of 25MPa, and pathway rate constants were found accordingly (See Table 3.12) [39] Table 3.12 Phenol- benzene decomposition reaction rate constants supercritical [39] Reaction Rate, s-1 kb.p kb.na kb.ga kb.ch kb.t kp.ga kp.t kp.c kp.ch kc.t kna.ch kt.ga 370 ℃, 25 MPa conditions 400 ℃, 25 MPa 450 ℃, 250 MPa 3,07E-04 9,04E-04 2,05E-04 3,86E-04 2,99E-04 7,03E-04 2,43E-02 0,00E+00 7,86E-04 0,00E+00 1,90E-03 1,77E-02 4,52E-04 1,78E-03 3,50E-04 7,60E-04 3,15E-04 1,69E-03 2,93E-02 1,67E-04 3,14E-03 8,70E-03 2,35E-03 1,90E-02 1,01E-03 3,83E-03 4,24E-04 1,44E-03 8,18E-04 2,52E-03 3,92E-02 5,03E-04 5,24E-03 1,57E-02 4,05E-03 1,92E-02 It was reported that ionic and radical reactions are likely to take place together for the phenol-benzene decomposition system [39]. Even though Arrhenius behavior was followed in the most cases, very low pre-exponentials indicate the suppression of the ionic reactions (e.g. dehydration, hydrolysis etc.) of the derivative compounds in supercritical water, which are likely to be present in the outlet stream of the simulation. Table 3.13 Arrhenius parameter of phenol-benzene decomposition in supercritical water Symbol b.p b.na b.ga b.ch Ea, kJ/mol 15,6 391,2 0,1 50,9 A, s-1 58,1 69,2 33,6 62,7 48 b.t p.ga p.t p.c p.ch c.t na.ch t.ga 3,5 54,8 1,8 1407,1 13718,5 44,4 1,9 27,7 50,8 59,5 23,1 89,2 87,9 47,8 37,2 3,7 Formic acid gasification kinetic parameters for 2 parallel reactions were calculated from the experiment findings of Yu et al. [63] for temperatures between 320 and 400 ℃, the results given in Table 3.14, formic acid is almost stable in subcritical water, even though rapid gasification reactions occur in supercritical water at elevated temperatures. Table 3.14 Arrhenius parameters of formic acid decomposition [63] conditions subcritical supercritical A Eact, kJ/mol A Eact, kJ/mol f.ga1 1,6E06 85,01 4,8E12 168,20 f.ga2 4,6E-05 NA 3,40E17 2,44E02 Arrhenius parameters of the acetic acid and propionic acid decomposition reactions in supercritical water were reported, hence they were directly taken from the literature [57], [64] (See Table 3.15). Acetic acid and propionic acid reactions are favored in the elevated temperature region in the supercritical water (𝑇 > 400 ℃), in subcritical region (for short residence time) they were accepted as stable compounds. Table 3.15 Arrhenius parameters of acetic and propionic acid decomposition Ea, kJ/mol Acetic Acid (aa.ga) 94 Propionic Acid (pa.ga) 89,4 A, s-1 2,5E+4 1,4E+5 Ref. [51] [61] Lactic acid decomposition rate data and reaction pathway was studied in near- and supercritical water for temperatures between 350 and 420 ℃ under 25 MPa pressure [65], kinetic data for the proposed reaction pathway is tabulated in Table 3.16 49 Table 3.16 Lactic acid decomposition reaction rate constants [65] Reaction Rate, 350 ℃, s-1 25 MPa Global 1,80E-03 koverall 1,10E-03 kla.aa 2,00E-04 kaa.la 9,40E-04 kla.aal 9,40E-02 kaal.ac 9,80E-04 kaa.hpa 5,00E-03 khpa.aa 7,99E-04 khpa.ga 2,43E-03 conditions 385 25 MPa 5,00E-03 2,79E-03 3,76E-04 4,22E-03 1,56E-03 1,18E-03 1,83E-03 8,98E-04 1,40E-03 Ref. ℃ 420 25 MPa 1,04E-02 4,17E-03 3,95E-04 1,68E-02 2,07E-03 4,78E-03 7,26E-02 1,59E-03 9,30E-05 ℃ [65] [65] [65] [65] [65] [65] [65] [65] [65] Overall lactic acid decomposition obeys the Arrhenius behavior in nearand supercritical water, even though formation reactions of acetic acid from acetaldehyde, acrylic acid from 3-HPA and glycolic acid-methanol solution from 3-HPA deviated from the Arrhenius relation. Average rate constants in the given temperature range for the indicated reactions were calculated by assuming negligible Activation energy, with the reasoning given in the guaiacol paragraph. Resulting kinetic parameters are displayed in Table 3.17. 50 Table 3.17 Arrhenius parameters of lactic acid decomposition supercritical water Symbols overall la.aa aa.la la.aal aal.ac aa.hpa hpa.aa aa.pa hpa.gly Ea, kJ/mol 90,2 68,8 35,4 147,7 0 80,1 0 34,9 0 A, s-1 6,71E04 6,89E02 2,00E-01 2,37E09 3,25E-02 4,13E03 2,65E-02 6,00E-01 1,31e-03 WGS reaction and methanation reactions were identified as main gas phase reactions. Water gas shift reaction was found to shift forward due to the presence of high water concentration resulting in very low CO concentration [47]. Methanation reaction effect on methane composition was found extremely low and methane formation was mainly linked to the glucose-xylose intermediates and guaiacol [47]. For the current study, one way WGS reaction was accepted and methanation reaction influence on reaction kinetics was neglected by aiming the consistency with experimental kinetic work. According to the experimental results of Sato, activation energy and pre-exponential factor of forward WGS reaction in non-catalytic hydrothermal conditions are 116 kJ/mol and 380189 s-1, respectively [66]. 51 4 Integration of Kinetic Data Set to the Reactor SimulationTest Run & Validation 4.1 Integration of Kinetic Data Set to the AspenPlusTM Simulation AspenPlusTM is a widely used process modeling tool for power, chemical, petrochemical, oil and gas industries for conceptual design, performance analysis and optimization purposes. It offers the world’s largest physical property database for conventional chemicals, electrolytes, solids and polymers, which is annually updated by U.S. National Institute of Standards and Technology (NIST). In addition, software has a diverse unit operation library including pump, compressor, valve, tank, heat exchanger, reactor (CSTR, PFR, Mixed Batch), flash column, absorber, filter, crystallizer, etc. Features of the software are suitable for the modeling of complex reacting systems with numerous chemical compounds at high pressure and temperature conditions. Furthermore, the scope of the work required the process flow-sheet build up with several units that are combined with reactors in a steady state system. This software was chosen in order to compile the outcome of kinetic approach to the SCWG process and to monitor the performance of the system at different operating parameters, in time efficient and precise way. A process simulation study was initiated by integrating kinetic data derived to the reaction database of the software. Based on kinetic studies obtained from the literature, biomass degradation was modeled by 1st order decomposition type reactions. Power law expression (built in the software) given in Equation 4-1 was applied for the reaction set, which relates the rate of reaction to the temperature, concentration of the reactant compound and stoichiometric coefficient. 𝑟 =𝐴∙𝑒 ( 1 1 −𝐸𝑎⁄ 𝑅 )∙[𝑇 −𝑇0 ] 𝑁 ∙ ∏ 𝐶𝑖 𝛼𝑖 Equation 4-1 𝑖=1 52 In order to achieve mass balance closure for guaiacol reactions, stoichiometric coefficient of the reactants were manipulated, even though the rate equations were kept as 1st order which was recorded as a limitation of the simulation, extent of the effects can be viewed in validation part. Rate constants were generally in a fair accordance with the Arrhenius law, hence the effect on results are expected to be minimal. In supercritical water, several non-Arrhenius behavioral reaction rate constants were preset, inserted to the software with the average value found in the given temperature range. Reactions and correspondent rate parameters were separated into two for sub- and supercritical regions due to the presence of separate studies for Dxylose and guaiacol. Having two different main reaction sets improves the running efficiency of the simulation program by splitting into liquid and vapor reacting phases. The set of reactions were applied to the T-Spec type RPlug reactor of the software with the assumption of radial mixing and process stream temperature specification for axial direction. At least 2 tubular flow reactors were used in the set of simulations due to the presence of 2 reaction sets for SCWG reactions (see Annexes A and B). The feed stream was assumed to be in liquid phase with the knowledge of high solubility of the model compounds in hot and compressed water. For thermodynamic and transport properties estimation of the simulation, selections of property methods are offered. In this study kinetic approach was accepted hence reaction equilibrium calculations were not performed, however mixing behavior of the compounds and equilibrium partial pressures are important for reactor and flash separator calculations. Among the various selections in AspenPlusTM property method menu, Ideal gas law EOS is the simplest method but insufficient of vapor-liquid equilibrium calculations near critical temperature, 373 ℃. Peng- Robinson (PR) and Redlich-Kwong (RK) property methods belong to the cubic EOS found insufficient for fluid property calculation at elevated temperatures, these EOS’ s are recommended for low to moderate pressures [10]. For the SCWG process simulations, Soave Redlich-Kwong (SRK) physical property method was selected because of its consistency in fluid property estimations near critical conditions. The general SRK expression used as property method is expressed as follows: 𝑃= 𝑅𝑇 𝑎(𝑇) − 𝑣̃ − 𝑏 𝑣̃(𝑣̃ + 𝑏) Equation 4-2 𝑎 𝑎𝑖 𝐺𝐸 𝑏 = ∑ 𝑥𝑖 − 1,546 ( + ∑ 𝑥𝑖 𝑙𝑛 ) 𝑏𝑅𝑇 𝑅𝑇𝑏𝑖 𝑅𝑇 𝑏𝑖 𝑖 Equation 4-3 𝑖 𝑅 2 𝑇𝐶2 𝑎 = 0,43 𝛼 𝑃𝐶 𝑖 Equation 4-4 53 𝑏 − 0,087 𝑅𝑇𝐶 𝑃𝐶 Equation 4-5 0,5 0,5 2 0,5 3 𝛼𝑖 = [1 + 𝑐1 (1 − 𝑇𝑟,𝑖 ) + 𝑐2 (1 − 𝑇𝑟,𝑖 ) + 𝑐3 (1 − 𝑇𝑟,𝑖 ) ] 2 Equation 4-6 For a co-volume parameter (b), linear mixing rule (Gmehling approach) is applied for a definition of the a and b parameters. Where 𝐺 𝐸 stands for a Gibbs free energy term, 𝑇𝑟 refers to the reduced temperature and 𝑐1,2,3 are the MathiasCopeman constants. 4.2 Results of the Reactor Test-Run In the preliminary runs for the reactor response verification, 2 isothermal plug flow reactors were used. The first reactor operates at 25 MPa and 370 ℃ and the second reactor operates at 25 MPa and 650 ℃. The main objective was to observe the consumption and formation rates of the organic compounds within the short residence time and the gas phase reaction trend in the 1st and 2nd reactor, respectively. Table 4.1 shows the parameters selected for the preliminary simulation runs. Table 4.1Simulation set parameters for preliminary run Biomass Biochemical Composition (Corn Cob) Cellulose Hemicellulose Lignin Ash Feed Flow Rate Biomass Load Residence time 1st reactor 2nd ractor (wt. %), dry 52 32 15 1 kg/hr 100 (wt. %) 4 s 3,4 6,7 According to Figure 4.1, cellobiose, model compound for cellulose decomposes rapidly in the presence of hot compressed water in the subcritical tubular flow reactor. The main decomposition products glycosyl-glyceraldehyde, glycosyl-glycolaldehyde and glucose are formed simultaneously. Then, glucose molar fraction starts decreasing slightly which indicates the higher rate of the overall glucose decomposition reactions as compared to the glucose formation 54 under the given process conditions. Molar fraction of the glycosylglyceraldehyde and glycosyl-glycolaldehyde, pyrolysis products of the cellobiose first increases then becomes steady, which is an indication of equal formation and decomposition reaction rates of the compounds. Fructose molar fraction develops in the beginning and then the fructose decomposition reactions dominate in the near-critical region, which results in a decrease of the compound similar to glucose. 0.0016 0.0014 Molar Fraction 0.0012 0.001 CELLOB 0.0008 GLYLGY 0.0006 GLYERY GLUCOS 0.0004 FRUCT 0.0002 0 0 1 2 3 Reactor Length (m) 4 5 Figure 4.1 Simulation results of cellobiose decomposition products in isothermal subcritical reactor at 370 ℃ Decomposition products’ formation and decomposition trends are shown in Figure 4.2. Erythrose and glyceraldehyde forms instantly since while anhydroglucose, dihydroxyacetone, pyruvaldehyde and 5-HMF form with a delay confirming the reaction network proposed for the cellobiose decomposition. These compounds are further decomposed to the acids, furfural or acetaldehyde. 1.20E-03 Molar Fraction 1.00E-03 ERYTH 8.00E-04 GLYCER GLYCOL 6.00E-04 ANHYDRO 4.00E-04 DIHYDR PYRUVAL 2.00E-04 HMF 0.00E+00 0 1 2 3 4 Reactor Length (m) 5 55 Figure 4.2 Simulation results of glucose decomposition products in isothermal subcritical reactor at 370 ℃ Using xylose as a starting compound, kinetics of hemicellulose decomposition was examined under given conditions. According to Figure 4.3, xylose is a stable compound and only a slight change in the molar fraction is observed, which indicates that for hemicellulose decomposition reactions to be favored supercritical condition is required. 0.0025 Molar Fraction 0.002 0.0015 XYLOSE METFOR 0.001 FURFURAL 0.0005 0 0 1 2 3 4 Reactor Length (m) 5 Figure 4.3 Simulation results of xylose decomposition products in isothermal subcritical reactor at 370 ℃ Decomposition of lignin in subcritical water was evaluated by performing the simulation for guaiacol with the built model. Guaiacol conversion reaches above 50% producing and bi-phenyl as a main compound. As shown in Figure 4.4, catechol, o-cresol and benzene formation are almost negligible due to the higher activation energy of the reactions, which will be favored in supercritical conditions. 56 0.0012 Molar Fraction 0.001 0.0008 GUAIACOL DIPHENYL 0.0006 CATECHOL 0.0004 OCRESOL BENZENE 0.0002 0 0 1 2 3 Reactor Length (m) 4 5 Figure 4.4 Simulation results of guaiacol decomposition products in isothermal subcritical reactor at 370 ℃ Organic acids are the main decomposition products of the biomass under hydrothermal conditions as hydrolysis products of aldehydes, ketones, furfurals and other organic intermediates. Figure 4.5 presents the molar composition development along the reactor length. The highest composition organic acids are lactic acid and acetic acid, both of which increase in composition almost steadily. Composition trend shown in Figure 4.5 is a result of reaction set applied for the subcritical condition, which disregards the hydrolysis of the acids below 370 ℃. 3.50E-03 Molar Fraction 3.00E-03 2.50E-03 2.00E-03 LACTIC 1.50E-03 ACETIC PROPIO 1.00E-03 ACRYLIC 5.00E-04 0.00E+00 0 1 2 3 Reactor Length (m) 4 5 Figure 4.5 Simulation results of organic acid decomposition products in isothermal subcritical reactor at 370 ℃ Gaseous product formation in the second reactor, which operates at 650 ℃ and 25 MPa is shown in Figure 4.6. Composition trends of the gaseous 57 Mole Fraction products are in agreement with the kinetics of the gasification reactions and also the water gas shift reaction since along the tube hydrogen, carbondioxide and methane molar compositions develop while the carbonmonoxide composition increases at the beginning then starts to diminish as time counts. Small fluctuation peaks at specific instants are estimated to be the results guaiacol and derivatives' pyrolytic gasification reactions; even though, the trend of gas compositions are mainly linked to the forward WGS reaction. 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 H2 CO2 CO CH4 C2H8 0 5 10 Reactor Length (m) 15 20 Figure 4.6 Simulation results of xylose decomposition products in isothermal supercritical reactor at 650 ℃ 58 4.3 Validation of the simulation Process gasification efficiency in the whole temperature region, up to 650 ℃ and also the intermediate yield response to the residence time and temperature variations are the results to be validated. The detailed and high contented set of experiments of Lu et al. [9], enabled the validation of the simulation system results in a range of pressure (20-30 MPa), residence time (15-47 s) and also the feedstock. Apart from the fruitful content of Lu and colleagues investigation, the temperature profile of a reacting medium was not described quantitatively. High gasification efficiency (88 – 105 %) of the experimental work [9] for short duration indicates rapid heating of the reacting medium, since it was found out earlier that gasification reactions are mostly favored at elevated temperatures and residence time spent under high temperature conditions directly influences the gasification efficiency. Accordingly, it was assumed that biomass slurry is rapidly preheated to 200 ℃ (~1 s) and spends 5 s in the subcritical reactor before reaching to the nearcritical temperature 370 ℃ (see Figure 4.7). 400 Temperature ( C ) 350 300 250 200 t vs. T 150 100 50 0 0 1 2 3 4 Residence time (s) 5 6 Figure 4.7 Temperature profile in subcritical reactor Supercritical reactor converges to the isothermal reactor with preheat to the maximum temperature of 650 ℃ within the first 1/10 th of the residence time and temperature stays constant during the rest of the operation. Variation 59 in overall residence time for comparison studies applied only to the supercritical reactor -at which almost all gas production occurs- by varying the tube length. Resulting heating trajectory and temperature versus time results table can be viewed in and Figure 4.8. 700 Temperature (C ) 600 500 400 300 t vs. T 200 100 0 0 5 10 15 Residence time (s) 20 25 Figure 4.8 Temperature profile in supercritical reactor Another important design parameter, flow type of the fluid inside the reactor is of great importance since biomass conversion in water consists mostly of diffusion driven reactions. 6 to 9 mm inner diameter reactor pipes with a highpressure stream [9], tubular flow reactor with the plug flow assumption were found to be suitable, which is possible to simulate using AspenPlusTM RPLUG Reactor. The experimental setup of Lu et al., is presented in Figure 4.9 in order to give more insight regarding the SCW gasification process in laboratory scale. The setup is mainly equipped with a water tank connected to the pumping system, two feeding tanks (one is at the process pressure while the other is atmospheric for a fresh feed), stainless steel tube reactor, external heating and thermocouples, outlet gas cooler, back pressure regulator valve liquid gas separator and outlet gas and liquid quantifying units. The feedstock used in experiments is mixed with carboxy methyl cellulose (CMC) in equal quantity in order to increase the viscosity of the fluid for pumping efficiency. In this study added CMC was accepted to have similar reacting behavior and physical properties with cellobiose. 60 Figure 4.9 Scheme of the experimental setup used in Lu’s work [9] 4.3.1 Effect of Residence Time Figure 4.10 shows the change in molar flow of the product gases with residence time. Residence time variation was performed by altering reactor tube length. For 27 s of operation, total tube length of 6,3 m was required with the input stream of 100 kg/hr. Other system parameters are set to the experimental ones, which are represented in Table 4.2. Table 4.2 Simulation operating parameters; comparisons for varying residence time Pressure (MPa) Temperature (K) Biomass feed load (%) Residence time (s) Feedstock 25 923 4 15-47 Wood sawdust +CMC The molar flow of hydrogen, carbondioxide and methane tend to increase while carbonmonoxide reduces in time. The results show that gasification reactions favor the increase in residence time and WGS reaction dominates the gas phase system (See Figure 4.10). 61 Molar Flow (kmol/hr) 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 H2 CO2 CO CH4 10 20 30 Residemce time (s) 40 50 Figure 4.10 Effect of residence time on product gas molar flow The validation study was started by comparing the conversion efficiencies; gasification efficiency and carbon gas efficiency results of the process simulation with the to-be-validated study [9]. GE refers to the percentage of the gaseous phase mass outlet to the dry biomass inlet. CE stands for the percentage of the number of moles of the carbon in the gaseous outlet to the biomass inlet to indicate the hydrogen production capacity of the system. Gasification efficiency (GE) and Carbon gas efficiency (CE) are expressed in Equation 4-7 and Equation 4-8, respectively. 𝐺𝐸 = 𝑚𝑔,𝑜 ∙ 100% 𝑚𝑏𝑖𝑜,𝑑𝑟𝑦 Equation 4-7 𝑛𝑐,𝑔 ∙ 100% 𝑛𝑐,𝑏𝑖𝑜 Equation 4-8 𝐶𝐸 = GE and CE tend to increase with the residence time, which is simply explained by enhanced production of gas products by serving more time to the gasification reactions. Average relative errors for the GE and CGE are found to be 13% and 8%, respectively. Besides the quantitative findings, trend of the efficiencies also show good accordance as can be viewed in Figure 4.11. 62 Efficiency (%) 120 100 80 CE 60 CE* GE 40 GE* 20 10 20 30 Residence time (s) 40 50 Figure 4.11 Comparison between simulation results GE, CE and experimental results GE*, CE* for residence time between 15 and 47 seconds Gaseous species compositions for the given residence time interval were also of great importance, by specifying the hydrogen or methane concentrations in the product gas, which are the main energy containing compounds at the process outlet. Figure 4.12 shows the yields of the main gaseous products of the process; hydrogen, carbondioxide, carbonmonoxide. The yield of each product gas was calculated as shown in the Equation 4-9. 𝑌𝑖 = 𝑛𝑖 𝑚𝑏𝑖𝑜,𝑑𝑟𝑦 ∙ 100% Equation 4-9 Where 𝑛𝑖 refers to the molar flowrate of a particular product gas and 𝑚𝑏𝑖𝑜,𝑑𝑟𝑦 refers to the mass flowrate of dry biomass to the reactor. A very small amount of ethane (10-8 mol gas/kg biomass) was detected but not shown in the chart, since Lu et al. did also not examine it for this study. Disagreement for short residence time is estimated to be a result of reduction in supercritical reactor (where mainly gas production takes place) while keeping subcritical reactor length constant. Hydrogen, carbon dioxide, and carbon monoxide formation with respect to residence time show the same trend as Lu’ s; yield of the components are also in a fair accordance, see Figure 4.12. Methane yield, on the other hand is higher than the experimental study, which suggests that in the experiment methane forms slower than the simulation conditions especially in the supercritical region. Hydrogen yield in the outlet stream, contrarily, is slightly lower than the validation results indicating the slower rate than the experiment. Good agreement in gasification efficiency but slight disagreement in gas compositions with Lu’s experiments suggest that gas forming reaction rates of the simulation are precise even though gas forming reaction equations are not fully correct. Lignin gasification path reaction stoichiometric coefficients were derived by performing RGibbs reactor at temperature of operation (for this case 650 ℃) since stoichiometric data was not published; which is the most likely reason of the disagreement. It is noteworthy that hydrogen, carbon dioxide and methane 63 composition increase time while the carbon monoxide concentration reduces by following the similar trend with the experimental results. Yield (mol gas/kg bio,dry) 22 H2 17 H2* CO2 12 CO2* 7 CO CO* 2 -3 10 20 30 Residence time (s) 40 50 CH4 CH4* Figure 4.12 Comparison between product yield results of the simulation and experimental results (*), for residence time between 15 and 47 seconds 4.3.2 Effect of Pressure Comparison of the gasification efficiencies and gas yields under the influence of varying operating pressure are shown in Figure 4.13 and Figure 4.14, respectively. In order to obtain the response of the system to the pressure, sensitivity analysis of simulation was performed with the model analysis tool of the AspenPlusTM. Gasification results were calculated by using the amount of product gases for specific pressure levels between 20 and 30 MPa. Simulation parameters are set to the experimental ones and summarized in Table 4.3. Table 4.3 Simulation operating parameters; comparisons for varying pressure Temperature (K) Biomass feed load (%) Residence time (s) Feedstock Pressure (MPa) 923 4 15-47 Wood sawdust + CMC 20-30 WGS reaction rate is favored by pressure due to the increase in SCW water density. Reactions of the intermediates in solvent is linked to the diffusion rate in water. Water may act like a physical barrier for reactions as a result of 64 high viscosity and reduce the rates. Increase in pressure therefore hinders these type of reactions with the “cage effect” of high viscosity water. CE and GE response to the pressure were found to be non-monotonic in Lu’s experiments due to the property change effects of water on different mechanisms. However, CE and GE trends with respect to the pressure are almost linear in the simulation results. Difference between experimental and the simulation results become smaller as pressure increases and at 30 MPa they are in very good agreement. Deviation, especially below 25 MPa can be explained by lack of the integration of the SCW property changes such as cage effect of high viscosity water, to the reaction kinetics of the simulation. Convergence of the results SCW region over 25 MPa (see Figure 4.13) approved the capability of the kinetic study application by AspenPlusTM simulation at elevated pressures. Efficiency (%) 120 100 80 CE 60 CE* GE 40 GE* 20 18 20 22 24 26 Pressure (MPa) 28 30 32 Figure 4.13 Comparison between simulation results GE, CE and experimental results GE*, CE* for residence reactor pressure between 200 and 300 bars According to the product gas yield comparison shown in Figure 4.14, a fair accordance was achieved over 25 MPa for carbon dioxide, hydrogen and carbon monoxide; on the other hand, disagreement in methane enlarged with pressure. Hydrogen, carbon dioxide and methane tend to increase with pressure in the given range of pressure in a similar, linear behavior except carbon monoxide, which decreases with pressure. Results of the study of Lu et al. [9] show that product gas yield response is not monotonic but is affected by pressure in contradictory ways due to the physical change in the solvent-solute mixture, explained in the efficiency comparison section. Yield values with respect to the pressure converges to the experiment values at 25 MPa and beyond. 65 Yield (mol gas/kg biodry) 22 H2 17 H2* CO2 12 CO2* 7 CO CO* 2 -3 18 20 22 24 26 Pressure (MPa) 28 30 CH4 32 CH4* Figure 4.14 Comparison between product yield results of the simulation and experimental results (*), for reactor pressure between 200 and 300 bars 4.3.3 Effect of Feedstock The comparison of the gasification of various feedstock improves the understanding of the quality of the simulation results for different structural compound compositions. Experiments were run at the 650 ℃ maximum temperature, 27 s residence time and 25 MPa operating pressure with a 4% biomass loading for rice straw, peanut shell, corn stalk corn cob and wood sawdust (2% CMC + 2% biomass). The composition data were obtained from Phyllis biomass database [5]; by calculating the elemental composition from the given biochemical composition and selecting the most similar samples with the Lu’s feedstock. It should be noted that 2% CMC was summed with the biomass cellulose content for the overall feed composition; furthermore, for the samples with less than 100% biochemical and ash content remaining part was equally added to the biochemical content. Resulting compositional mass flow rate of the biomass feedstock (for 100 kg/hr feed flow) are shown in Table 4.4. Table 4.4 Biochemical and ash mass compositions of dry biomass samples used in the simulation Rice Straw kg/hr Peanut Shell kg/hr Corn Stalk Corn Cob kg/hr kg/hr Wood Sawdust kg/hr Cellulose Hemicellulose Lignin Ash 2,79 0,50 0,31 0,40 2,77 0,44 0,67 0,12 2,91 0,67 0,37 0,06 3,04 0,64 0,30 0,02 2,99 0,47 0,53 0,01 Total 4 4 4 4 4 GE and CGE results show good accordance except for peanut shell and corn stalk as shown in Figure 4.15. Corn cob simulation has the CE and GE values almost identical with the Lu’s results [9]. The main reason for the different 66 results for corn stalk and wood sawdust are the use of different composition samples. Lu et al. provide the elemental analysis of the samples, even though the simulation is based on the chain decomposition reactions of cellulose, hemicellulose and lignin model compounds by the mentioned kinetic approach. Biochemical and ash analysis of the samples were collected from Phyllis biomass database [5], resulted elemental composition of peanut shell and corn stalk are different than from Lu’s samples, which is a common issue for biomass related studies since their composition depends on many variables; soil, fertilizer, weather condition and air quality. For set of the simulation runs, effect of the ash content on reaction kinetics were neglected due to the negligible ash content in wood sawdust; even though for rice straw, peanut shell and corn stalk constitutes high ash content, which is believed to have catalyzing or inhibiting effect on decomposition and gasification reactions in hydrothermal media. 120 Efficiency (%) 100 80 GE 60 GE* CGE 40 CGE* 20 0 Rice Straw Peanut Shell Corn Stalk Corn Cob Wood Sawdust Figure 4.15 Comparison between simulation results GE, CE and experimental results GE*, CE* for different feedstock; rice straw, peanut shell, corn stalk, corn cob and wood sawdust. Comparison between Lu’s SCW biomass gasification work and the current process simulation findings enabled the evaluation of the proposed model’s strengths and weaknesses. First of all, the kinetic mechanism based SCWG simulation enables to analyze the response of the system to the variation of parameters, such as residence time, pressure and feed composition; not only for the outlet conditions but also at a specific instant during the operation, as a result of the kinetic approach. Overall gasification and carbon gas efficiency results were in a very good agreement with the experimental results, both in quantity and regarding qualitative response trend. This depicts the ability of simulation to predict the degree of gasification of biomass feed, found as a fruitful outcome for further process design works. Differences for low pressures (𝑃 < 25 MPa) can be explained by kinetic studies done for 25 and 30 MPa, and 67 also with the lack of integration of pressure related inhibiting or promoting effects of water (ion product, cage effect). Product gas yield comparison with the experimental data is in a fair agreement. The differences between the simulation results and Lu’s experimental results were in a reasonable range considering the assumptions made on, organic acid types formed, direct gasification reaction stoichiometry of lignin and derivatives and non-Arrhenius reactions. It should also be noted that process conditions in the experimental work possibly scatters from the ones set for simulation, e.g. residence time, heating trend biomass feed load etc., which are suspected to lead to deviations. Simulation runs for different types of biomass feedstock resulted in good agreement with experimental except for the biomass samples with significantly different elemental compositions, such as peanut shell and corn stalk. The approach of classifying main biomass biochemicals and their reaction pathways improved the simulation precision for various feedstocks with different biochemical compositions, which was also an encouraging outcome of the current study. 68 5 Sensitivity Analysis 5.1 Sensitivity Analysis for Reactor Heating Rate Effect: Biomass decomposition and gasification in sub- and supercritical water occurs through numerous parallel and series reactions, in this study the decomposition reactions of main biochemical compounds were the scope of interest. Cellulose, hemicellulose and lignin goes into direct gasification or decompose into the liquid intermediates, the latter path continues with recombination or further decomposition to the gas products unless they are stable enough to leave the set of reactor unconverted. High gasification efficiency pathway for each biochemical compound exists if the suppression of tar, char and stable intermediates is achieved. Tar and char formation minimization are the main challenges of biomass gasification systems. Tar barely gasifies when it is formed, which leads to the plugging and loss of efficiency and is mainly sourced by molecular furfural and similar structured polymers. In addition to tar and char, formation of slow decomposing intermediates lower gasification efficiency, for a given residence time. Biomass constituents mainly convert through radical and ionic mechanisms. As described in the Chapter 2, the former follows the Arrhenius behavior and reaction rate increases with temperature regardless of the physical condition of water. Ionic reactions are affected by the dielectric constant and ion product of water. Ionic reactions occur at high rate when the water has a high ion product, which is the case in the near critical region. Evaluation of heating rate effect is vital in order to develop the reactor and process conditions for a high gaseous throughput without reactor plugging problems. Organic acids and derivatives were detected to be rather stable compounds, requiring longer residence times even in supercritical water. This part of the work aims to simulate the complete reaction network with kinetic approach, which allows gathering information at a specific time and length of the tubular reactor. Even though reduction of the rates with temperature in supercritical water was not applicable due to the power law type reaction set of AspenPlusTM, keeping the rate constants of the ionic reactions at their average value while others increase with temperature in supercritical water, helps to understand advantageous temperature profile for the outlet condition. Furthermore, Arrhenius parameters of the radical type reactions determine the rate constant dependence to the temperature, which can only be evaluated by kinetics approach followed in this set of work. In order to observe the changes in gasification efficiency, product gas yield and amount of char formed according to the temperature trend inside the tubular reactor, three different scenarios were simulated. For the sensitivity analysis the highest temperature of the sub- and supercritical reactor was designed to reach at its half-length instead of a sharp increase in the beginning and short residence time is given for the reactions; the former approach was applied to generate meaningful results for the higher capacity applications (higher heating requirements), the latter approach was followed to evaluate the efficiency results better at moderate values and. Corn cob was chosen as feedstock sample according to the promising results obtained in the validation 69 section. Inlet conditions were selected as 100 kg/hr feed flow rate 12% biomass load at 25 oC and 25 MPa. Simulation set parameters are presented in Table 5.1. Table 5.1 Simulation set parameters Feed flow rate Biomass type Dry bio load Biochemical & Ash flow rate Cellobiose Xylose Guaiacol Ash Reactor dimensions Subcritical Reactor Supercritical Reactor kg/hr Diameter Length Max. Temp. Diameter Length Max. Temp. wt. % 100 Corn Cob 12 kg/hr kg/hr kg/hr kg/hr 9,12 1,92 0,90 0,06 m m oC m m oC 0,014 1,40 370 0,05 4,60 650 The first scenario was simulated as a linearly increasing temperature profile in the whole tube length and preheating in order to reach the maximum temperature halfway the tube length followed by isothermal operation in the other half for subcritical and supercritical reactors, respectively (Figure 5.1 and Figure 5.2). 70 0 1 Residence time (s) 3 4 2 5 6 7 400 350 300 250 200 150 100 50 0 L vs T t vs T 0 0.5 1 1.5 Tube Length (m) Figure 5.1 Temperature profile in subcritical reactor for the 1st scenario 0 5 10 Residence time (s) 15 20 25 30 700 600 500 400 L vs T 300 t vs T 200 100 0 0 1 2 3 4 5 Length (m) Figure 5.2 Temperature profile in supercritical reactor for the 1st scenario The second scenario involved simulation for both of the reactors as; preheating in order to reach the maximum temperature at the halfway followed by isothermal operation in the other half (Figure 5.3 and Figure 5.4). 71 0 Residence time (s) 2 3 1 4 5 400 350 300 250 200 150 100 50 0 L vs T t vs T 0 0.5 1 1.5 Tube Length (m) Figure 5.3 Temperature profile in subcritical reactor for the 2nd scenario 0 Residence time (s) 10 15 20 5 25 30 700 600 500 400 L vs T 300 t vs T 200 100 0 0 1 2 3 4 5 Length (m) Figure 5.4 Temperature profile in supercritical reactor for the 2nd scenario The third scenario was simulated as; preheating in order to reach the maximum temperature at the halfway followed by isothermal operation in the other half and linear increasing temperature profile in the whole tube length for subcritical and supercritical reactors, respectively (Figure 5.5 and Figure 5.6). 72 0 Residence time (s) 2 3 1 4 5 400 350 300 250 200 150 100 50 0 L vs T t vs T 0 0.5 1 1.5 Tube Length (m) Figure 5.5 Temperature profile in subcritical reactor for the 3rd scenario 0 5 10 Residence time (s) 15 20 25 30 35 700 600 500 400 L vs T 300 t vs T 200 100 0 0 1 2 3 4 5 Length (m) Figure 5.6 Temperature profile in supercritical reactor for the 3rd scenario Gasification efficiency and carbon efficiency results for different reactor temperature trends should be evaluated with the knowledge of reaction mechanisms of cellulose, hemicellulose and lignin decomposition. Glucose and fructose are the main hydrolysis product of cellulose, which convert further into aldehydes, ketones, and acids or to the aromatics such as furfural and 5HMF[53]. The mechanism of the former path is mostly radical which follows the Arrhenius behavior in both sub- and supercritical water, while the ionic reactions take place mostly at the near critical water due to the high concentration of H3O+ and OH- ions (catalytic effect). Produced aromatics in subcritical region are more stable and prone to form higher molecular polymers that cannot (easily) be gasified. On the contrary, organic acids formed in the subcritical region are degradable to the product gases in the high temperature supercritical region and this results in a relatively high gasification efficiency [9]. The highest gasification and carbon efficiency is obtained in the first scenario, which has low average temperature in the subcritical reactor and high average temperature in the supercritical reactor (preheating in the first half). Results (in 73 Figure 5.7) agree with the kinetic study findings; the temperature profile of the first scenario suppresses the 5-HMF and furfural formation in the subcritical reactor and favors the radical type gasification reactions in the supercritical reactor, which result in high gasification and carbon efficiencies. The lowest efficiencies obtained in the third scenario shows the importance of the gasification reactions that take place in the high temperature supercritical region, since most of the radical type liquid to gas reactions have an elevated activation energy barrier. 90 84.27 82.19 80 66.97 Efficiency (%) 70 64.98 59.10 60 49.57 50 40 30 20 10 0 GE CE 1st 2nd 3rd Figure 5.7 GE and CE values of 1st 2nd and 3rd scenarios The percentage of carbon in unconverted liquid effluent indicates the degree of formation of rather stable products; furthermore, it is a result of short time spent in the maximum temperature region. An inverse relation between gasification efficiency and unconverted liquid holds for both analytic and kinetic reasons; hence the results can be justified with the same reasons mentioned above. Char formation from biomass in hydrothermal gasification is caused mainly by conversion of lignin and derivatives [61]. Char from guaiacol and its derivatives is formed by crosslinking of the active sites and creation of high molecular weight structures which is favored in the supercritical region at high temperatures. Carbon molar flow by means of char was calculated by subtracting the carbon molar flow rates of intermediate liquid compounds and gaseous products from carbon in biomass feed. As can be seen in Figure 5.8, the first and second scenario result in higher chars formation, which agrees with the kinetic findings. Carbonunc. liq (%) refers to the molar ratio of carbon contained in the 74 liquid effluent to the carbon fed with the biomass. NCarbon (char) is the molar carbon flow in the outlet (sub stream), which was calculated by subtracting the number of moles of carbon in gas and liquid from biomass. 1st 2nd 3rd 45 14 40 12 35 25 8 20 6 15 Carbon char (%) Carbon unc. liq. (% ) 10 30 4 10 2 5 0 0 1st 2nd 3rd Figure 5.8 Carbon mole ratio of unconverted liquid in the effluent, Carbon (unc.liq)% and molar flowrate of carbon contained in char for 1st 2nd and 3rd scenarios In Figure 5.9, product gas composition results from the multiple scenarios are shown. High gas yields in the first two scenarios are linked with gasification efficiency results. It should be noted that Carbon efficiency results represent the abundance of carbon containing gas in the product. Higher composition of carbon monoxide in the third scenario can be explained by the presence and importance of WGS reaction. Supercritical region and high temperature favors the WGS reaction, hence its effect is more significant in the first two scenarios. 75 100% 90% Composition (%) 80% 70% 60% 50% CH4 40% CO 30% CO2 20% H2 10% 0% 1st 2nd 3rd CH4 7.9 8.3 22.1 CO 19.5 18.8 15.5 CO2 34.9 34.9 24.8 H2 37.6 38.0 37.7 Figure 5.9 Product gas composition values for first, second and third scenarios 76 5.2 SCWG Process Scheme Build-Up & Sensitivity Analysis for Parameter Variations on SCWG Plant Simulation Plant design and simulation of the SCWG process was done in order to analyze the combustible gas production rate, efficiency and self-sufficiency of the system with more realistic feed rate and also to create simple process flowsheet including auxiliary units. By doing so, biomass to product gas and liquid effluent pathway was completed; additionally, the influence of various process variables such as maximum operating temperature, biomass load in the feed and type of biomass (effect of biochemical composition) were quantified. 5.2.1 Design Parameters and Unit Selections For the overall process scheme build up, early found favorable process conditions were applied with the literature examples [43], [67]. Main process units used (with the aid of AspenPlusTM) were mixer, tubular flow reactor, liquidliquid separator, vapor-liquid separator, water pump and pressure relief valves. The basic plant scheme is shown in the Annex-D. The feed stream was pumped to the operating pressure and fed to the first reactor. It should be noted that, tubular flow reactors are at the same time heat-exchanging units, which enables to compute the reaction rates and intermediate evolution during the heat-up phase similar to the real application. The reactor system is separated into three parts according to the temperature levels; in the first reactor the stream is heated up to 300 ℃ (Preheat), in the second reactor the critical condition is almost reached with an outlet temperature of 370 ℃ (Superheat) and the third reactor is where supercritical phenomena happen from 374℃ to the maximum temperature (550 – 650 ℃). The outlet stream from the 3rd reactor (Heatrcy1) is a mixture of supercritical water, intermediate liquid compounds and product gases. Heatrcy1 stream is recycled for in-direct thermal heat exchange with the reacting stream. Routlet1 stream leaving the 1st tubular reactor after the heat rejection section is first cooled down to 35 ℃ (by a fresh cold water feed possibly from a river or possible design of air-water cooling systems) and fed the liquid separator unit (Separ), where high molecular weight insoluble compounds or polymers are separated from the gas-rich pressurized water stream. In the current study, all ungasified hydrocarbons were separated aiming to improve vapor-liquid equilibrium calculation performance of the simulation, which is prone to be disrupted by the presence of complex hydrocarbons. Water gas mixture (H2O+gas) then flows to the high-pressure vapor-liquid flash separator, where product gas rich in H2 is separated from liquid down flow. The liquid stream is then de-pressurized to atmospheric conditions and enters to the low-pressure vapor-liquid flash separator. In this column, most of the CO2 solved in high-pressure water is separated from water with other remaining product gases. Product gas streams can be classified as, H2 rich high-pressure gas (PRODGAS1), CO2 rich low pressure gas (PRODGAS2). For commercial applications, the high pressure stream can be upgraded further for fuel cell uses, while low pressure stream can be utilized to produce extra heat to be used internally by the supercritical reactor if found economically viable. Unconverted organics and part of water can be recycled to the feed; in the current work mass 77 recycling was not considered due to very low concentrations of the unconverted organic compounds, which can be a further development on the subject especially for the biomass types containing higher lipid or protein, such as algae. Component molar flows in key process streams listed in Annex-D. Heat Transfer Analysis Heat exchange between recovering stream (Heatrcy1) and reacting flow is the main drive for the self-sufficiency of high-energy demand process. Design and scaling of the heat exchangers (in-direct flow passing in tubular reactors) with the most accurate results is believed to be possible with the differential approach to the heat transfer between rapid property changing pseudo liquid/vapor phase fluids and also the heat of reactions with the simultaneous moles of the organic compounds. With the consideration of the main object of the work; integration, monitoring and analysis of the kinetic model, a rather simplified method was applied for the heat exchanger design; by applying a constant overall heat transfer coefficient value, in accordance with the built in “counter-cool” heat exchangers of the software. Design temperature estimations or calculations of the heat exchange (between the thermal fluids) type tubular reactors are still not available in the literature owing to the novelty of the process, except for simplified process schemes of the pilot plant designs, please see paragraph 1.5 [7]. By making a subjective assumption based on the reactor sizing conditions and density fluctuations and with the associated tubular flow reactor scheme [7], a 150 ℃ difference between the outlet temperature of the gasifier was selected. Inside the horizontal heat exchanger, no significant pressure loss was assumed, and the hot fluid outer tube, as well as the cold fluid inner tube design was specified. Heat transfer calculation results are directly linked with the reactor tube sizings ( Table 5.2). The section of the heat exchange process is shown in Figure 5.10. Average convective heat transfer coefficients of the hot stream (hhot @255 ℃ < 𝑇ℎ𝑜𝑡 < 650 ℃), cold stream (hcold @ 25 ℃ < 𝑇𝑐𝑜𝑙𝑑 < 500 ℃) and conductive heat transfer coefficient of steel tube (kwall) are 60,000 W/m2. K, 20,000 W/m2. K, 19 W/m.K, respectively [68]. Figure 5.10 Simplified heat exchange unit scheme 78 Log-mean temperature difference (∆𝑻𝑳𝑴 ) was achieved within given temperature range for the counter-current heat exchanger (Equation 5-1). The overall heat transfer coefficient (𝑼) was calculated by using convective and conductive heat transfer resistances, as expressed by Equation 5-2. A typical residence time-reactor length was known for a hydrothermal gasification process; hence, inner and outer diameters of the tube inside were calculated (thickness was checked to be inline with the security limitations), as shown in Table 5.2. The heat transfer rate was calculated from heat duty of the water heating between 500 and 650 ℃ and 25 MPa with the aid of AspenPlusTM simulation, to be consistent. ∆𝑇𝐿𝑀 = (𝑇𝐻,𝑖𝑛 − 𝑇𝐶,𝑜𝑢𝑡 ) − (𝑇𝐻,𝑜𝑢𝑡 − 𝑇𝐶,𝑖𝑛 ) (𝑇 − 𝑇𝐶,𝑜𝑢𝑡 ) ln ( 𝐻,𝑖𝑛 ⁄ ) (𝑇𝐻,𝑜𝑢𝑡 − 𝑇𝐶,𝑖𝑛 ) 𝑈= 1 𝑟 . ln(𝑟𝑜 − 𝑟𝑖 ) 1 1 + 𝑖 + ℎ𝑐𝑜𝑙𝑑 𝑘𝑤𝑎𝑙𝑙 ℎℎ𝑜𝑡 𝑄 = 𝑈. (𝜋. 𝐷𝑡𝑢𝑏𝑒 . 𝐿𝑡𝑢𝑏𝑒 ). ∆𝑇𝐿𝑀 Equation 5-1 Equation 5-2 Equation 5-3 Table 5.2 Tubular reactor sizing parameters Transfer parameters ∆𝑇𝐿𝑀 ℎ𝑐𝑜𝑙𝑑 ℎℎ𝑜𝑡 𝑘𝑤𝑎𝑙𝑙 𝑈 (K) (W/m2.K) (W/m2.K) (W/m.K) (W/m2.K) 187 20.000 60.000 19 1090 douter(mm) 12 12 26,7 dinner(mm) 8 8 15,6 Reactors 1st Reactor 2nd Reactor 3rd Reactor length (m) 22,5 16 22 Residencetime (s) 7,2 3,3 22,9 79 Gasification and Carbon efficiency definitions are shown in the validation paragraph 4.3 Cold gas efficiency calculations (CGE) were made by dividing thermal energy achieved from product gas to the thermal energy of the biomass load (by achieving hypothetically the full recovery from the lower heating values) and externally supplied heat duty (see Equation 5-4). 𝐶𝐺𝐸 = ∑𝑖 𝑚𝑔𝑎𝑠,𝑖 . 𝐿𝐻𝑉𝑖 ∙ 100% 𝑚𝑏𝑖𝑜,𝑑𝑟𝑦 . 𝐿𝐻𝑉𝑏𝑖𝑜,𝑑𝑟𝑦 + 𝑄𝑠𝑢𝑝𝑝,𝑡ℎ Equation 5-4 Where 𝐿𝐻𝑉 refers to the lower heating value and (MJ/kg) and 𝑚 refers to the mass flowrate (kg/s). 5.2.2 SCWG of Cellulosic Biomass Plant Process Simulation For the base case simulation, Corn cob was selected as a feed sample based on the comparatively good agreement in the validation chapter. Biomass mass concentration in feed was chosen to be 15%, since it was found as a minim load for thermally self-sufficient operation in another simulation study using the Gibbs free energy minimization approach [42], demanded to be further analyzed by a kinetic approach. Reactor pressure was selected to be 25 MPa for all runs in 3 consecutive reactors (for SCWG process scheme please see Annex-D). Maximum temperature of 650 ℃ was applied for the base simulation, due to its promising gasification efficiency results in the previous chapter. The mixer is fed with water biomass mixture at 25 ℃ and 1 bar with 500 L/hr flow total rate. Recovered stream after final cooling to the 35 ℃, is separated from ungasified liquid compounds, and then gas-liquid separation was done by high and low pressure flash separators. Results were calculated from the molar flow rate values of the associated streams by using Equation 4-7, Equation 4-8, Equation 4-9 and Equation 5-4. Table 5.3 Base case simulation results Simulation (Base Case) Efficiency Values (%) GE CE CGE Gas Yield (molgas,i/kgbiomass,dry) H2 CO2 CH4 CO Gas Composition (molgas,i/molgas,tot %) H2 CO2 CH4 92,2 81,6 59,3 17,1 16,1 8 1,9 39,7 37,3 18,5 80 CO 4,5 5.2.3 Effect of Maximum Temperature Efficiency (%) Gas phase and intermediate to gas reaction rates are favored by high temperature in the supercritical region. External heating demand, on the other hand, increases with temperature of the supercritical reactor, which would negatively impact the thermal efficiency. The degree of change of GE and CE at the third reactor outlet temperature (Tmax) will therefore be beneficial for process parameter decision-making. Sensitivity analysis for Tmax was run for 3 cases, 650 ℃, 600 ℃ and 550 ℃. Resulting GE and CE values are presented in Figure 5.11. 100.00 90.00 80.00 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 92.24 81.59 76.41 70.37 58.70 56.98 GE CE 650 ℃ 600 ℃ 550 ℃ Figure 5.11 GE and CE results of the process simulation for Tmax= 650℃, 600 ℃, 550 ℃ GE and CE values increase with Tmax, a natural result of high activation energy requiring gasification reactions, which are mostly stagnant at low temperatures. Product gas formation shows a monotonic relation with Tmax, although optimum level can be understood with the further heat transfer analysis. 81 5.2.4 Effect of Biomass Load Hydrothermal biomass conversion processes generally suffer from only allowing low biomass feed shares in the water. In this set of analysis four different cases were evaluated according to the biomass load, 10, 12, 15, 18 and 20 %. Aim of the set of runs was to analyze the best load value in terms of gasification and thermal efficiency of the process. Figure 5.12 illustrates the response of GE, CE and GCE to the load variation. Figure 5.13 and Figure 5.14 show supplied external thermal energy (Qsupp) and net thermal output (Qth, net); specific supplied external thermal energy (Qsupp, spec) and specific net thermal output (Qth, net, specific) responses to the biomass load, respectively. Qth, net calculation was done according to Equation 5-5. Qth, net = ∑ 𝑚𝑔𝑎𝑠,𝑖 . 𝐿𝐻𝑉𝑖 − Q𝑠𝑢𝑝𝑝,𝑡ℎ Equation 5-5 𝑖 𝑄 𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐 = GE Qth, net 𝑚𝑏𝑖𝑜,𝑑𝑟𝑦 CE Equation 5-6 CGE 100 Efficiency (%) 90 80 70 60 50 40 8 10 12 14 16 Biomass Load (%) 18 20 22 Figure 5.12 Biomass load effect on GE and CE 82 According to Figure 5.12, GE between 10 to 15 % of biomass load does not change significantly; at 18% of load it increases from 92% to 96% then decreases back to the same level at 20%. GE mainly defines how efficiently the biomass is converted into gaseous product. Results obtained show that if reaction kinetics and reactor is designed to prevent clogging issues suffered mostly with the dense streams, biomass load as high as 18% would result in the most efficient biomass to gas conversion. The difference between GE and CE results from the H2 composition in the product gas stream, in other words, CE is a measure of the efficiency of the Carbon containing gases (e.g. CO2, CH4), which has its optimum value at 15%, meaning the highest biomass to gas carbon conversion at that point is expected to be realized. CGE trend illustrates that with high load thermal efficiency of the system is favored even though the mass flow of the dry biomass feed (denominator of the CGE definition) increases; which points out favoring effect of concentration on SCWG. For CGE, 18% load is the optimum value beyond which a decrease in CGE is observed. Definite kinetic explanation of the resulted responses is trivial to interpret due to high number and complexity of the reaction network. However, a significant effect of concentration on decomposition reaction rates can be seen from the results obtained. Qsupplied Qth, net 0.3 0.25 Q (MW) 0.2 0.15 0.1 0.05 0 8 10 12 14 16 Biomass Load (%) 18 20 22 Figure 5.13 Biomass load effect on thermal energy supplied and net thermal energy production rate As illustrated in Figure 5.13, net thermal output (Qth, net) of the SCWG process increases, while supplied thermal energy for heating up the fluid (Qsupp.) stays constant with biomass load. The horizontal trend of Qsupp. vs biomass load line is a result of water like behavior of the stream with dilute biomass content, hence unchanged heat transfer parameters. High water content in the stream absorbs to a large extent the energy supplied, and leaves the heating demand of biomass insignificant. According to the results obtained, energy production from SCWG process is favored by high biomass load as long as clogging due to high load is prevented. 83 Qsupplied,specific Qth, net, specific 10 9 Q specific (MJ/kg) 8 7 6 5 4 3 2 1 0 8 10 12 14 16 Biomass Load (%) 18 20 22 Figure 5.14 Biomass load effect on specific thermal energy supplied and net thermal energy production rate As can be viewed in Figure 5.14 Specific thermal output increases with load from 10 to 18% where it reaches a maximum. A similar trend of the Qth, net, specific and GE depicts the consistency of the sensitivity analysis since both parameters evaluate the biomass to gas conversion performance of the system, for unit feed. Qsupplied, spec decreases with biomass load, simply explained by almost constant numerator (Qsupplied) and increasing denominator (biomass feed rate). 5.2.5 Effect of Biomass Type The reaction kinetic approach of the current work enables to analyze the compositional effect of biochemical constituents in real biomass. Reaction network and kinetic parameters of cellulose (cellobiose), hemicellulose (xylose) and lignin (guaiacol) are significantly different, resulting in gasification rate and outlet composition variances. Comparison between SCWG of corn cob and rice husk will enable one to evaluate the influence of biomass constituents on process performance. Rice husk-comes up with high moisture content owing to the condition of the crop land- is a good example of high moisture biomass, a possible target of the real application. Corn cob/Rice straw has cellulose, hemicellulose, lignin and ash content of 52/39,3, 32/25, 15/15,9 and 1/19,8 %, respectively (see Table 4.4) . Significantly high ash of rice straw results in low biochemical content. In reaction sets ash was accepted as non-reacting inorganic, 84 and possible catalytic or inhibiting effects were disregarded, due to the variance in ash compositions and deficient data in the literature. Furthermore, lignin composition (excluding ash) of rice straw is notably higher than corn cob. Process operating parameters are set to the base case as can be viewed in Table 5.1. Figure 5.15 and Figure 5.16 present the GE, CE, CGE and product gas compositions for each biomass type, respectively. 100.00 92.24 90.00 81.59 78.82 Efficiency (%) 80.00 68.89 70.00 53.62 60.00 47.55 50.00 40.00 30.00 20.00 10.00 0.00 GE CE Corn Cob CGE Rice Husk Figure 5.15 Biomass type effect on GE, CE and CGE Figure 5.15 shows that, GE and CE values of corn cob are roughly 10 % higher than the ones of rice husk. Poor biomass to gas conversion efficiency of rice husk is primarily linked to the high non-reacting ash content. Dominating effect of ash content theory is supported by the comparison of carbon fraction of the liquid intermediate and biomass feed. Figure 5.15 displays that more carbon conversion is achieved from biomass to gas, even though inert ash dominates the efficiency results. CGE results follow the trends of the former ones, since poor gasification efficiency results in low combustible gas production with almost the same reactor heat duties. 85 = 120 Gas Composition (%) 100 4.45 4.17 18.5 19.4 37.3 37.8 39.7 38.7 80 60 40 20 0 Corn Cob Rice Husk H2 CO2 CH4 CO Figure 5.16 Biomass type effect on product gas compositions CO2, CH4, and CO compositions are higher for rice husk, whilst lower for H2 as depicted in Figure 5.16. The effect of the biochemical load to the product gas composition is an expected and fruitful outcome of the kinetic approach. Reaction network of cellulose is prone to end up in high H2 content product gas; on the other hand, lignin produces more carbon containing gases. 86 Conclusions and Further Developments This work has addressed the following questions: Q1. Are the reported experimental kinetic data for individual model biomass compounds capable of predicting the real lignocellulosic biomass (mainly agricultural residues) behavior? SCWG of biomass, reaction kinetic model and pathway was developed according to the experimental findings in the literature for main biomass constituents; cellulose, hemicellulose and lignin. Current data pool in the literature was found to be sufficient to propose a network for the whole reacting system, even though precision of the model can always be improved with possible future findings especially about; cellulose decomposition reactions in supercritical water over 400 ℃ (with detailed compositional data of intermediates), organic acid types formed from glucose-fructose derivatives, equation stoichiometry of lignin and derivatives gasification reactions, 5-HMF decomposition reactions in supercritical water. Furthermore, reaction rate equations that include H3O+ and OH- ions will enable to tune water properties to minimize aromatics and char forming dehydration and polymerization reactions in sub- and near-critical water. Precision of the composition calculations can be further improved with the addition of nth order reactions for required cases. Q2 AspenPlusTM software capable of simulating the SCWG of model compound combinations and deliver valid results based on the reaction kinetics approach? Validation test results showed that detailed research on kinetic data in the literature pays off with realistic process simulation outcomes. Simulation was capable of running not only with model compounds, but also with real biomass as a combination of cellulose, hemicellulose and lignin. Gasification and carbon efficiency results were promising with relative error range of 13% and 8% for residence times between 15-47 seconds. Pressure variation effect on gasification efficiency (GE) and carbon efficiency (CE) showed agreement with the experimental findings for pressures of 25 MPa and 30 MPa. Product gas yield results were promising due to the similar response to the residence time and pressure modifications except CH4, which was off range. SCWG simulation was found successful for giving precise results for various biomass types, such as rice straw, corn cob and wood sawdust. Char and tar formation sourced by glucose and fructose as well as other ionic mechanisms can be modeled more precisely by implementing rate constant - water dielectric constant ion product relation. Fluidized bed simulation with the updated version of AspenPlusTM can be the 87 further study to examine the influence of high mass-heat transfer as well as catalyst bedding material. Q3. What can be the possible process conditions to favor high gas and low tar/char yields? Sensitivity analysis results pointed out that SCWG of biomass efficiency parameters, product gas yields/compositions and tar/char formation rates were strongly influenced by process conditions, such as temperature profile in the reactors, maximum temperature reached, biomass concentration in feed and type of the feed. The highest GE and CE results were attained for rapid heating to 650 ℃ in the supercritical reactor (compared to a linear temperature profile inside the supercritical reactor), 18% biomass load with corn cob. Highest char formation was observed for the case where subcritical and supercritical reactors are both preheated, linked to favoring the aromatics to char reactions at elevated temperatures. Q4. Is biomass gasification based on the SCWG process feasible with reasonable yields and thermal self-sufficiency? According to the SCWG process simulation for 650 ℃ maximum temperature, reactor pressure, 500 L/hr feed rate with 15% biomass (corn cob) concentration 184,4 kW of net thermal energy output (Qth,net) was achieved (by assuming the exploitation of product gas LHV with full efficiency), hence selfsustainability of system was depicted. Qth,net increase was observed with higher biomass content, at 20% biomass load the maximum energy production rate of 263,8 kW was attained; it should be noted that, plugging issues related with high biomass load should be studied experimentally. It was also found that high ash and lignin content in biomass has inhibiting effect on gasification efficiency and energy output of the process. 88 References: [1] A. Demirbaş, “Biomass resource facilities and biomass conversion processing for fuels and chemicals,” Energy Convers. Manag., vol. 42, no. 11, pp. 1357– 1378, Jul. 2001. [2] J. Ramage and J. Scurlock, Biomass. Renewable energy-power for a sustainable future. Oxford: Oxford University Press, 1996. [3] C. Promdej and Y. Matsumura, “Temperature Effect on Hydrothermal Decomposition of Glucose in Sub- And Supercritical Water,” pp. 8492–8497, 2011. [4] A. Kru and A. Ga, “B io m a s s Co n v e rs io n in Wa te r a t 330-410 ° C a n d 30-50 MP a . Id e n tific a tio n o f Ke y Co m p o u n d s fo r In d ic a tin g D iffe re n t Ch e m ic a l Re a c tio n P a th w a y s,” pp. 267–279, 2003. [5] “Phyllis2 - Database for biomass and waste.” [Online]. Available: https://www.ecn.nl/phyllis2/. [Accessed: 27-Mar-2015]. [6] A. Kruse and E. Dinjus, “Hot compressed water as reaction medium and reactant,” J. Supercrit. Fluids, vol. 39, no. 3, pp. 362–380, Jan. 2007. [7] Y. MATSUMURA, T. MINOWA, B. POTIC, S. KERSTEN, W. PRINS, W. VANSWAAIJ, B. VANDEBELD, D. ELLIOTT, G. NEUENSCHWANDER, and A. KRUSE, “Biomass gasification in near- and super-critical water: Status and prospects,” Biomass and Bioenergy, vol. 29, no. 4, pp. 269–292, Oct. 2005. [8] G. Brunner, Hydrothermal and Supercritical Water Processes. Elsevier, 2014. [9] Y. LU, L. GUO, C. JI, X. ZHANG, X. HAO, and Q. YAN, “Hydrogen production by biomass gasification in supercritical water: A parametric study,” Int. J. Hydrogen Energy, vol. 31, no. 7, pp. 822–831, Jun. 2006. [10] J. A. M. Withag, “On the Gasification of Wet Biomass in Supercritical Water,” University of Twente, 2013. [11] M. J. . J. Antal, “Tower power - Producing fuels from solar energy,” vol. 32, pp. 58–62, Apr. 1976. [12] M. J. J. Antal and T. B. Reed, “Symposium on advances in synthetic fuels presented before the Division of Petroleum Chemistry, American Chemical Society, Miami Beach meeting, September 10--15, 1978.” 89 [13] M. J. ANTAL, H. L. FRIEDMAN, and F. E. ROGERS, “Kinetics of Cellulose Pyrolysis in Nitrogen and Steam,” Combust. Sci. Technol., vol. 21, no. 3–4, pp. 141–152, Apr. 2007. [14] W. S.-L. Mok and M. J. Antal, “Effects of pressure on biomass pyrolysis. II. Heats of reaction of cellulose pyrolysis,” Thermochim. Acta, vol. 68, no. 2–3, pp. 165–186, Oct. 1983. [15] M. J. Antal, “Effects of reactor severity on the gas-phase pyrolysis of celluloseand kraft lignin-derived volatile matter,” Ind. Eng. Chem. Prod. Res. Dev., vol. 22, no. 2, pp. 366–375, Jun. 1983. [16] J. Herguido, J. Corella, and J. Gonzalez-Saiz, “Steam gasification of lignocellulosic residues in a fluidized bed at a small pilot scale. Effect of the type of feedstock,” Ind. Eng. Chem. Res., vol. 31, no. 5, pp. 1274–1282, May 1992. [17] R. P. Overend, T. A. Milne, and L. K. Mudge, Eds., Fundamentals of Thermochemical Biomass Conversion. Dordrecht: Springer Netherlands, 1985. [18] M. J. Antal, T. Leesomboon, W. S. Mok, and G. N. Richards, “Mechanism of formation of 2-furaldehyde from d-xylose,” Carbohydr. Res., vol. 217, pp. 71– 85, Sep. 1991. [19] W. S. L. Mok and M. J. Antal, “Uncatalyzed solvolysis of whole biomass hemicellulose by hot compressed liquid water,” Ind. Eng. Chem. Res., vol. 31, no. 4, pp. 1157–1161, Apr. 1992. [20] O. Bobleter, “Hydrothermal degradation of polymers derived from plants,” Prog. Polym. Sci., vol. 19, no. 5, pp. 797–841, Jan. 1994. [21] H. R. Holgate, J. C. Meyer, and J. W. Tester, “Glucose hydrolysis and oxidation in supercritical water,” AIChE J., vol. 41, no. 3, pp. 637–648, Mar. 1995. [22] B. M. Kabyemela, T. Adschiri, R. Malaluan, and K. Arai, “Degradation Kinetics of Dihydroxyacetone and Glyceraldehyde in Subcritical and Supercritical Water,” Ind. Eng. Chem. Res., vol. 36, no. 6, pp. 2025–2030, Jun. 1997. [23] B. M. Kabyemela, T. Adschiri, R. M. Malaluan, and K. Arai, “Kinetics of Glucose Epimerization and Decomposition in Subcritical and Supercritical Water,” Ind. Eng. Chem. Res., vol. 36, no. 5, pp. 1552–1558, May 1997. [24] E. Jakab, K. Liu, and H. L. C. Meuzelaar, “Thermal Decomposition of Wood and Cellulose in the Presence of Solvent Vapors,” Ind. Eng. Chem. Res., vol. 36, no. 6, pp. 2087–2095, Jun. 1997. [25] A. A. Peterson, F. Vogel, R. P. Lachance, M. Fröling, M. J. Antal, Jr., and J. W. Tester, “Thermochemical biofuel production in hydrothermal media: A review 90 of sub- and supercritical water technologies,” Energy Environ. Sci., vol. 1, no. 1, p. 32, Jul. 2008. [26] D. Yu, M. Aihara, and M. J. Antal, “Hydrogen production by steam reforming glucose in supercritical water,” Energy & Fuels, vol. 7, no. 5, pp. 574–577, Sep. 1993. [27] X. Xu and M. J. Antal, “Gasification of sewage sludge and other biomass for hydrogen production in supercritical water,” Environ. Prog., vol. 17, no. 4, pp. 215–220, 1998. [28] M. J. Antal, S. G. Allen, D. Schulman, X. Xu, and R. J. Divilio, “Biomass Gasification in Supercritical Water †,” Ind. Eng. Chem. Res., vol. 39, no. 11, pp. 4040–4053, Nov. 2000. [29] Y. Matsumura, X. Xu, and M. J. Antal, “Gasification characteristics of an activated carbon in supercritical water,” Carbon N. Y., vol. 35, no. 6, pp. 819– 824, Jan. 1997. [30] X. Xu, Y. Matsumura, J. Stenberg, and M. J. Antal, “Carbon-Catalyzed Gasification of Organic Feedstocks in Supercritical Water †,” Ind. Eng. Chem. Res., vol. 35, no. 8, pp. 2522–2530, Jan. 1996. [31] H. Schmieder, J. Abeln, N. Boukis, E. Dinjus, A. Kruse, M. Kluth, G. Petrich, E. Sadri, and M. Schacht, “Hydrothermal gasification of biomass and organic wastes,” J. Supercrit. Fluids, vol. 17, no. 2, pp. 145–153, Apr. 2000. [32] A. Kruse, A. Krupka, V. Schwarzkopf, C. Gamard, and T. Henningsen, “Influence of Proteins on the Hydrothermal Gasification and Liquefaction of Biomass. 1. Comparison of Different Feedstocks,” Ind. Eng. Chem. Res., vol. 44, no. 9, pp. 3013–3020, Apr. 2005. [33] A. Kruse and E. Dinjus, “Influence of Salts During Hydrothermal Biomass Gasification: The Role of the Catalysed Water-Gas Shift Reaction,” Zeitschrift für Phys. Chemie, vol. 219, no. 3–2005, pp. 341–366, Mar. 2005. [34] B. Potic, S. R. A. Kersten, M. Ye, M. A. van der Hoef, J. A. M. Kuipers, and W. P. M. van Swaaij, “Fluidization with hot compressed water in micro-reactors,” Chemical Engineering Science. Elsevier, 01-Mar-2005. [35] S. R. A. Kersten, B. Potic, W. Prins, and W. P. M. Van Swaaij, “Gasification of Model Compounds and Wood in Hot Compressed Water,” Ind. Eng. Chem. Res., vol. 45, no. 12, pp. 4169–4177, Jun. 2006. [36] D. Knežević, D. Schmiedl, D. Meier, S. Kersten, and W. Van Swaaij, “HighThroughput Screening Technique for Conversion in Hot Compressed Water: Quantification and Characterization of Liquid and Solid Products,” Ind. Eng. Chem. Res., vol. 46, no. 6, pp. 1810–1817, Mar. 2007. 91 [37] A. KATO and Y. MATSUMURA, “Hydrothermal Pulping of Wet Biomass as Pretreatment for Supercritical Water Gasiificalion Studied Using Cabbage as a Model Compound.,” J. Japan Inst. Energy, vol. 82, no. 2, pp. 97–102, Jun. 2003. [38] Y. MATSUMURA, M. HARADA, D. LI, H. KOMIYAMA, Y. YOSHIDA, and H. ISHITANI, “Biomass Gasification in Supercritical Water with Partial Oxidation,” J. Japan Inst. Energy, vol. 82, no. 12, pp. 919–925, Jun. 2003. [39] T. L.-K. Yong and Y. Matsumura, “Kinetics analysis of phenol and benzene decomposition in supercritical water,” J. Supercrit. Fluids, vol. 87, pp. 73–82, Mar. 2014. [40] X. Hao, L. Guo, X. Zhang, and Y. Guan, “Hydrogen production from catalytic gasification of cellulose in supercritical water,” Chem. Eng. J., vol. 110, no. 1– 3, pp. 57–65, Jun. 2005. [41] L. GUO, Y. LU, X. ZHANG, C. JI, Y. GUAN, and A. PEI, “Hydrogen production by biomass gasification in supercritical water: A systematic experimental and analytical study,” Catal. Today, vol. 129, no. 3–4, pp. 275–286, Dec. 2007. [42] L. Fiori, M. Valbusa, and D. Castello, “Supercritical water gasification of biomass for H2 production: process design.,” Bioresour. Technol., vol. 121, pp. 139–47, Oct. 2012. [43] O. Yakaboylu, J. Harinck, K. Smit, and W. de Jong, “Supercritical Water Gasification of Biomass: A Literature and Technology Overview,” Energies, vol. 8, no. 2, pp. 859–894, Jan. 2015. [44] I.-G. Lee, M.-S. Kim, and S.-K. Ihm, “Gasification of Glucose in Supercritical Water,” Ind. Eng. Chem. Res., vol. 41, no. 5, pp. 1182–1188, Mar. 2002. [45] W. Schwald and O. Bobleter, “Hydrothermolysis of Cellulose Under Static and Dynamic Conditions at High Temperatures,” J. Carbohydr. Chem., vol. 8, no. 4, pp. 565–578, Sep. 1989. [46] P. D’Jesús, N. Boukis, B. Kraushaar-Czarnetzki, and E. Dinjus, “Influence of Process Variables on Gasification of Corn Silage in Supercritical Water,” Ind. Eng. Chem. Res., vol. 45, no. 5, pp. 1622–1630, Mar. 2006. [47] F. L. P. Resende and P. E. Savage, “Kinetic model for noncatalytic supercritical water gasification of cellulose and lignin,” AIChE J., p. NA–NA, 2010. [48] J. S. Kim, Y. Y. Lee, and R. W. Torget, “Cellulose Hydrolysis Under Extremely Low,” Appl. Biochem. Biotechnol., vol. 91–93, no. 1–9, pp. 331–340, 2001. [49] N. Ortega, “Kinetics of cellulose saccharification by Trichoderma reesei cellulases,” Int. Biodeterior. Biodegradation, vol. 47, no. 1, pp. 7–14, Jan. 2001. 92 [50] M. Sasaki, “Kinetics of Cellulose Conversion at 25 MPa in Sub- and Supercritical Water,” vol. 50, no. 1, 2004. [51] B. M. Kabyemela, M. Takigawa, T. Adschiri, R. M. Malaluan, and K. Arai, “Mechanism and Kinetics of Cellobiose Decomposition in Sub- and Supercritical Water,” vol. 5885, no. V, pp. 357–361, 1998. [52] K. GOTO, K. TAJIMA, M. SASAKI, T. ADSCHIRI, and K. ARAI, “Reaction mechanism of sugar derivatives in subcritical and supercritical water,” Kobunshi ronbunshu, vol. 58, no. 12. Tsukiji daisan nagaoka, pp. 685–691. [53] M. Ka and C. Th, “Glu c o s e a n d F ru c to s e D e c o m p o s itio n in S u bc ritic a l a n d S u p e rc ritic a l Wa te r : D e ta ile d Re a c tio n P a th w a y , Me c h a n is m s , a n d Kin e tic s,” pp. 2888–2895, 1999. [54] D. A. Cantero, M. D. Bermejo, and M. J. Cocero, “The Journal of Supercritical Fluids Kinetic analysis of cellulose depolymerization reactions in near critical water,” J. Supercrit. Fluids, vol. 75, pp. 48–57, 2013. [55] M. SASAKI, T. HAYAKAWA, K. ARAI, and T. ADSCHIRI, “MEASUREMENT OF THE RATE OF RETRO-ALDOL CONDENSATION OF D-XYLOSE IN SUBCRITICAL AND SUPERCRITICAL WATER,” in Hydrothermal Reactions and Techniques, WORLD SCIENTIFIC, 2003, pp. 169–176. [56] J. Qi and L. U. X. G. Pel, “Kinetics of Non-catalyzed Decomposition of D-xylose in High Temperature Liquid Water *,” vol. 15, no. 20476089, pp. 666–669, 2007. [57] A. K. Goodwin and G. L. Rorrer, “Reaction rates for supercritical water gasification of xylose in a micro-tubular reactor,” Chem. Eng. J., vol. 163, no. 1–2, pp. 10–21, Sep. 2010. [58] A. K. Goodwin and G. L. Rorrer, “Conversion of Xylose and Xylose−Phenol Mixtures to Hydrogen-Rich Gas by Supercritical Water in an Isothermal Microtube Flow Reactor,” Energy & Fuels, vol. 23, no. 7, pp. 3818–3825, Jul. 2009. [59] T. Kanetake, M. Sasaki, and M. Goto, “Decomposition of a Lignin Model Compound under Hydrothermal Conditions,” Chem. Eng. Technol., vol. 30, no. 8, pp. 1113–1122, Aug. 2007. [60] M. Sasaki and M. Goto, “Conversion of biomass model compound under hydrothermal conditions using batch reactor,” Fuel, vol. 88, no. 9, pp. 1656– 1664, Sep. 2009. [61] T. L. Yong and M. Yukihiko, “Kinetic Analysis of Guaiacol Conversion in Suband Supercritical Water,” 2013. 93 [62] F. S. Asghari and H. Yoshida, “Kinetics of the Decomposition of Fructose Catalyzed by Hydrochloric Acid in Subcritical Water : Formation of 5Hydroxymethylfurfural , Levulinic , and Formic Acids,” pp. 7703–7710, 2007. [63] J. Yu and P. E. Savage, “Decomposition of Formic Acid under Hydrothermal Conditions,” Ind. Eng. Chem. Res., vol. 37, no. 1, pp. 2–10, Jan. 1998. [64] J. C. Meyer, P. A. Marrone, and J. W. Tester, “Acetic acid oxidation and hydrolysis in supercritical water,” AIChE J., vol. 41, no. 9, pp. 2108–2121, Sep. 1995. [65] V. F. Chemie, “On the dehydration of lactic acid in near- and supercritical water,” 2013. [66] T. Sato, S. Kurosawa, R. L. Smith, T. Adschiri, and K. Arai, “Water gas shift reaction kinetics under noncatalytic conditions in supercritical water,” J. Supercrit. Fluids, vol. 29, no. 1–2, pp. 113–119, Apr. 2004. [67] A. Kruse, “Hydrothermal biomass gasification,” J. Supercrit. Fluids, vol. 47, no. 3, pp. 391–399, Jan. 2009. [68] K. Yamagata, K. Nishikawa, S. Hasegawa, T. Fujii, and S. Yoshida, “Forced convective heat transfer to supercritical water flowing in tubes,” Int. J. Heat Mass Transf., vol. 15, no. 12, pp. 2575–2593, Dec. 1972. 94 Annexes A- Subcritical Region Reactions Reaction Rxn No. type 1 Kinetic 2 Kinetic 3 Kinetic ge.g Kinetic gg.g Kinetic g.f Kinetic g.e Kinetic g.a Kinetic g.gly Kinetic f.e Kinetic f.gly Kinetic gly.dih Kinetic dih.gly Kinetic gly.p Kinetic dih.p Kinetic f.acid Kinetic p.acid Kinetic a.acid Kinetic e.acid Kinetic glyo.acid Kinetic fa.ga1 Kinetic g.5 Kinetic f.5 Kinetic 5.lf Kinetic 5.ff Kinetic xy.fu Kinetic xy.gm Kinetic fu.aa Kinetic gu.t Kinetic gu.b Kinetic gu.c gu.oc c.t Kinetic Kinetic Kinetic c.oc Kinetic Stoichiometry CELLOBIO(MIXED) --> GLYCOERY(MIXED) + GLYCOLAL(MIXED) CELLOBIO(MIXED) --> GLYCOGLY(MIXED) + ERYTHROS(MIXED) CELLOBIO(MIXED) + WATER(MIXED) --> 2 DEXTR-01(MIXED) GLYCOERY(MIXED) + WATER(MIXED) --> DEXTR-01(MIXED) + ERYTHROS(MIXED) GLYCOGLY(MIXED) + WATER(MIXED) --> DEXTR-01(MIXED) + GLYCOLAL(MIXED) DEXTR-01(MIXED) --> D-FRU-01(MIXED) DEXTR-01(MIXED) --> 1.5 ERYTHROS(MIXED) DEXTR-01(MIXED) --> LEVOG-01(MIXED) + WATER(MIXED) DEXTR-01(MIXED) --> 2 GLYCERAL(MIXED) D-FRU-01(MIXED) --> 1.5 ERYTHROS(MIXED) D-FRU-01(MIXED) --> 2 GLYCERAL(MIXED) GLYCERAL(MIXED) --> DIHYDROX(MIXED) DIHYDROX(MIXED) --> GLYCERAL(MIXED) GLYCERAL(MIXED) --> PYRUV-01(MIXED) + WATER(MIXED) DIHYDROX(MIXED) --> PYRUV-01(MIXED) + WATER(MIXED) D-FRU-01(MIXED) --> 2 LACTI-01(MIXED) PYRUV-01(MIXED) + WATER(MIXED) --> LACTI-01(MIXED) LEVOG-01(MIXED) + WATER(MIXED) --> 3 ACETI-01(MIXED) ERYTHROS(MIXED) --> 2 ACETI-01(MIXED) GLYCOLAL(MIXED) --> ACETI-01(MIXED) FORMI-01(MIXED) --> CARBO-01(MIXED) + HYDRO-01(MIXED) DEXTR-01(MIXED) --> 5-HYD-01(MIXED) + 3 WATER(MIXED) D-FRU-01(MIXED) --> 5-HYD-01(MIXED) + 3 WATER(MIXED) 5-HYD-01(MIXED) + 2 WATER(MIXED) --> LEVUL-01(MIXED) + FORMI-01(MIXED) 5-HYD-01(MIXED) --> FURFU-01(MIXED) + FORMA-01(MIXED) D-XYL-01(MIXED) --> FURFU-01(MIXED) + 3 WATER(MIXED) D-XYL-01(MIXED) --> GLYCERAL(MIXED) + METHYLFO(MIXED) FURFU-01(MIXED) + 2 WATER(MIXED) --> ACETI-01(MIXED) + ACRYL-01(MIXED) 1.7143 GUAIA-01(MIXED) --> BIPHE-01(MIXED) + 1.857 WATER(MIXED) + 0.786 OXYGE-01(MIXED) 0.8571 GUAIA-01(MIXED) --> BENZE-01(MIXED) + 0.4286 WATER(MIXED) + 0.6429 OXYGE-01(MIXED) 0.8571 GUAIA-01(MIXED) + 0.3571 OXYGE-01(MIXED) --> CATECHOL(MIXED) + 0.4286 WATER(MIXED) GUAIA-01(MIXED) --> O-CRE-01(MIXED) + 0.5 OXYGE-01(MIXED) 2 CATECHOL(MIXED) --> BIPHE-01(MIXED) + 1.5 OXYGE-01(MIXED) + WATER(MIXED) 1.16667 CATECHOL(MIXED) + 0.5 WATER(MIXED) --> O-CRE-01(MIXED) + 0.916667 OXYGE-01(MIXED) 95 t.ph Kinetic oc.t t.b b.ch la.acry acry.la la.acet acet.aa acry.hpa hpa.acry acry.pa hpa.glyco fa.ga2 l.la Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic BIPHE-01(MIXED) + 0.5 OXYGE-01(MIXED) + WATER(MIXED) --> 2 PHENO01(MIXED) 1.714 O-CRE-01(MIXED) + 0.0716 OXYGE-01(MIXED) --> BIPHE-01(MIXED) + 1.857 WATER(MIXED) BIPHE-01(MIXED) + WATER(MIXED) --> 2 BENZE-01(MIXED) + 0.5 OXYGE-01(MIXED) BENZE-01(MIXED) --> CHAR(NC) LACTI-01(MIXED) --> ACRYL-01(MIXED) + WATER(MIXED) ACRYL-01(MIXED) + WATER(MIXED) --> LACTI-01(MIXED) LACTI-01(MIXED) --> ACETA-01(MIXED) + CARBO-02(MIXED) + WATER(MIXED) ACETA-01(MIXED) + WATER(MIXED) --> ACETI-01(MIXED) + HYDRO-01(MIXED) ACRYL-01(MIXED) + WATER(MIXED) --> 3HPA(MIXED) 3HPA(MIXED) --> ACRYL-01(MIXED) + WATER(MIXED) ACRYL-01(MIXED) + WATER(MIXED) --> PROPI-01(MIXED) + 0.5 OXYGE-01(MIXED) 3HPA(MIXED) + 0.75 OXYGE-01(MIXED) --> 1.5 GLYCO-01(MIXED) FORMI-01(MIXED) --> CARBO-02(MIXED) + WATER(MIXED) LEVUL-01(MIXED) + WATER(MIXED) --> LACTI-01(MIXED) + ACETA-01(MIXED) B- Supercritical Region Reactions Rxn No. 1 2 3 ge.g gg.g g.f g.e g.a g.gly f.e f.gly gly.dih dih.gly gly.p dih.p f.acid p.acid a.acid e.acid glyo.acid g.5 f.5 5.lf g.ff Reaction type Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Stoichiometry CELLOBIO(MIXED) --> GLYCOERY(MIXED) + GLYCOLAL(MIXED) CELLOBIO(MIXED) --> GLYCOGLY(MIXED) + ERYTHROS(MIXED) CELLOBIO(MIXED) + WATER(MIXED) --> 2 DEXTR-01(MIXED) GLYCOERY(MIXED) + WATER(MIXED) --> DEXTR-01(MIXED) + ERYTHROS(MIXED) GLYCOGLY(MIXED) + WATER(MIXED) --> DEXTR-01(MIXED) + GLYCOLAL(MIXED) DEXTR-01(MIXED) --> D-FRU-01(MIXED) DEXTR-01(MIXED) --> 1.5 ERYTHROS(MIXED) DEXTR-01(MIXED) --> LEVOG-01(MIXED) + WATER(MIXED) DEXTR-01(MIXED) --> 2 GLYCERAL(MIXED) D-FRU-01(MIXED) --> 1.5 ERYTHROS(MIXED) D-FRU-01(MIXED) --> 2 GLYCERAL(MIXED) GLYCERAL(MIXED) --> DIHYDROX(MIXED) DIHYDROX(MIXED) --> GLYCERAL(MIXED) GLYCERAL(MIXED) --> PYRUV-01(MIXED) + WATER(MIXED) DIHYDROX(MIXED) --> PYRUV-01(MIXED) + WATER(MIXED) D-FRU-01(MIXED) --> 2 LACTI-01(MIXED) PYRUV-01(MIXED) + WATER(MIXED) --> LACTI-01(MIXED) LEVOG-01(MIXED) + WATER(MIXED) --> 3 ACETI-01(MIXED) ERYTHROS(MIXED) --> 2 ACETI-01(MIXED) GLYCOLAL(MIXED) --> ACETI-01(MIXED) DEXTR-01(MIXED) --> 5-HYD-01(MIXED) + 3 WATER(MIXED) D-FRU-01(MIXED) --> 5-HYD-01(MIXED) + 3 WATER(MIXED) 5-HYD-01(MIXED) + 2 WATER(MIXED) --> LEVUL-01(MIXED) + FORMI-01(MIXED) 5-HYD-01(MIXED) --> FURFU-01(MIXED) + FORMA-01(MIXED) 96 xy.wshs xy.fu Kinetic Kinetic aa.ga fu.wshs wgsr Kinetic Kinetic Kinetic gu.t Kinetic gu.c gu.ch gu.oc Kinetic Kinetic Kinetic c.oc t.ch b.ph b.na b.ch b.t ph.t ph.c ph.ch c.t na.ch la.acry acry.la la.acet acet.aa acry.hpa hpa.acry acry.pa hpa.glyco fa.ga1 fa.ga2 aa.ga pa.ga glyco.ga mf.aa fu.ch l.la Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic Kinetic D-XYL-01(MIXED) --> ACETI-01(MIXED) + ACRYL-01(MIXED) + WATER(MIXED) D-XYL-01(MIXED) --> FURFU-01(MIXED) + 3 WATER(MIXED) ACETI-01(MIXED) + ACRYL-01(MIXED) + WATER(MIXED) --> 5 CARBO-02(MIXED) + 5 HYDRO-01(MIXED) FURFU-01(MIXED) + 2 WATER(MIXED) --> ACRYL-01(MIXED) + ACETI-01(MIXED) CARBO-02(MIXED) + WATER(MIXED) --> CARBO-01(MIXED) + HYDRO-01(MIXED) 1.714 GUAIA-01(MIXED) --> BIPHE-01(MIXED) + 1.857 WATER(MIXED) + 0.786 OXYGE-01(MIXED)* 0.8571 GUAIA-01(MIXED) + 0.3571 OXYGE-01(MIXED) --> CATECHOL(MIXED) + 0.4286 WATER(MIXED) GUAIA-01(MIXED) --> CHAR(NC) GUAIA-01(MIXED) --> O-CRE-01(MIXED) + 0.5 OXYGE-01(MIXED) 1.16667 CATECHOL(MIXED) + 0.5 WATER(MIXED) --> O-CRE-01(MIXED) + 0.916667 OXYGE-01(MIXED) BIPHE-01(MIXED) --> CHAR(NC) BENZE-01(MIXED) + 0.5 OXYGE-01(MIXED) --> PHENO-01(MIXED) 2 BENZE-01(MIXED) --> NAPHT-01(MIXED) + ETHYL-01(MIXED) BENZE-01(MIXED) --> CHAR(NC) 2 BENZE-01(MIXED) + 0.5 OXYGE-01(MIXED) --> BIPHE-01(MIXED) + WATER(MIXED) 2 PHENO-01(MIXED) --> BIPHE-01(MIXED) + WATER(MIXED) + 0.5 OXYGE-01(MIXED) PHENO-01(MIXED) + 0.5 OXYGE-01(MIXED) --> CATECHOL(MIXED) PHENO-01(MIXED) --> CHAR(NC) 2 CATECHOL(MIXED) --> BIPHE-01(MIXED) + WATER(MIXED) + 1.5 OXYGE-01(MIXED) NAPHT-01(MIXED) --> CHAR(NC) LACTI-01(MIXED) --> ACRYL-01(MIXED) + WATER(MIXED) ACRYL-01(MIXED) + WATER(MIXED) --> LACTI-01(MIXED) LACTI-01(MIXED) --> ACETA-01(MIXED) + CARBO-02(MIXED) + WATER(MIXED) ACETA-01(MIXED) + WATER(MIXED) --> ACETI-01(MIXED) + HYDRO-01(MIXED) ACRYL-01(MIXED) + WATER(MIXED) --> 3HPA(MIXED) 3HPA(MIXED) --> ACRYL-01(MIXED) + WATER(MIXED) ACRYL-01(MIXED) + WATER(MIXED) --> PROPI-01(MIXED) + 0.5 OXYGE-01(MIXED) 3HPA(MIXED) + 0.75 OXYGE-01(MIXED) --> 1.5 GLYCO-01(MIXED) FORMI-01(MIXED) --> CARBO-01(MIXED) + HYDRO-01(MIXED) FORMI-01(MIXED) --> CARBO-02(MIXED) + WATER(MIXED) ACETI-01(MIXED) --> METHANE(MIXED) + CARBO-01(MIXED) PROPI-01(MIXED) + WATER(MIXED) --> 3 CARBO-02(MIXED) + 4 HYDRO-01(MIXED) GLYCO-01(MIXED) + WATER(MIXED) --> 2 CARBO-01(MIXED) + 3 HYDRO-01(MIXED) METHYLFO(MIXED) --> ACETI-01(MIXED) FURFU-01(MIXED) --> CHAR(NC) LEVUL-01(MIXED) + WATER(MIXED) --> LACTI-01(MIXED) + ACETA-01(MIXED) *For reactions in the presence of Oxygen: It should be noted that Oxygen was used as mass balance agent without being in rate equations and purged after the reactor, hence does not have influence on kinetics. 97 C- RGibss reaction stoichiometries (for Lignin Pathway) 1- 650 ℃, 250 bars Rxn Reaction No. type Stoichiometry GUAIA-01(MIXED) + 5.36262 WATER(MIXED) --> 2.81059 HYDRO-01(MIXED) + 3.63874 CARBO-01(MIXED) + 0.0851501 CARBO-02(MIXED) + 3.27573 METHANE(MIXED) + 1 Kinetic 0.000195485 ETHAN-01(MIXED) BIPHE-01(MIXED) + 11.2359 WATER(MIXED) --> 3.61427 HYDRO-01(MIXED) + 5.54694 CARBO-01(MIXED) + 0.141988 CARBO-02(MIXED) + 6.30995 METHANE(MIXED) + 2 Kinetic 0.000563771 ETHAN-01(MIXED) PHENO-01(MIXED) + 5.04449 WATER(MIXED) --> 2.16141 HYDRO-01(MIXED) + 2.98613 CARBO-01(MIXED) + 0.0722316 CARBO-02(MIXED) + 2.94123 METHANE(MIXED) + 3 Kinetic 0.000204897 ETHAN-01(MIXED) BENZE-01(MIXED) + 5.40379 WATER(MIXED) --> 1.87729 HYDRO-01(MIXED) + 2.66718 CARBO-01(MIXED) + 0.0694326 CARBO-02(MIXED) + 3.26281 METHANE(MIXED) + 4 Kinetic 0.000290222 ETHAN-01(MIXED) 2- 600℃, 250 bars Rxn Reaction No. type Stoichiometry GUAIA-01(MIXED) + 5.9489 WATER(MIXED) --> 4.06268 HYDRO-01(MIXED) + 3.8921 CARBO-01(MIXED) + 0.164702 CARBO-02(MIXED) + 2.94283 METHANE(MIXED) + 1 Kinetic 0.000187762 ETHAN-01(MIXED) BIPHE-01(MIXED) + 11.9951 WATER(MIXED) --> 5.26526 HYDRO-01(MIXED) + 5.86036 CARBO-01(MIXED) + 0.274421 CARBO-02(MIXED) + 5.86408 METHANE(MIXED) + 2 Kinetic 0.000574953 ETHAN-01(MIXED) PHENO-01(MIXED) + 5.49768 WATER(MIXED) --> 3.13528 HYDRO-01(MIXED) + 3.17897 CARBO-01(MIXED) + 0.13973 CARBO-02(MIXED) + 2.68089 METHANE(MIXED) + 0.000201878 3 Kinetic ETHAN-01(MIXED) BENZE-01(MIXED) + 5.79849 WATER(MIXED) --> 2.732 HYDRO-01(MIXED) + 2.83188 CARBO-01(MIXED) + 0.134729 CARBO-02(MIXED) + 3.0328 METHANE(MIXED) + 0.000296391 4 Kinetic ETHAN-01(MIXED) D- Component List Component ID Type Component name WATER Conventional WATER CELLOBIO Conventional Alias H2O * 98 D-XYL-01 GUAIA-01 Conventional D-XYLOSE Conventional GUAIACOL GLYCINE GLYCOGLY GLYCOERY DEXTR-01 Conventional GLYCINE Conventional Conventional Conventional DEXTROSE D-FRU-01 GLYCERAL GLYCOLAL ERYTHROS Conventional D-FRUCTOSE Conventional Conventional Conventional 1,3-DIHYDROXY-2Conventional PROPANONE DIHYDROX LEVOG-01 PYRUV-01 Conventional LEVOGLUCOSAN Conventional PYRUVIC-ALDEHYDE 55-HYD-01 Conventional HYDROXYMETHYLFURFURAL FURFU-01 Conventional FURFURAL FORMA-01 Conventional FORMALDEHYDE METHYLFO Conventional METHYL-FORMATE MAPLELAC Conventional PHENO-01 Conventional PHENOL O-CRE-01 Conventional O-CRESOL BENZE-01 Conventional BENZENE CATECHOL Conventional 1,2-BENZENEDIOL NAPHT-01 Conventional NAPHTHALENE PHENYLPH Conventional BIPHE-01 Conventional DIPHENYL FORMI-01 Conventional FORMIC-ACID ACETI-01 Conventional ACETIC-ACID GLYCO-01 Conventional GLYCOLIC-ACID LACTI-01 Conventional LACTIC-ACID PROPI-01 Conventional PROPIONIC-ACID ACRYL-01 Conventional ACRYLIC-ACID ACETA-01 Conventional ACETALDEHYDE 3HPA Conventional LEVUL-01 Conventional LEVULINIC-ACID METHANOL Conventional METHANOL AMMONIA Conventional AMMONIA HYDRO-01 Conventional HYDROGEN CARBO-01 Conventional CARBON-DIOXIDE METHANE Conventional METHANE CARBO-02 Conventional CARBON-MONOXIDE ETHAN-01 Conventional ETHANE ASH Conventional POTASSIUM-CHLORIDE C5H10O5 C7H8O2-E1 C2H5NO2D1 C6H12O6 C6H12O6N1 C3H6O3-N2 C6H10O5N1 C3H4O2 C6H6O3-N5 C5H4O2 CH2O C2H4O2-2 C6H6O C7H8O-3 C6H6 C6H6O2-E1 C10H8 C12H10 CH2O2 C2H4O2-1 C2H4O3-D1 C3H6O3-D1 C3H6O2-1 C3H4O2-1 C2H4O-1 C5H8O3-D1 CH4O H3N H2 CO2 CH4 CO C2H6 KCL 99 CHAR ETHYL-01 OXYGE-01 NITROGEN ARGON Nonconventional Conventional ETHYLENE Conventional OXYGEN Conventional NITROGEN Conventional ARGON C2H4 O2 N2 AR *Empty Alias section means user defined compounds 100 101