MeT 01/2016 COMPONENTS DATABANK FOR THE INDUSTRIAL SIMULATIONS IN THE VIRTUAL SUGARCANE BIOREFINERY (VSB)

Transcription

1 Centro Nacional de Pesquisa em Energia e Materiais Laboratório Nacional de Ciência e Tecnologia do Bioetanol MeT 01/2016 COMPONENTS DATABANK FOR THE INDUSTRIAL SIMULATIONS IN THE VIRTUAL SUGARCANE BIOREFINERY (VSB) Isabelle Sampaio Edvaldo Morais Tassia Junqueira Vera Gouveia Antonio Bonomi

2 ABSTRACT The simulation of an industrial process, within the Virtual Sugarcane Biorefinery platform, is based on mass and energy balances carried out through process simulators, such as Aspen Plus, SuperPro and EMSO. Considering the process simulation in Aspen Plus, the inclusion of components present in a sugarcane biorefinery (e.g., first and second generation processes) is detailed. Some components and additional data were introduced in the simulator databank due to their absence or the need to complement the physicochemical properties with information available in literature. Guidelines for the definition of thermodynamic models in the unit operation models is also included in this technical memorandum. The appropriate selection is essential to improve process representativeness, since the thermodynamic models are the basis for the calculation of phase equilibrium computation and energy balance. Keywords: biorefinery; process simulation; properties; thermodynamic models. 2

3 SUMMARY 1 INTRODUCTION COMPONENT DATABANK ASPEN PLUS PROPERTY METHODS FINAL REMARKS REFERENCES

4 1 Introduction The Virtual Sugarcane Biorefinery (VSB) is an innovative framework that integrates computer simulation platforms with economic, social and environmental evaluation tools to assess technical and sustainability impacts of different sugarcane biorefinery alternatives/routes integrating all the stages of the biomass chain: agricultural production, transport, industrial conversion, use and final disposal of the products. The first steps in the process simulation include definition of feedstock composition components (feedstock, inputs, products, etc.), physicochemical properties and thermodynamic models. Feedstock composition was previously described in MeT 22/2015, the other steps are covered in this technical memorandum. This technical memorandum presents the current component databank constructed for the VSB simulations, the properties that were modified or inserted when necessary, their respective sources and the thermodynamic property methods selected for the different unit operations of the process. 2 Component databank Sugarcane composition is comprised of water, fibers, extractives and ashes (see MeT 22/2015). Different chemical compounds were chosen to represent these categories of components on the databank as described in Morais et al. (2016): Ashes are assumed to be comprised by: Minerals: represented by K2O; Salts: represented by KCl, since potassium salts are those present in greater proportion (approximately 60% of the total salts in the ash); Soil: represented by SiO2; Extractives are assumed to be comprised by: Organic acids: represented by aconitic acid. The concentration of this acid on sugarcane is around 3 times higher than that of all the other organic acids; Sucrose; Glucose; Phosphoric acid: represents the phosphates present in the sugarcane plant. 4

5 It is important to keep in mind that not only the components present on the inputs of the process should be inserted (such as sugars from sugarcane), but all chemical compounds should be considered, whether they are final products (ethanol), undesired fermentation products (e.g. acetic acid, glycerol) or reactants (such as lime for juice treatment). Table 1 presents all the components inserted into the simulations databank. As a general rule, the same databank is used in all simulations constructed within the VSB, with occasional modifications depending on specific demands of the simulated process (e. g.: insertion of new chemical compounds, modification of specific properties of the compounds). Some components were already available at the Aspen Plus databank and did not have their properties altered; other compounds were not present at the databank (identified as user defined in Table 1) or needed additional properties. Table 1: Current chemical compounds databank of the simulations. Aspen Plus ID Complete name Type Alias on Aspen Plus databank ACET-AC Acetic Acid Conventional ACETIC-ACID ACETATE Cellulose Acetate Solid a ACETIC-ACID CAL-ACON Calcium Aconitate Solid a USER DEFINED CAL-PHOS Calcium Phosphate Solid a CALCIUM-PHOSPHATE CAO Calcium Oxide Conventional CALCIUM-OXIDE CAOH2 Lime Conventional CALCIUM-HYDROXIDE CELLULOS Cellulose Solid a USER DEFINED CO2 Carbon Dioxide Conventional CARBON-DIOXIDE ENZYME Enzymes Solid a USER DEFINED ETHANOL Ethanol Conventional ETHANOL FLOCCUL Flocculant Solid a POLY(ACRYLAMIDE-STYRENE) FURFURAL Furfural Conventional FURFURAL GLUCOLIG Glucose Oligomers Conventional DEXTROSE GLUCOSE Glucose Conventional DEXTROSE GLYCEROL Glycerol Conventional GLYCEROL H2O Water Conventional WATER H2SO4 Sulfuric Acid Conventional SULFURIC-ACID H3PO4 Phosphoric Acid Conventional ORTHOPHOSPHORIC-ACID 5

6 HMF Hydroxymethylfurfural Conventional USER DEFINED ISOAMIL Isoamyl alcohol Conventional 3-METHYL-1-BUTANOL ISOBUTOH Isobutanol Conventional ISOBUTANOL LGNSOL b Soluble lignin Solid a USER DEFINED LIGNIN Lignin Solid a USER DEFINED MINERALS Minerals Conventional POTASSIUM-OXIDE N2 Nitrogen Conventional NITROGEN NAOH Caustic Soda Conventional SODIUM-HYDROXIDE NH4OH Ammonium Hydroxide Conventional AMMONIUM-HYDROXIDE O2 Oxygen Conventional OXYGEN ORG-AC Organic Acids Conventional TRANS-ACONITIC-ACID SALTS Salts Conventional POTASSIUM-CHLORIDE SOIL Soil Conventional SILICON-DIOXIDE SUCROSE Sucrose Conventional SUCROSE XYLAN Xylan Solid a USER DEFINED XYLOLIG Xylose Oligomers Conventional D-XYLOSE XYLOSE Xylose /Arabinose Conventional D-XYLOSE YEAST Yeast Solid a USER DEFINED a Solid components do not take part in phase equilibrium. b Solubilization represented by a phase change (solid to mixed). For most chemical compounds presented on Table 1, Aspen Plus default physicochemical properties were used. For other components, respective properties used on the simulations databank (Aspen Plus v ) were defined as explained below: ACETATE (acetyl group): represented by acetic acid, having its standard enthalpy of combustion changed to match the one calculated for the reaction equation: Acetate(s) + 2O2 2H2O + 2 CO2; CAL-ACON (calcium aconitate): properties collected from gypsum (CaSO4-2H2O) available at NREL databank (WOOLEY AND PUTSCHE, 1996), except by its molecular weight (MW), which was calculated based on the molecular formula of calcium aconitate; CELLULOSE: Cellulose was inserted as a user defined component. The values for molecular weight, enthalpy of formation of the solid (DHSFRM), solid heat capacity (CPSPO1) and molar volume of the solid (VSPOLY) were found on Wooley and Putsche (1996); 6

7 ENZYME: the enzyme component was inserted as a solid component with user defined properties. The values for molecular weight (molecular formula: CH1.59O0.42N0.24S0.01), enthalpy of formation, solid heat capacity and volume of the solid were retrieved from NREL (2011); FLOCCUL (Flocculant polymer): represented on the simulation databank by polyacrylamide; GLUCOLIG (glucose oligomers): inserted as glucose (dextrose), with recalculated values for molecular weight and standard enthalpy of formation, using data from cellulose (WOOLEY AND PUTSCHE, 1996); H3PO4 (phosphoric acid): the value for the dipole moment (MUP) of the compound was retrieved from Colby College (2012); HMF (hydroxymethylfurfural): this component was not present at the Aspen Plus databank (version 7.3.2). The molecular structure of HMF was imported from CHEMSPIDER (2012). Since version 8.6, this component is available in the Aspen Plus databank (2015); LGNSOL (soluble lignin): similar to LIGNIN with same physicochemical properties. The solubilization of lignin is represented by a phase change (solid to mixed). LIGNIN: some data for lignin were retrieved from Wooley and Putsche (1996) (solid heat capacity, molar volume of the solid). Other values were modified to better represent the sugarcane lignin, as the values found on Wooley and Putsche (1996) represent wood lignin. The lignin structure was also modified to correctly represent sugarcane lignin (molecular formula: C9O2.9H8.6(OCH3)). The correct value for the enthalpy of formation of this compound was estimated using the enthalpy of combustion (27000 kj/kg) given by Stanmore (2010); NAOH: the dipole moment for the compound was retrieved from Colby College (2012); NH4OH: Molecular weight, critical pressure (Pc), critical temperature (Tc) and normal boiling point (Tb) were retrieved from SuperPro Designer databank. Dipole moment was assumed equal to the dipole moment of NaOH, found on Colby College (2012). The enthalpy of formation was retrieved from an older version of the Aspen Plus databank (v. 7.1), since this property was missing in v ; 7

8 ORG-AC (organic acids): This compound represents the organic acids present on the sugarcane and was inserted as trans-aconitic-acid on the databank. The parameters from the Aspen Databank for acetic acid were used for the radius of gyration (RGYR), dipole moment and enthalpy of formation. XYLAN: Xylan was inserted as a user defined component. Values for solid heat capacity were obtained in NREL (2011); values for molar volume of the solid and enthalpy of formation of the solid were found on Wooley and Putsche (1996); XYLOLIG (xylose olygomers): inserted as xylose (d-xylose), with recalculated values for molecular weight and standard enthalpy of formation, using data from xylan (WOOLEY AND PUTSCHE, 1996). The parameters from the Aspen Plus databank for glucose were used for the radius of gyration and dipole moment; XYLOSE: inserted as xylose (d-xylose). The values for dipole moment and radius of gyration of glucose were used for xylose; YEAST: molecular weight and solid standard enthalpy of formation were retrieved from SuperPro Designer databank. Solid molar volume was calculated based on its density (1180 kg/ m 3 ) from Scherrer et al. (1977) and molecular weight. The solid heat capacity polynomial function coefficients were obtained from NREL (2011). Some components, even those retrieved from one of the Aspen Plus databases, had a few properties modified or inserted when necessary. The modifications provided by the users are stored in USERDEF Table (a specific table for the main properties of user defined compounds); CPSPO1 (Solid heat capacity) and VSPOLY (Solid molar volume). The inserted values are shown in Table 2, Table 3 and Table 4, respectively. 8

9 Table 2: User defined properties. Component Molecular weight Solid standard enthalpy of formation Standard enthalpy of formation Pc Dipole moment Tb Tc Radius of gyration Aspen code MW DHSFRM DHFORM PC MUP TB TC RGYR Unity g/mol J/kmol kcal/mol bar debye C C meter ACETATE -4.56E+08 CAL-ACON 462-2E+09 CELLULOS E+08 ENZYME a E+07 GLUCOLIG H3PO HMF LGNSOL LIGNIN E E+08 NAOH NH4OH ORG-AC E-10 XYLAN E+08 XYLOLIG E-10 XYLOSE E-10 YEAST E+07 a Enzyme properties may change according to the cocktail used. Table 3: Solid molar volume* VSPOLY CAL-ACON CELLULOS ENZYME LGNSOL LIGNIN XYLAN YEAST C C C C C C C * Solid Volume Polynomial (VSPOLY/1 ) in m 3 /kmol V*(T) = C 1 + C 2T + C 3T 2 + C 4T 3 + C 5T 4 for C 6 T C 7 9

10 Table 4: Solid heat capacity* CPSPO1 CAL- ACON CELLULOS ENZYME LGNSOL LIGNIN XYLAN YEAST C C C C C C C C * Solid Heat Capacity (CPSPO1/1 8) in J/kmol.K Cp*(T) = C 1 + C 2T + C 3T 2 + C 4/T+ C 5/T 2 + C 6/ T for C 7 T C 8 It is important to mention that when opening a simulation in a different version from that in which it was originally built, sometimes loss of properties may occur. In these cases, it is recommended to open the simulation on the version of the simulator in which it was constructed and check the Review table of properties to copy and insert the missing data to the new version of the simulation. 3 Aspen Plus Property methods Alongside with the definition of the chemical compounds, defining the property methods that will be used is an important step when beginning a new simulation. Property methods are a collection of methods and models that are used on Aspen Plus to estimate thermodynamic and transport properties. (ASPENTECH, 2009) According to the Aspentech (2009), the thermodynamic properties estimated by the simulator are: Fugacity coefficient (K-values) Enthalpy Entropy Gibbs free energy Volume The transport properties are: 10

11 Viscosity Thermal conductivity Diffusion coefficient Surface tension When selecting property methods for a simulation, it is necessary to take into account different factors such as the chemical function of the components (polar, non-polar, electrolyte), process pressure and temperature, physical state, among others. Property methods used on VSB simulations As previously explained for the chemical compounds selected for the construction of the simulations, property methods may also differ depending on specific details of the process that is going to be simulated; the following description of selected methods applies to simulations with standard configurations for first and second generation ethanol production. - NRTL-RK: The NRTL (Non-random Two-Liquid) model estimates activity coefficients for the liquid phase. The Redlich-Kwong (RK) equation of state is used to estimate the properties of the vapor phase for sections of the plant with moderate operating pressures (up to 10 bar) and high sugar concentrations where deviations from the ideal behavior for the vapor phase may be expected. - NRTL-HOC: For sections of the plant where short chain organic acids are present in higher concentrations (fermentation, distillation), Hayden O'Connell (HOC) was selected for calculation of the properties of the vapor phase to account for possible dimerization of these acids. - RKS-BM (Redlich-Kwong-Soave and Boston-Mathias alpha function): According to Aspentech (2009), this model is suitable for representing combustion processes. Therefore, this model was selected for the boiler where biomass is burnt. - STEAMNBS: This model was selected for processes that involved expansion or compression of steam. Aspentech (2009) indicates this model for steam cycles, turbines and compressors. In the VSB, processes involving steam cycles are found on the cogeneration, heat and power (CHP) sector of the plant. 11

12 4 Final Remarks This technical memorandum is useful for those that are beginning to build a process model for a sugarcane-based biorefinery or similar process, e.g. using other feedstocks or producing different products. In addition, this organized information support future publications including process simulations carried out within the VSB. The proper definition of feedstock composition, physicochemical properties and thermodynamic models are essential to obtain reliable mass and energy balances for different routes and technologies, allowing evaluation of different biorefinery alternatives. References ASPENTECH. Physical Property Methods CHEMSPIDER, 2012, Chemspider Databank. Available at < Accessed in: March COLBY COLLEGE, 2012, Computational Chemistry Lab. Available at: < Accessed in: March MeT 22. Sugarcane biomass composition for the industrial simulations in the Virtual Sugarcane Biorefinery (VSB). Campinas, CNPEM, MORAIS, E. R. et al. Biorefinery alternatives. In: BONOMI, A.; CAVALETT, O.; CUNHA, M. P.; LIMA, M. A. P. (eds). Virtual Biorefinery: An Optimization Strategy for Renewable Carbon Valorisation. Springer, NREL, 2011, Process Design for Biochemical Conversion of Biomass to Ethanol DW1111A Aspen Plus simulation file. Avaiable at: < Accessed in: March, SCHERRER, R.; BERLIN, E.; GERHARDT, P. Density, Porosity, and Structure of Dried Cell Walls Isolated from Bacillus megaterium and Saccharomyces cerevisiae. J Bacteriol 129(2):1162-4,

13 STANMORE, B. R. Generation of Energy from Sugarcane Bagasse by Thermal Treatment. Waste Biomass Valorization 1(1):77-89, doi: /s WOOLEY, R.J.; PUTSCHE, V., 1996, Development of an Aspen Plus Physical Property Database for Biofuels Components. NREL/MP National Renewable Energy Laboratory. 13