Process integration and exergy analysis of the autothermal reforming of glycerol using supercritical water

Depósito de Investigación de la Universidad de Sevilla https://idus.us.es/ This is an Accepted Manuscript of an article published by Elsevier in Energy, Vol. 42 (1), on June 2012, available at: https://doi.org/10.1016/j.energy.2012.03.069 © 2012 Elsevier. En idUS Licencia Creative Commons CC BY-NC-ND 1 Process Integration and Exergy Analysis of the Autothermal Reforming of Glycerol using Supercritical Water F. J. Gutiérrez Ortiz, P. Ollero, A. Serrera, S. Galera Departamento de Ingeniería Química y Ambiental, Universidad de Sevilla Camino de los Descubrimientos, s/n. 41092 Sevilla, Spain Phones: + 34 95 448 72 68/65/60 Fax: + 34 95 446 17 75 Abstract The most thermodynamically favorable operating conditions at which glycerol can be converted into hydrogen with maximum hydrogen yield by autothermal reforming using supercritical water were identified in a previous paper. As a second part of the study, a conceptual design based on energy integration and exergy analysis of the whole process has been performed. In the proposed scheme, the huge pressure energy of the gas product just at the outlet of the reforming reactor is converted into electrical power and a fraction of the expanded gas used to provide energy support for the process by burning it in a furnace, if needed. By using the optimal conditions found in the previous work, a severe deficit of energy arises in the process. Thus, both water-to-glycerol and oxygen-to-glycerol mole ratios at which thermoneutral conditions are achieved in the reformer are computed by burning all the product gas from the reformer, both for pure and pretreated crude glycerol, at different reforming and preheating temperatures. The pressure used is 240 atm. The effects of the main operating parameters are investigated by sensitivity analysis to identify optimal conditions to maximize power production under autothermal conditions, evaluating the results by energy and exergy analyses. The computations are made with the aid of AspenPlus™, using the predictive Soave-Redlich-Kwong equation of state as the thermodynamic method in the simulation of the supercritical region. Keywords Autothermal reforming, supercritical water, glycerol, hydrogen, exergy analysis, process integration, biodiesel 2 1. Introduction Glycerol is obtained as a by-product in biodiesel production using vegetable oils by a base-catalyzed transesterification reaction. The purification of this glycerol is largely based on its final purity requirements. Glycerol of high purity is an important industrial feedstock for applications in food, cosmetics, pharmaceutical and other industries; however, it is costly to refine crude glycerol, especially for medium and small-sized plants. In any case, since glycerol production and utilization have a notable impact on the economics and sustainability of biodiesel production, the development of novel processes for glycerol valorization is essential. Among all possible routes, glycerol conversion into hydrogen as an energy carrier is one of the most attractive. Different reforming processes along this line have been studied and still go on [1-6]. Likewise, glycerol bioconversion to chemicals, such as succinic acid or acetic acid, is also being studied [7, 8]. This paper continues the study started in a previous paper, focused on autothermal reforming (ATR) – reforming plus partial oxidation – using supercritical water (SCW), which is defined as water that is heated and compressed above its critical temperature (374 ºC) and pressure (22.1 MPa). Supercritical water has many advantageous properties [9-11]; among others, glycerol reforming using supercritical water in a catalyst-free process arises to be a very attractive option. Autothermal reforming using SCW may be an excellent method because of its ability to achieve thermoneutral conditions by adjusting the feed ratios, which can save on energy required for heating the reactor and make it possible to avoid the use of a catalyst, as aforementioned. Studies on the ATR of glycerol can be found in the literature [12- 16]; also on glycerol reforming in supercritical water, but non autothermal [17]. In the previous study [18], hydrogen production by autothermal reforming of pure and crude glycerol using SCW was investigated, and optimal conditions for hydrogen production from a thermodynamic point of view were obtained, but without performing an energy assessment of the process or using a process integration. In this new paper, a conceptual design based on energy integration and exergy analysis of the whole process is proposed and assessed. Although a final purified hydrogen-rich gas to use in a fuel cell may be considered as the main aim, an easier option is to convert the huge pressure energy of the reformate product at the outlet of the reforming reactor into electrical energy by means of an expander. The product gas makes it possible to sustain the energy through the overall process by further 3 energy integration. In this way no external heat source is required and the oxygen entering the reforming reactor is controlled in order to not oxidize all the glycerol fed into the process. 2. Methodology 2.1. Exergy analysis The exergy balance is a statement of the law of degradation of energy, and an exergy analysis quantifies the efficiency lost in a process due to the loss in energy quality, indicating where the process can be improved and suggesting what areas should be given consideration. In other words, exergy analysis makes it possible to evaluate the irreversibilities of the process as the exergy flow loss or lost work, once the suitable control volume is selected. In this study, the terms energy flow and exergy flow refer to energy and exergy per unit of time (power). Likewise, the terms heat flow and work flow will be used, although power will also be used for the latter. The studied process is continuous and steady-state; it is open to interactions with the surroundings, and potential and kinetic energy are not considered. Likewise, the reactors are in equilibrium condition. Also, the product gases are at the reactor temperature. The deadstate conditions (of surroundings) are 25ºC and 1 atm. All flows (mass, energy, exergy) entering the system are considered as positive, while flows exiting are indicated by a negative sign. By combining energy and entropy balances, the lost work (power), WL , can be computed as follows:     inout system outin WWBBWL  )( [1] I.e., the lost work is the net sum of exergy flows of inlet and outlet streams minus the net power (work flow) produced in the system, since the whole system is adiabatic. Otherwise, the heat flows ( j Q ) transferred to the system (positive) and by the system (negative) affected by the Carnot factor should be taken into account:         j Carnot T T0 1 [2] where T0 is 298 K and Tj is the temperature of the jth heat source or sink outside the system. Additionally, the chemical energy in the reformate gas provides the energy and exergy source. 4 Some good reviews provide further discussion about exergy terms [19]. As the final objective of reforming is to get maximum work (power) through the huge pressure energy of reformate products, the exergy efficiency of the process to be inspected is the following: WLW W net net      [3] The term Lost Work (flow) includes both internal and external exergy losses. By taking into account that the cooling water heated in the process could be used as a local source of heating, a combined heat-power exergy efficiency (h-p) may be defined as follows:         EX watercoolingnet EX watercoolingnet ph BWLW BW   [4] where watercooling inoutwatercooling BBB   )(  [5] Finally, the thermal or energy efficiency is defined as GlyGly net TLHVm W     [6] Table 1 shows the summary of the exergy analysis for each component of the process. 2.2. Process Integration and Simulation A conceptual design of the process is proposed by including a number of heat exchangers suitably located in the process to improve the energy use. The strategy followed is based on minimizing the lost work or exergy losses, so small temperature-driving forces must be achieved, by using countercurrent flow and small temperature approaches at the ends of the exchangers. In this 'from inside to outside' approach to the heat exchanger network, the internal streams close to each other at high thermal levels are the first to contact each other (the high-temperature hot streams heat the high-temperature cold streams) and the external more separated streams at low thermal levels are the last to contact each other (the lowtemperature hot streams warm up the low-temperature cold streams), by taking the reformer as a central point, where maximum temperature is required. Then, by assessing the redistribution of exchanger duties in alternative flow sheets, a final flow sheet that provides 5 the best heat integration performance is finally selected. As a consequence, the flow sheet of the process is considerably more complex than that used in our previous work [18], as illustrated in Fig. 1. A glycerol-water mixture is pumped into the supercritical reformer after heating as much as possible by three heat exchangers in order to minimize glycerol oxidation by incoming air (using a minimum O2/Glycerol mole ratio). In the reactor, non-oxidized glycerol is reformed at a temperature and pressure previously established. Hence, the reforming reactor operates by providing the energy needed to sustain the endothermic reaction and to reach the specified temperature. The reforming reactor has been simulated as an R-Gibbs reactor, in which the gas product composition and the heat of the overall reaction are calculated under conditions minimizing the Gibbs free energy. The product gas is expanded in a turbine, converting the pressure energy into electrical energy. The gas purification for obtaining a hydrogen-rich gas to be used in a fuel cell is not carried out. The product gas is cooled to 60ºC by pumping in cooling water, and condensate water present in the product gas is removed by means of a gasliquid separator. In order to reduce the glycerol fraction to oxidize in the reformer, a fraction of the gas stream at the separator outlet enters a furnace-combustor along with preheated air. The furnace-combustor is simulated as a stoichiometric reactor where the entire gas fraction that can be oxidized is combusted. The furnace is modeled as an isothermal reactor in such a way that flue gas is at 1000ºC (a real value) and the heat released is sent to a heater located just at the reformer inlet. This heater would be physically integrated inside the furnace, like steam tubes in a boiler; i.e., furnace plus heater are a single unit. In the furnace, the gas is burnt to preheat the feed up to a set preheating temperature of the glycerol-water mixture before entering the reforming reactor. The obtained flue gas is used for first heating the glycerol-water mixture, and then the compressed air needed in the reformer. The scheme includes eight process-integrated heat exchangers. Air enters the process at two points. High-pressure air is needed in the reformer, using a five-stage compressor (compression ratio around 3 per stage) with intermediate cooling (to 40ºC per stage) to avoid excessively high temperatures and optimize the compression from a thermodynamic point of view. Since the compressor is rated with an isentropic efficiency, it is modeled as an isentropic machine, so an isentropic power is calculated; in fact, it is an adiabatic but not reversible machine. The same is true for the expander. A fan is used to feed the air required in the furnace-combustor, so there is an O2 content of 3%vol in the flue gas. Specifications of the elements used in the simulation are shown in Table 2. 6 The equilibrium compositions in the reforming reactor have been calculated for given operating conditions. The computation has been done with the aid of AspenPlus™ version 2006.5 [20]. The thermodynamic method used has been the predictive Soave-Redlich-Kwong (PSRK), since it provides the most accurate binary interaction parameters and gives more satisfactory results for mixtures of non-polar and polar components, as in the case of the crude/pure glycerol and water mixture [18]. Hydrogen, carbon monoxide, carbon dioxide, methane, ethane, propane, water, methanol, ethanol, glycerol and oxygen as well as pure carbon were added manually, as they were considered the possible species in autothermal reforming of glycerol using SCW. The simulation did not predict coke formation for any of the experimental conditions in this paper, and it did not add any other compound (mole fractions lower than 10-12). The contents of ethane, propane and ethanol were also so low that they have not been included in the analyses. For all operating conditions simulated, glycerol and methanol conversions are always 100% at equilibrium condition. As in the previous paper [18], two feeds have been studied: pure and pretreated crude glycerol. The pretreated crude glycerol feed of glycerol of variable methanol content (10, 20 and 30 wt%), without water. Based on reforming reactions (7) and (8), two hydrogen yields are computed using equations (9) and (10): Glycerol reforming: C3H8O3 + 3H2O = 3CO2 + 7H2 [7] Methanol reforming: CH3OH + H2O = CO2 + 3H2 [8] Pure glycerol: 7 1 )(/ / 353 2 2 OHHChmol Hhmol H [9] Pretreated crude glycerol: 3)(/7)(/ / 3353 2 2  OHCHhmolOHHChmol Hhmol H [10] 2.3. Case studies Equilibrium compositions of reforming gas obtained were determined as a function of reforming temperature (600–1000 ºC), preheating temperature of aqueous solution of glycerol fed to the reformer (from 500 to reforming temperature for each value of this variable), water- 7 to-pure or crude glycerol mole ratio (W/G) from 3 to 99, and oxygen-(entering as air)-to-pure or crude glycerol mole ratio (O2/G), as well as MeOH content in the crude glycerol from 10 to 30 wt%), as a measure of the glycerol purity of the crude feed. A sensitivity analysis has been carried out by varying these parameters. The total mass flow rate of glycerol fed to the system is always 1000 kg/h. The required W/G ratio is achieved by adding more or less water. Pressure did not significantly affect the process in the range from 200 to 300 atm and is not included in the sensitivity analysis. Thus, a fixed value of 240 atm chosen taking into account that it is convenient to have a margin above the critical pressure (218 atm), thus accounting for pressure drops in the equipment and through the pipes. Under given operating conditions, simulations run to calculate the O2 needed in the reforming reactor for achieving thermoneutral conditions, at which the process takes place without external heat to sustain the reactions and the overall heat flow in the reformer is null. Nevertheless, a fraction of the product gas energy must be fed back in order to preheat the feed and to achieve the thermoneutral conditions in some cases, as shown below. Besides, oxygen entering the reformer must be limited to avoid the complete oxidation of glycerol fed to the reformer. Thus, the O2/G ratio should be lower than 3.5; however, for reforming to take place, the O2/G ratio should be considerably lower. In the computations, this computed parameter never reaches 2.0. Accordingly, simulations have been also run to calculate the fraction of the reformate gas to be burnt in the furnace-combustor. This fraction will be higher for high reforming temperature, low preheating temperature and high W/G mole ratio. 3. Results and discussion The optimal thermodynamic conditions obtained in the previous work [18] were first tested under the scheme proposed, i.e., 900ºC and 240 atm for 5 wt% pure glycerol in aqueous solution (equivalent to a W/G mole ratio of 99). Tables 3 and 4 show the overall energy balance at reforming temperatures of 900 and 800ºC, respectively, using an O2/G of 1. It can be observed that a heat flow of 17289 kW is needed in the reformer, but the net power is 6588 kW when reforming at 900ºC; i.e., the overall heat flow in the reformer cannot become null. Therefore, it is not possible to make the process autothermal, since the energy demanded is much higher than that provided by the system itself. In fact, there is a severe deficit of energy within the overall process, which cannot be compensated for even by converting the power obtained in the expander into thermal energy, since the reformer requires much more energy; an external energy source would be necessary and the process cannot be autothermal. By simulating at O2/G mole ratios of 2 and 3.5, the results are qualitatively similar. Thus, it is 8 necessary to know what the maximum W/G mole ratio is. This has been carried out for all reforming and preheating temperature ranges. 3.1. Limits of water-to-glycerol mole ratio for a thermoneutral condition Depending on the feed temperature at the reformer inlet, for a given reforming temperature, the constraint of attaining null overall heat in the reformer cannot be achieved at high water-to-glycerol mole ratios:  At very high W/G, the autothermal process is not possible because an unavailable energy input is required, even though the O2/G mole ratio is increased up to what is required to burn all the glycerol fed in, due to the high water flow rate. Thus, there is an unreachable absolute value of W/G (about 23 for 600ºC reforming temperature).  When the preheating temperature is very different from the reforming temperature, as W/G increases the required oxygen becomes relatively elevated since the energy input needed in the reformer is higher, especially at high reforming temperatures. This establishes a practical W/G maximum for reforming a fraction of the glycerol. Furthermore, there is also a practical minimum for the W/G mole ratio, in order to compare all the ATR cases within the same framework. For low W/G ratio values and when the preheating temperature differs significantly from the reforming temperature, only a fraction of the reformate gas is needed, and the unused fraction should be taken into account in the global assessment. This is not as simple as may be thought since the gas excess must be purified and its chemical energy converted into electrical energy; this is outside the scope of the present paper, but will be treated in the next one, although using a much more complex simulation scheme. Likewise, the cases with very low W/G ratio and high preheating temperatures (equal to or near the reforming temperature) lead to significant methane formation and low conversion of glycerol to hydrogen in equilibrium conditions. The methanation reactions of CO and CO2 (equations 11 and 12) are exothermic and thus, the exothermicity in the reformer makes it unnecessary to add oxygen. There is even a heat flow surplus, which would be released to the environment. Then, the process is no longer autothermal. CO + 3H2 = CH4 + H2O – methanation of CO – [11] CO2 + 4H2 = CH4 +2H2O – methanation of CO2 – [12] Table 5 shows the range of possible W/G mole ratios for every reforming temperature and preheating temperature, as well as the O2/G mole ratio range corresponding to the W/G 15 The operating conditions of the process have been assessed by a complete analysis of energy and exergy, carried out for every component of the process. AspenPlusTM has been used for computation, using the predictive Soave-Redlich-Kwong equation of state as the thermodynamic method in the simulation of the supercritical region. The optimal conditions found in a previous study [18] were tested first, but the resulting process is not feasible from the energy point of view since there is a severe energy deficit. Another procedure has been then tried, in which the water and the oxygen entering the reformer are calculated to achieve thermoneutral conditions aided by the energy coming from the combustion of all the product gas in a furnace. The constraint imposed is null net heat within the reformer. The preheating temperature of the inlet feed (water plus glycerol) and the reforming temperature were taken as the main parameters to change. Energy and exergy have been assessed for numerous cases, illustrated in different Figures and Tables, both for pure and pretreated crude glycerol. One unpredicted result comes from the preheating temperature, which is that better exergy efficiency and higher net power are achieved when using a low preheating temperature. The next paper of this series will be aimed at maximizing hydrogen production after purifying the product gas not burnt in the furnace in order to use it in a fuel cell to obtain supplementary electrical power apart from that obtained from the expander. As described in the previous study mentioned above, the studied ATR process using SCW may be realistic and experimental investigation is required to better understand this process. Acknowledgment This research is supported by the Science and Technology Ministry of Spain under the research project ENE2009-13755, as a Project of Fundamental Research inside the framework of the National Plan of Scientific Research, Development and Technological Innovation 2008-2011. Nomenclature B  Exergy flow (kW) H  Enthalpy flow (kW) LHV Lower Heating Value (kJ/kg) WL  Lost work flow (power) or exergy flow loss (kW) m  Mass flow rate (kg/s) 16 Q  Heat flow (kW) S  Entropy flow (kJ/(s K)) T Temperature (K) W  Work flow or Power (kW) Greek  Efficiency (%) Subscripts Gly Glycerol i, j stage in referred to some entering the control volume considered (equipment, subsystem or system) max maximum min minimum out referred to some leaving the control volume considered (equipment, subsystem or system) F-C Furnace-Combustor R Reforming reactor References [1] Adhikari, S, Fernando S, Gwaltney SR, To SDF, Bricka RM, Steele PH, Haryanto A. A thermodynamic analysis of hydrogen production by steam reforming of glycerol. Int J of Hydrogen Energ 2007; 32: 2875-80 [2] Davda RR, Shabaker JW, Huber GW, Cortright RD, Dumesic, JA. A review of catalytic issues and process conditions for renewable hydrogen and alkanes by aqueous phase reforming of oxygenated hydrocarbons over supported metal catalysts. Appl Catal B 2006; 56: 171–86. [3] Vagia EC, Lemonidou AA. Thermodynamic analysis of hydrogen production via autothermal steam reforming of selected components of aqueous bio-oil fraction. Int J of Hydrogen Energ 2008; 33: 2489 – 500. 17 [4] Luo N, Zhao X, Cao F, Xiao T, Fang D. Thermodynamic study on hydrogen generation from different glycerol reforming processes. Energ Fuel 2007; 21: 3505–12 [5] Vaidya PD, Rodrigues AE. Glycerol reforming for hydrogen production: A review. Chem Eng Technol 2009; 32: 1463–69 [6] Adhikari, S, Fernando S, Haryanto A. Hydrogen production from glycerol: An update. Energ Convers Manage 2009; 50: 2600–04 [7] Vlysidis A, Binns M, Webb C, Theodoropoulos C. A techno-economic analysis of biodiesel biorefineries: Assessment of integrated designs for the co-production of fuels and chemicals. Energy 2011; 36: 4671–83 [8] Chatzifragkou A, Makri A, Belka A, Bellou S, Mavrou M, Mastoridou M, Mystrioti P, Onjaro G, Aggelis G, Papanikolaou S. Biotechnological conversions of biodiesel derived waste glycerol by yeast and fungal species. Energy 2011; 36: 1097-108 [9] Savage, PE. Organic chemical reactions in supercritical water. Chem Rev 1999; 99: 603–21 [10] Peterson AA, Vontobel P, Vogel F, Tester JW. In situ visualization of the performance of a supercritical water salt separator using neutron radiography. The J Supercrit Fluid, 2008; 43: 490-99 [11] Calzavara Y, Dubien CJ, Boissonnet G, Sarrade S. Evaluation of biomass gasification in supercritical water process for hydrogen production. Energ Convers Manage 2005; 46: 615-631 [12] Dauenhauer PJ, Salge JR, Schmidt LD. Renewable hydrogen by autothermal steam reforming of volatile carbohydrates. J Catal 2006; 244: 238–47 [13] Swami, SM, Abraham MA. Integrated catalytic process for conversion of biomass to hydrogen. Energ Fuel 2006; 20: 2616-22 [14] Vagia EC, Lemonidou AA. Thermodynamic analysis of hydrogen production via autothermal steam reforming of selected components of aqueous bio-oil fraction. Int J of Hydrogen Energ 2008; 33: 2489 – 500. [15] Authayanun S, Arpornwichanop A, Paengjuntuek W, Assabumrungrat S. Thermodynamic study of hydrogen production from crude glycerol autothermal reforming for fuel cell applications. Int J of Hydrogen Energ 2010; 35: 6617-23 18 [16] Wang H, Wang X, Li M, Li S, Wang S, Ma X. Thermodynamic analysis of hydrogen production from glycerol autothermal reforming. Int J of Hydrogen Energ 2009; 34: 5683-90 [17] Byrd AJ, Gupta RB, Pant KK. Hydrogen production from glycerol by reforming in supercritical water over Ru/Al2O3 catalyst. Fuel 2008; 87: 2956 - 60. [18] Gutiérrez Ortiz FJ, Ollero P, Serrera A. Thermodynamic analysis of the autothermal reforming of glycerol using supercritical water. Int J of Hydrogen Energ 2011; 36: 12186-99 [19] Rosen, MA. Advances in hydrogen production by thermochemical water decomposition: A review. Energy 20110; 35: 1068–76 [20] Aspen Technology Inc., http://www.aspentech.com 19 Tables and Figures EQUIPMENT Exergy Analysis (exergy losses and exergy efficiency) Mixers (MIX)   inoutlost BBBWL  Splitters (SPL) and Mechanical Separators (SEP) inoutlost BBBWL    Pumps (P) and Fans (FAN) 0) minmin  inoutuseful BBWW  0)  lostinoutrealrealuseful BBBWW  WLBB BB BBB BB inout inout lostinout inout P          )( )( )( Multi-stage Compressor (COMP) + intermediate cooling stages j coolj inout jinout stagei iinoutlost TT T HHWBBBWL             int, 0 ,2/)( 1)(  Expanders (TURB) 0)() maxmax  inoutuseful BBWW  0)()  lostinoutrealrealuseful BBBWW  )( )( )( )( inout inout inout lostinout TURB BB WLBB BB BBB           Heat exchangers (HE) fluidhot inout fluidcold inout fluidhot inout fluidhot inout EX BB BB BB WLBB           )( )( )( )(          outin fluidcold inout fluidhot inout BBBBBBWL  )()( or also      inout fluidcold inout fluidhot inout STSTSSTSSTWL  0000 )()( Reforming reactor (R) 0;   Rinoutlost QBBBWL   Furnace-Combustor (F-C) 0 inoutCF HHQ   CF CF inoutlost Q T T BBBWL            0 1 Heater CF CF inoutlost Q T T BBBWL            0 1 Table 1. Exergy analysis of the individual process units 20 Code Equipment Specifications MIX Mixer Pressure drop: 0 atm. P1 Pump Efficiency: 80% Outlet pressure: 240 atm HE1 Heat exchanger Pressure drop: 0 atm. ΔT (hot fluid outlet - cold fluid outlet) = 2ºC HE2 Heat exchanger Pressure drop: 0 atm. ΔT (hot fluid outlet - cold fluid outlet) = 3ºC HE3 Heat exchanger Pressure drop: 0 atm. ΔT (hot fluid outlet - cold fluid outlet) = 3ºC HEATER Heat exchanger Pressure drop: 0 atm. Outlet temperature = Variable (500 – 800 ºC) R1 Reactor Inlet temperature: Variable (800 ºC as nominal) Pressure drop: 0 atm TURB Expander Isentropic turbine (isentropic efficiency: 0.72) Outlet Pressure: 1.5 atm. HE4 Heat exchanger Pressure drop: 0 atm. ΔT (hot fluid outlet - cold fluid outlet) = 3ºC HE5 Heat exchanger Pressure drop: 0 atm. Outlet temperature of hot fluid = 60ºC SEP Gas-liquid separator Adiabatic Pressure drop: 0 atm. F-C Furnace-Combustor Combustion of everything able to be oxidized Operating temperature: 1000 ºC Operating pressure: 1 atm Surplus reaction heat  HEATER C1 – C5. Compressor (five stages) Isentropic machine Compression ratio: 2.992 per stage Final pressure at the outlet: 240 atm. Intermediate Coolers for C1-C5 Four heat exchangers Pressure drop: 0 atm. Outlet temperature in each stage = 40ºC C6 Compressor Isentropic machine (isentropic efficiency: 0.72) Outlet pressure: 1.1 atm HE6 Heat exchanger Pressure drop: 0 atm. ΔT (hot fluid outlet - cold fluid outlet) = 3ºC HE7 Heat exchanger Pressure drop: 0 atm. ΔT (hot fluid outlet - cold fluid outlet) = 3ºC P2 Pump Efficiency: 80% Outlet pressure: 1.1 atm SPLIT Splitter Split fraction stream TGE: variable on heat required in Heater; final value to get null SGOUT Pressure drop: 0 atm Table 2. Specifications of the components for the simulation 21 Work/Heat entering the system (kW) Enthalpy of the inlet-streams (kW) P1 220 . 36 Glycerol - 2053 . 65 P2 1.11 Water -83766.50 (C1-C5) + inter. cool. 72.99 Air -0.09 C6 14.93 Air3 -0.28 Reformer (R1) 17289.35 CW1 -875940.00 Work leaving the system (kW) Enthalpy of the outlet-streams (kW) TURB -6897.39 GC6 -7949.23 W1 -81131.44 CW4 -861978.61 OVERALL 10701.35 10701.24 Table 3. Overall energy balance for autothermal reforming of glycerol using SCW at 900ºC, 240 atm and W/G of 99 by burning the entire product gas, using an O2/G of 1.0 Work/Heat entering the system (kW) Enthalpy of the inlet-streams (kW) P1 220 . 36 Glycerol - 2053 . 65 P2 0.83 Water -83766.50 (C1-C5) + inter. cool. 72.99 Air -0.09 C6 14.93 Air3 -0.28 Reformer (R1) 15761.07 CW1 -825246.39 Work leaving the system (kW) Enthalpy of the outlet-streams (kW) TURB -6170.67 GC6 -7984.78 W1 -81089.42 CW4 -812093.06 OVERALL 9899.52 9899.65 Table 4. Overall energy balance for autothermal reforming of glycerol using SCW at 800ºC, 240 atm and W/G of 99 by burning the entire product gas, using an O2/G of 1.0 Reforming temperature Preheating temperature W/G and O2/G ranges 600 700 800 900 1000 (ºC) (mol/mol) (ºC) (ºC) (ºC) (ºC) (ºC) 500 W/G 10 – 23 3 – 20 3 – 17 3 – 15 3 – 13 O2/G 0.017 – 0.320 0.032 – 0.797 0.319 – 1.20 0.624 – 1.478 1.028 – 1.699 600 W/G Non an ATR case 7 – 20 3 – 17 3 – 15 3 - 13 O2/G 0.013 – 0.390 0.136 – 0.843 0.422 – 1.163 0.837 – 1.428 700 W/G 19 – 20 4 – 18 3 - 16 3 – 14 O2/G 0.012 – 0.025 0.039 – 0.544 0.257 – 0.905 0.674 – 1.217 800 W/G 8 – 18 3 – 16 3 – 14 O2/G 0.013 – 0.216 0.104 – 0.606 0.520 – 0.965 900 W/G 5 – 17 3 – 14 O2/G 0.035 – 0.317 0.368 – 0.711 1000 W/G 3 – 15 O2/G 0.212 – 0.459 Table 5. W/G and O2/G mole ratio ranges for different preheating and reforming temperature (pure glycerol at 240 atm.); upper limits of W/G correspond to the maximum integer and O2/G is computed for these values 22 Work entering the system (kW) Enthalpy of the inlet-streams (kW) P1 48.69 Glycerol - 2053.65 P2 0.17 Water -15987.49 (C1-C5) + inter. cool. 15.78 Air -0.02 C6 19.62 Air3 -0.10 CW1 -155340.25 Work leaving the system (kW) Enthalpy of the outlet-streams (kW) TURB -1478.14 GC6 -7428.99 W1 -14482.31 CW4 -152864.33 OVERALL -1393.89 -1394.13 Table 6. Overall energy balance for autothermal reforming of glycerol using SCW for thermoneutral condition by burning the entire product gas (240 atm. 800ºC reforming and preheating temperature) Heat exchanger Q (kW) Tin) cool Tout) cool Tin) hot Tout) hot cool m  (kg/h) hot m  (kg/h) HE01 288.65 28.67 82.99 92.03 85.00 4626.30 35234.42 HE02 525.04 82.99 183.34 390.49 186.33 4626.30 7756.24 HE03 1716.73 183.34 394.81 1000.00 397.81 4626.30 7756.24 HE04 19.36 194.39 387.50 397.81 390.49 322.42 7756.24 HE05 2764.42 25.00 92.03 286.11 60.00 35234.42 4948.71 HE06 38.02 60.00 107.69 126.09 110.69 1632.58 7756.24 HE07 149.94 36.40 123.09 186.33 126.09 6123.66 7756.24 HEATER 2113.71 394.81 802.39 - - 4626.30 - Temperatures in ºC. Table 7. Heat flows, mass flow rates and temperatures for all the seven plus one heat exchangers for thermoneutral condition by burning the entire product gas (240 atm. 800ºC reforming and preheating temperature) EQUIPMENT B   (kW) in W  (kW) out W  (kW) in Q  (kW) out Q  (kW) cool S   (kW/K) hot S   (kW/K) lost B  (kW) MIX1 -25.80 25.80 P1 38.98 48.69 9.71 HE01 288.65 -288.65 0.88 -0.80 24.33 HE02 525.04 -525.04 1.30 -0.94 105.71 HE03 1716.73 -1716.73 2.96 -1.81 342.91 R1 -474.90 474.90 TURB -1862.70 -1478.14 0.00 0.00 384.56 HE05 0.00 2764.42 -2764.42 8.37 -7.31 314.95 SEP -1.73E-04 0.00 HE6 38.02 -38.02 0.11 -0.10 2.84 C6 14.30 19.62 5.32 HE7 149.94 -149.94 0.43 -0.35 23.03 FURNACE -2910.75 0.00 -2113.71 1292.10 HE4 19.36 -19.36 0.03 -0.03 1.69 P2 0.13 0.17 0.00 0.00 0.03 COMP+cool-inter. 46.53 73.12 0.00 -57.34 40.04 HEATER 1329.02 0.00 2113.71 0.00 289.70 Total Exergy Losses Total Exergy Losses 3337.62 Table 8 Changes in exergy flows, work and heat flows, and exergy loss in the individual process units for thermoneutral condition by burning the entire product gas (240 atm, 800ºC reforming and preheating temperature) 23 Preheating temperature W/G ratio O2/G ratio lost B  out W  net WThermal Exergy combined H2 (ºC) (mol/mol) (mol/mol) (kW) (kW) (kW) (%) (%) (%) (%) Reforming temperature of 600 ºC 500 23.15 0.32 3526.63 1333.41 1231.10 27.70 25.88 29.49 19.65 600 Non autothermal case Reforming temperature of 700 ºC 500 20.11 0.80 3547.21 1502.09 1397.80 30.95 27.94 31.17 30.38 600 20.50 0.39 3458.65 1427.49 1327.26 29.86 27.73 31.07 31.79 700 20.83 0.04 3381.85 1361.91 1284.67 28.91 27.53 30.96 32.88 Reforming temperature of 800 ºC 500 17.52 1.20 3552.25 1658.47 1510.72 33.99 29.84 32.77 38.00 600 17.85 0.88 3482.62 1601.23 1473.91 33.16 29.74 32.72 40.65 700 18.20 0.54 3408.75 1539.21 1433.73 32.26 29.61 32.66 43.04 800 18.54 0.22 3337.62 1478.14 1393.89 31.36 29.46 32.59 45.03 Reforming temperature of 900 ºC 500 15.58 1.48 3547.11 1777.29 1615.02 36.34 31.29 34.03 39.50 600 16.01 1.21 3480.22 1721.25 1578.79 35.52 31.21 34.00 43.87 700 16.40 0.90 3416.17 1666.20 1542.65 34.71 31.11 33.94 48.36 800 16.78 0.60 3353.72 1611.00 1506.09 33.89 30.99 33.87 52.32 900 17.12 0.32 3297.03 1559.48 1471.60 33.11 30.86 33.78 55.75 Reforming temperature of 1000 ºC 500 13.63 1.70 3527.73 1886.96 1714.19 38.57 32.70 35.30 35.65 600 14.02 1.43 3467.54 1837.87 1682.50 37.86 32.67 35.29 41.45 700 14.38 1.22 3414.66 1793.91 1654.07 37.22 32.63 35.29 45.83 800 14.72 0.96 3360.78 1749.09 1624.86 36.56 32.59 35.27 51.01 900 15.03 0.73 3307.82 1704.21 1595.52 35.90 32.54 35.24 55.68 1000 15.37 0.46 3255.71 1659.28 1565.74 35.23 32.47 35.19 60.75 Table 9. Water to glycerol mole ratio, oxygen to glycerol mole ratio, exergy losses, turbine and net power; energy, exergy and combined efficiencies and hydrogen yield for different reforming and preheating temperatures Figure 1. Flow-sheet of the simulated process 01 02 GLY W AIRIN 06 SG0 1 SG0 2 CW3 CW1 CW2 AIR GC3 GC4 04 SG0 3 AIR3IN SGIN2 GC Q GC2 AIR3 C GC5 SGIN GC6 05 AIR3 SG04 W1 SGO UT P1 M IX R1 T URB HE01 P2 C5 C4 C3 C1 HE02 HE05 FURNACE HE04 HE07 HE06 HE03 C6 SEP C2 HEATE R SPL IT Supercritical Water Re forme r CW4 24 0,0 0,2 0,4 0,6 0,8 1,0 1,2 3 5 7 9 11 13 15 17 19 21 23 25 W/G mole ratio Oxygen to glycerol mole ratio 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 Fraction of gas burnt 700ºC 800ºC 900ºC 1000ºC 700_gas burnt 800_gas burnt 900_gas burnt 1000_gas burnt Figure 2. O2/G mole ratio and fraction of product gas to be burnt versus W/G (up to the allowed limits), for different reforming temperatures when preheating the feed to the reforming temperature Figure 3. O2/G mole ratio and fraction of product gas to be burnt versus W/G (up to the allowed limits), for different reforming temperatures when preheating the feed at 500ºC 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 600 700 800 900 1000 Reforming temperature (ºC) Turbine and net power (kW) Wturbine Wnet Wturbine_500 Wnet_500 Figure 4. Turbine work and net work flows versus the reforming temperature when preheating at reforming temperature and preheating at 500ºC (pure glycerol, thermoneutral condition and 240 atm) 0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4 1,6 1,8 3 5 7 9 11 13 15 17 19 21 23 25 W/G mole ratio Oxygen to glycerol mole ratio 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 Fraction of gas burnt 600 700ºC 800ºC 900ºC 1000ºC 600_gas_burnt 700_gas burnt 800_gas burnt 900_gas burnt 1000_gas burnt Preheating temperature = 500ºC