Supercritical fluid extraction of oil sand bitumens from the Uinta Basin, Utah

The supercritical fluid extraction (SFE) of bitumens from the Whiterocks, Asphalt Ridge, PR Spring and Sunnyside oil sand deposits has been investigated in a semicontinuous system. The extraction experiments were conducted with the Asphalt Ridge and Sunnyside bitumens at five different operating conditions using commercial propane as the solvent. The results indicted that the cumulative extraction yields increased with an increase in pressure at constant temperature and decreased with increase in temperature at constant pressure. The extraction yields increased with an increase in solvent density. The composition of the feedstock was a major factor in controlling the extraction yields. The four bitumens from the Uinta Basin, Utah, varied significantly in their physical and chemical properties. The propane extraction yields were inversely proportional to the bitumen asphaltene content and directly proportional to the bitumen resin content. The cumulative extraction yields increased with an increase in bitumen volatility and saturates and aromatics contents for the Whiterocks, PR Spring and Sunnyside bitumens. The asphaltenes appeared to concentrate in the residual fraction and were not extracted. Furthermore, they hindered the extraction of other solubility classes. The extracted phases were upgraded relative to the bitumens as indicated by their volatilities. The volatilities of the extract phases were considerably higher than those of the bitumens. The fractionation of the residual fractions into solubility fractions indicated that saturates and aromatics were preferentially extracted from the bitumen relative to the asphaltenes and resins. This phenomenon was confirmed by the reduction in the measured hydrogen/carbon ratios of the residual fractions. The SFE of Asphalt Ridge and Sunnyside bitumen was modeled using continuous thermodynamics principles and the Peng-Robinson equation of state. A process flow diagram was suggested to upgrade bitumens using supercritical fluid extraction and separation technology. Suitable operating conditions such as pressure, temperature and solvent-to-feed ratio were identified for the proposed extraction and separation process concept. The modeling successfully fit the experimental observations. A high temperature simulated distillation technique was developed along with a software to extend the ASTM D2887 and D5307 techniques to estimate the boiling point distribution of heavy oils from 811 K to 973 K.

SUPERCRITICAL FLUID EXTRACTION OF OIL SAND BITUMENS FROM THE UINTA BASIN, UTAH by ? Murugesan Subramanian A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Chemical and Fuels Engineering The University of Utah December 1996 ABSTRACT The supercritical fluid extraction (SFE) of bitumens from the Whiterocks, Asphalt Ridge, PR Spring and Sunnyside oil sand deposits has been investigated in a semicontinuous system. The extraction experiments were conducted with the Asphalt Ridge and Sunnyside bitumens at five different operating conditions using commercial propane as the solvent. The results indicted that the cumulative extraction yields increased with an increase in pressure at constant temperature and decreased with increase in temperature at constant pressure. The extraction yields increased with an increase in solvent density. The composition of the feedstock was a major factor in controlling the extraction yields. The four bitumens from the Uinta Basin, Utah, varied significantly in their physical and chemical properties. The propane extraction yields were inversely proportional to the bitumen asphaltene content and directly proportional to the bitumen resin content. The cumulative extraction yields increased with an increase in bitumen volatility and saturates and aromatics contents for the Whiterocks, PR Spring and Sunnyside bitumens. The asphaltenes appeared to concentrate in the residual fraction and were not extracted. Furthermore, they hindered the extraction of other solubility classes. The extracted phases were upgraded relative to the bitumens as indicated by their volatilities. The volatilities of the extract phases were considerably higher than those of the bitumens. The fractionation of the residual fractions into solubility fractions indicated that saturates and aromatics were preferentially extracted from the bitumen relative to the asphaltenes and resins. This phenomenon was confirmed by the reduction in the measured hydrogen/carbon ratios of the residual fractions. The SFE of Asphalt Ridge and Sunnyside bitumen was modeled using continuous thermodynamics principles and the Peng-Robinson equation of state. A process flow diagram was suggested to upgrade bitumens using supercritical fluid extraction and separation technology. Suitable operating conditions such as pressure, temperature and solvent-to-feed ratio were identified for the proposed extraction and separation process concept. The modeling successfully fit the experimental observations. A high temperature simulated distillation technique was developed along with a software to extend the ASTM D2887 and D5307 techniques to estimate the boiling point distribution of heavy oils from 811 K to 973 K. v TABLE OF CONTENTS ABSTRACT................................................................................................................ iv LIST OF FIGURES................................................................................................... ix LIST OF TABLES.................................................................................................... xvii N O M EN C LA TU R E............................................................................................... xxii ACKNOWLEDGEMENTS.......................................................................................xxv Chapter I. INTRODUCTION.................................................................................................. 1 Worldwide and USA Oil Sands Resources.................................................. 1 Nature of Bitumen........................................................................................... 6 Recovery Processes.......................................................................................9 Research Objectives.................................................................................... 13 II. LITERATURE REVIEW.................................................................. ...................15 Supercritical Fluid Extraction Fundamentals.............................................. 15 Experimental Observations.............................................................. 16 Theory of Supercritical Fluid Extraction..........................................17 Applications of Supercitical Fluid Extraction.............................................. 28 Phase Behavior Studies at Supercritical Conditions................................. 31 Supercritical Fluid Extraction of Oil Sands................................................. 31 Commercial Supercritical Fluid Extraction Processes............................... 35 ROSE Process.................................................................................. 35 HSC-ROSE Process..........................................................................42 Demex Process................................................................................. 42 SOLVAHL Process............................................................................ 48 Modeling the Bitumen Extraction Process................................................. 52 Pseudocomponent Lumping Scheme.............................................. 53 Continuous Thermodynamics...........................................................57 Gamma Distribution...........................................................................68 Beta Distribution................................................................................ 70 SemiContinuous Mixtures................................................................. 72 Flash Calculations............................................................................. 76 Quadrature Method........................................................................... 81 Equations of State............................................................................. 83 Perturbed Hard Chain Equation of State......................................... 84 Phase Behavior Studies on Bitumen Systems........................................... 96 Modeling Approach.....................................................................................130 III. EXPERIMENTAL APPARATUS AND PROCEDURES................................. 132 Supercritical Fluid Supply System............................................................ 132 Supercritical Fluid Extractor and Densitometer Assembly......................136 Data Acquisition System............................................................................ 138 SFE Separator Assembly........................................................................... 139 Calibration of the Densitometer................................................................ 140 Oil Sands Bitumen Preparation................................................................. 141 Experimental Procedures........................................................................... 145 Product Analysis.........................................................................................148 Liquid Product Analysis...................................................................148 Gas Analysis................................................................................... 156 IV. RESULTS AND DISCUSSION........................................................................158 Feedstock Characterization.......................................................................158 Physical Properties.........................................................................160 Chemical Properties........................................................................164 Simulated Distillation................................................................................. 168 Discussion........................................................................................171 Findings............................................................................................186 Preliminary Process Experiments............................................................. 199 Supercritical Fluid Extraction of Oil Sands Bitumens.............................. 200 Supercritical Fluid Extraction of Asphalt Ridge Bitumen.........................210 Pressure Effect................................................................................ 211 Temperature Effect..........................................................................217 Solvent Density Effect.....................................................................220 Carbon Number Distribution for AR Bitumen Extract Phases..... 224 Reproducibility................................................................................. 230 Supercritical Fluid Extraction of Sunnyside Bitumen.............................. 230 Pressure Effect................................................................................ 233 Temperature Effect..........................................................................238 Solvent Density Effect.....................................................................241 Carbon Number Distribution for SS Bitumen Extract Phases..... 244 Reproducibility................................................................................. 250 Comparison of SFE of Bitumens from Uinta B asin................................. 250 Pressure Effect................................................................................ 253 Temperature Effect..........................................................................256 Solvent Density Effect.................................................................... 259 vii Effect of Bitumen Asphaltene Content........................................... 262 Effect of Bitumen Resin Content....................................................266 Effect of Bitumen Saturate and Aromatics Content......................267 Compositional Analyses of Residual Fraction.......................................... 275 Saturates Content of Residual Fractions......................................275 Asphaltene Content of Residual Fractions....................................289 Elemental Analyses.....................................................................................300 Modeling SFE Using Continuous Thermodynamics Principle................ 310 Choice of Continuous Distribution Function................................. 310 Calculation Procedure Using Quadrature M ethod.......................314 Modeling Supercritical Extraction Process............................................... 325 Primary Extractor............................................................................. 329 Separator 1........................................................................................345 V. CONCLUSIONS............................................................................................... 378 Appendices A. SUPERCRITICAL FLUID EXTRACTION DATA............................................ 381 B. SIMULATED DISTILLATION DATA............................................................... 388 C. ANALYTICAL TEST PROCEDURE FOR SARA ANALYSIS........................404 D. FIGURES PERTAINING TO MODELING.......................................................411 E. SIMULATED DISTILLATION SOFTWARE CODE........................................426 REFERENCES.......................................................................................................483 VITA.........................................................................................................................502 viii Copyright© Murugesan Subramanian 1996 All Rights Reserved LIST OF FIGURES Figure Page 1.1 Schematic of Various Processes for Bitumen Extraction from Oil Sands............................................................................................................. 10 2.1 Effect of Temperature on Enhancement Factor For Propane and n-Butane and its Binary Mixtures with n-Decane.......................................23 2.2. Effect of Temperature on the Enhancement Factor for PropaneDecane and Butane-Decane Binary Systems............................................ 26 2.3 Schematic of the Demex Process............................................................... 45 2.4. Schematic of the SOLVAHL Process.......................................................... 49 2.5. Representation of Discrete and Continuous Distributions for Complex Hydrocarbon Mixtures.................................. ................................61 2.6 Typical Gamma and Beta Distribution Functions.......................................73 2.7 Schematic for Flash Calculation of Continuous Mixtures..........................77 2.8 Feed Distribution for Oil and Resin used by Cotterman............................91 2.9 Molecular Weight Distributions for Oil and Resin Fractions in Vapor and Liquid as Predicted by Cotterman for Mixture at 8 MPa (Pr=1.9) and 398 K (Tr=1.0 8 )........................................................... 93 2.10 Measured and Predicted Vapor Phase Composition of a Saturates-rich O il.......................................................................................... 97 2.11 Measured and Predicted Vapor Phase Composition of an Aromatics-rich Oil..........................................................................................99 2.12 Solubility of Carbon Dioxide in Athabasca Bitumen................................ 105 2.13 Solubility of Carbon Dioxide in Peace River Bitumen............................. 107 2.14 Comparison of Carbon Dioxide Solubility in Cold Lake Bitumen and Saturated Liquid Density by Modified Martin EOS with Measured Values.......................................................................................................... 113 2.15 Measured and PHC Predicted Weight Fraction of C 0 2 in Bitumen...... 117 2.16 Measured, Soave and PR Predicted Weight Fraction of C 0 2 in Bitumen........................................................................................................119 2.17 Measured and PHC Predicted Weight Fraction of Bitumen in C 0 2 ...... 121 2.18 Measured, Soave and PR Predicted Weight Fraction of Bitumen in C 0 2........................................................................................................... 123 2.19 Comparison Between Experimental and Predicted Extracted Phase Composition for PR Spring Bitumen-Propane System at 10.4 MPa (P,=1.2) and 380 K (T,=1.0 3 )............................................................ 127 3.1 Schematic of the Supercritical Fluid Extraction System..........................133 3.2 Schematic of the Gas Chromatograph System........................................149 3.3 Carbon Number (C5 to C90) versus Retention Time Calibration Curve for Simulated Distillation Analysis................................................. 153 4.1 Relationship Between Temperature and Viscosity for the Bitumens.... 161 4.2 Comparison of the Solubility Fractions of the Bitum ens.........................165 4.3 Chromatograms for Whiterocks Bitumen................................................. 174 4.4 Boiling Point Distribution for Whiterocks Bitum en.................................. 176 4.5 Chromatograms for the Whiterocks Bitumen Extract.............................. 182 4.6 Boiling Point Distribution for the Whiterocks Bitumen Extract............... 184 4.7 Chromatograms for the Whiterocks Bitumen Residual Fraction............187 4.8 Boiling Point Distribution for the Whiterocks Bitumen Residual Fraction........................................................................................................189 4.9 Carbon Number Distributions for the Whiterocks Bitumen and the Saturates, Aromatics and Resins Solubility Fractions............................191 x 4.10 Carbon Number Distributions for the Asphalt Ridge Bitumen and the Saturates, Aromatics and Resins Solubility Fractions......................193 4.11 Carbon Number Distributions for the PR Spring Bitumen and the Saturates, Aromatics and Resins Solubility Fractions............................ 195 4.12 Carbon Number Distributions for the Sunnyside Bitumen and the Saturates, Aromatics and Resins Solubility Fractions..................... 197 4.13 Effect of Solvent Flowrate on the Attainment of Thermodynamic Equilibrium for Hexadecane-C02 system at 10.4 MPa (Pr=1.41) and 311 K (Tr=1.02)........................................................................................... 202 4.14 Operating Conditions for SFE of Bitumens using Propane as Solvent.........................................................................................................205 4.15 Propane Density at Various Temperatures and Pressures.................... 208 4.16 Effect of Pressure on SFE Yields with Asphalt Ridge Bitumen..............212 4.17 Measured Extract Phase Density During SFE of Asphalt Ridge Bitumen with Propane as Solvent............................................................. 214 4.18 Effect of Temperature on SFE Yields with Asphalt Ridge Bitumen....... 218 4.19 Effect of Solvent Density on SFE Yields with Asphalt Ridge Bitumen........................................................................................................221 4.20 Carbon Number Distributions for the Asphalt Ridge Bitumen, Extracts and Residual Fractions Obtained from SFE at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03)......................................................................225 4.21 Effect of Pressure and Temperature on the Carbon Number Distributions of the Second Extraction Window Obtained during SFE of the Asphalt Ridge Bitumen............................................................ 228 4.22 Reproducibility for SFE with the Asphalt Ridge Bitumen at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92)............................................................ 231 4.23 Effect of Pressure on SFE with Sunnyside Bitumen............................... 234 4.24 Measured Extract Phase Densities during SFE of the Sunnyside Bitumen with Propane as Solvent............................................................. 236 4.25 Effect of Temperature on the SFE Yields with the Sunnyside Bitumen........................................................................................................ 239 4.26 Effect of Solvent Density on Extraction Yields With the Sunnyside Bitumen........................................................................................................242 4.27 Carbon Number Distributions for the Sunnyside Bitumen and the Extract and Residual Fractions Obtained from SFE at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03)......................................................................246 4.28 Effect of Pressure and Temperature on the Carbon Number Distribution of the Second Extraction Windows obtained from SFE with the Sunnyside Bitumen.......................................................................248 4.29 Reproducibility for SFE with Sunnyside Bitumen at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92)......................................................................251 4.30 Effect of Pressure on the Extraction Yields for the Four Bitumens from the Uinta Basin at Constant Temperature 380 K (Tr=1.0 3 )..........254 4.31 Effect of Temperature on the Extraction Yields for the Four Bitumens from the Uinta Basin at Constant Pressure 10.4 MPa (Pr=2.3)........................................................................................................257 4.32 Effect of Solvent Density on the Extraction Yields for the Four Bitumens from Uinta Basin.........................................................................260 4.33 Relationship Between Asphaltene Content and Extraction Yield for the Four Uinta Basin Bitumens................................................................. 264 4.34 Relationship Between Resin Content and Extraction Yield for the Four Uinta Basin Bitumens........................................................................268 4.35 Relationship Between Saturates Content and Extraction Yield for the Four Uinta Basin Bitumens................................................................. 272 4.36 Effect of Solvent Density on the Extraction of Saturates and Aromatics from the Whiterocks Bitumen.................................................. 280 4.37 Effect of Solvent Density on the Extraction of Saturates and Aromatics from the PR Spring Bitumen.....................................................282 4.38 Effect of Solvent Density on the Extraction of Saturates from the Asphalt Ridge Bitumen............................................................................... 285 4.39 Effect of Solvent Density on the Extraction of Saturates from the Sunnyside Bitumen..................................................................................... 287 4.40 Effect of Solvent Density on the Extraction of Asphaltenes from the Whiterocks Bitumen....................................................................................290 4.41 Effect of Solvent Density on the Extraction of Asphaltenes from the Asphalt Ridge Bitumen............................................................................... 292 4.42 Effect of Solvent Density on the Extraction of Asphaltenes from the PR Spring Bitumen......................................................................................294 4.43 Effect of Solvent Density on the Extraction of Asphaltenes from the Sunnyside Bitumen.....................................................................................296 4.44 Relationship Between Pure Solvent Density and the Residual Fraction H/C Atomic Ratio for the Whiterocks Bitumen..........................302 4.45 Relationship Between Pure Solvent Density and the Residual Fraction H/C Atomic Ratio for the Asphalt Ridge Bitumen..................... 304 4.46 Relationship Between Pure Solvent Density and the Residual Fraction H/C Atomic Ratio for the PR Spring Bitumen............................ 306 4.47 Relationship Between Pure Solvent Density and the Residual Fraction H/C Atomic Ratio for the Sunnyside Bitumen...........................308 4.48 Continuous Molecular Weight Distribution Function for the Asphalt Ridge Bitumen Using a Gamma Distribution............................................ 312 4.49 Continuous Molecular Weight Distribution Function for the Asphalt Ridge Bitumen Using a Higher Order Polynomial Function................... 315 4.50 Relationship Between Molecular Weight and Boiling Point (Alkane Based).......................................................................................................... 319 4.51 Optimization Results to Choose the Number of Quadrature Points...... 323 4.52 Process Flow Diagram for a Supercritical Fluid Extraction and Separation Process for Upgrading Bitumens........................................... 326 4.53 Simulated Extraction Yield Results for the Asphalt Ridge Bitumen at 339 K (Tr=0.92).......................................................................................333 4.54 Simulated Extraction Yield Results for the Asphalt Ridge Bitumen at 380 K (Tr=1.03).......................................................................................335 4.55 Simulated Extraction Yield Results for the Asphalt Ridge Bitumen at 422 K (Tr=1.14).......................................................................................337 4.56 Simulated Extraction Yield Results for the Sunnyside Bitumen at 339 K (Tr=0.92)........................................................................................... 339 4.57 Simulated Extraction Yield Results for the Sunnyside Bitumen at 380 K (Tr=1.0 3 )........................................................................................... 341 4.58 Simulated Extraction Yield Results for the Sunnyside Bitumen at 422 K (Tr=1.14)........................................................................................... 343 4.59 Continuous Distributions for the Saturates, Aromatics and Resin Ensembles for the Asphalt Ridge Bitumen Using a Higher Order Polynomial Function................................................................................... 347 4.60 Continuous Distributions for the Saturates, Aromatics and Resin Ensembles for the Sunnyside Bitumen Using a Higher Order Polynomial Function................................................................................... 349 4.61 Predicted Selectivity of Saturates and Aromatics over Resins in the Vapor Phase for the AR Bitumen.............................................................. 356 4.62 Predicted Selectivity of Saturates over Resins in the Vapor Phase for the AR Bitumen......................................................................................358 4.63 Predicted Selectivity of Aromatics over Resins in the Vapor Phase for the AR Bitumen......................................................................................360 4.64 Predicted Selectivity of Resins over Saturates and Aromatics in the Liquid Phase for the AR Bitumen.............................................................. 362 4.65 Predicted Selectivity of Saturates and Aromatics over Resins in the Vapor Phase for the Sunnyside Bitumen................................................. 364 4.66 Predicted Selectivity of Saturates over Resins in the Vapor Phase for the Sunnyside Bitumen.........................................................................366 4.67 Predicted Selectivity of Aromatics over Resins in the Vapor Phase for the Sunnyside Bitumen.........................................................................368 xiv 4.68 Predicted Selectivity of Resins over Saturates and Aromatics in the Liquid Phase for the Sunnyside Bitumen..................................................370 4.69 Comparison of the Molecular Weight Distribution of the Saturates, Aromatics and Resins for the Asphalt Ridge and Sunnyside Bitumens......................................................................................................374 C.1 Schematic of the Adsorption Chromatography Technique......................409 D.1 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Asphalt Ridge Bitumen at 339 K (Tr=0.92) using Separate Solubility Ensembles................................................................. 414 D.2 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Asphalt Ridge Bitumen at 380 K (Tr=1.03) using Separate Solubility Ensembles................................................................. 415 D.3 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Asphalt Ridge Bitumen at 422 K (Tr=1.14) using Separate Solubility Ensembles................................................................. 416 D.4 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Sunnyside Bitumen at 339 K (Tr=0.92) using Separate Solubility Ensembles.................................................................................. 417 D.5 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Sunnyside Bitumen at 380 K (Tr=1.03) using Separate Solubility Ensembles...................................................................................418 D.6 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Sunnyside Bitumen at 422 K (Tr=1.14) using Separate Solubility Ensembles.................................................................................. 419 D.7 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor Phase for Asphalt Ridge Bitumen at 339 K (Tr=0.92) using Separate Solubility Ensembles................................................................. 420 D.8 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor Phase for Asphalt Ridge Bitumen at 380 K (Tr=1.03) using Separate Solubility Ensembles................................................................. 421 D.9 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor xv Phase for Asphalt Ridge Bitumen at 422 K (Tr=1.14) using Separate Solubility Ensembles................................................................. 422 D.10 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor Phase for Sunnyside Bitumen at 339 K (Tr=0.92) using Separate Solubility Ensembles...................................................................................423 D.11 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor Phase for Sunnyside Bitumen at 380 K (Tf=1.03) using Separate Solubility Ensembles...................................................................................424 D.12 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor Phase for Sunnyside Bitumen at 422 K (Tr=1.14) using Separate Solubility Ensembles.................................................................................. 425 xvi LIST OF TABLES Table Page 1.1 Estimated Oil Sands Resources W orldwide................................................ 3 1.2 Oil Sands Resources of the United States....................................................4 1.3 Estimate of Oil Sands Resources of the State of U ta h ............................. 5 1.4 Physical and Chemical Properties of Uinta Basin Bitumens......................7 2.1 Comparison of the Phenanthrene Concentration in Vapor Phase for Various Solvents.......................................................................... 18 2.2 Applications of Supercritical Extraction...................................................... 29 2.3 Quality of Yield Obtained from ROSE Unit for Lagunillas Vacuum Resid.............................................................................................................. 37 2.4 Quality of Yield Obtained from ROSE-SR Unit for Coal Liquefaction Resid from Bituminous C o a l........................................................................38 2.5 Quality of Yield Obtained from ROSE Unit for Resid from Retorting Oil Bearing Diatomacious Earth Using Iso-Butane as S olvent................ 39 2.6 Quality of Product and Yield Obtained from ROSE Unit for nButane and n-Pentane Extractions of Athabasca Oil Sands Bitumen (623 K Plus Resid)......................................................................... 40 2.7 Quality of Product and Yield Obtained from ROSE Unit Using Zuata Crude as Feed and Propane as Solvent.......................................... 41 2.8 Quality of Syncrude Produced by the HSC-ROSE Process..................... 43 2.9 Products Yield and Quality from the Demex Process............................... 47 2.10 Yield and Quality for the Arabian Light Vacuum Resid Using the SOLVAHL Process................................................................................. 51 2.11 Theoretical Number of Isomers for Paraffins............................................. 54 2.12 Various Single Variable Distribution Functions.......................................... 69 2.13 Properties of Saturates and Aromatics-Rich Oils Used by Cotterman................................................................................................. 95 2.14 Pseudocomponents for Athabasca and Peace River Bitumen and their Estimated Properties.......................................................................... 102 2.15 Estimated Critical Properties of Athabasca Bitumen Pseudocomponents................................................................................... 103 2.16 Estimated Critical Properties of Peace River Bitumen Pseudocomponents....................................................................................104 2.17 Critical Parameters Used by Lu et al. to Predict Solubility of C 02 in Athabasca Bitumen.........................................................................110 2.18 Properties of C 02, Heavy Oils and Bitumen Used by Jamaluddin et a l..........................................................................................112 2.19 Estimated Properties of Bitumen Pseudocomponents............................ 116 2.20 Comparison Between Predicted and Experimentally Obtained Extract Phase Compositions for Whiterocks Bitumen Using Propane as Solvent at 10.4 MPa (Pr=2.3) and 380 K (Tr=1.03).............126 3.1 Temperature Programs for Determination of the Residual Toluene Content in Bitumen by Gas Chromatography........................................... 144 3.2 Temperature Programs Used to Obtain Carbon Number Distributions of Bitumen and Bitumen Products by Gas Chromatography.........................................................................................152 3.3 Analyses of the Gases Used as SFE Solvents........................................157 4.1 Physical and Chemical Properties of Uinta Basin Bitumens.................. 159 4.2 The Eact for Viscous Flow for Four Bitumens from Uinta Basin (Utah)...........................................................................................................163 4.3 Comparison of the Extended Method Results..........................................179 4.4 Simulated Distillation Analyses for Bitumens Analyzed..........................180 xviii 4.5 Results of the Hexadecane-Carbon Dioxide Experiments Performed at 10.4 MPa (Pr=1.41) and 311 K (Tr=1.02)........................... 201 4.6 Measured Densities of Commercial Propane........................................... 207 4.7 Comparison of Boiling Fractions for Four Bitumens............................... 271 4.8 Summary of Extraction Yields and Residual Fractions Analyses for the Whiterocks Bitumen............................................................ .................276 4.9 Summary of Extraction Yields and Residual Fractions Analyses for the Asphalt Ridge Bitumen.........................................................................277 4.10 Summary of Extraction Yields and Residual Fractions Analyses for the PR Spring Bitumen............................................................................... 278 4.11 Summary of Extraction Yields and Residual Fractions Analyses for the Sunnyside Bitumen.............................................................................. 279 4.12 Quadrature Points and Weight Factor for Gauss-Legendre Integration................................................................................................... 318 4.13 Quadrature Components Properties for Optimization of Number of Components for Asphalt Ridge Bitumen...................................................322 4.14 Properties of the Asphalt Ridge Bitumen at the Quadrature Points...... 330 4.15 Properties of the Sunnyside Bitumen at the Quadrature Points.............331 4.16 Properties of the Saturates Fraction at the Quadrature Points for the Asphalt Ridge Bitumen......................................................................... 352 4.17 Properties of the Aromatics Fraction at the Quadrature Points for the Asphalt Ridge Bitumen.........................................................................353 4.18 Properties of the Resins Fraction at the Quadrature Points for the Asphalt Ridge Bitumen............................................................................... 354 A.1 Extraction Yields for Supercritical Fluid Extraction of Uinta Basin Bitumens at 5.6 MPa (Pr=1.2) and 380 K (Tr=1.03)................................. 382 A.2 Extraction Yields for Supercritical Fluid Extraction of Uinta Basin Bitumens at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92)............................... 383 xix A.3 Extraction Yields for Supercritical Fluid Extraction of Uinta Basin Bitumens at 10.4 MPa (Pr=2.3) and 380 K (Tr=1.03)............................... 384 A.4 Extraction Yields for Supercritical Fluid Extraction of Uinta Basin Bitumens at 10.4 MPa (Pr=2.3) and 422 K (Tr=1.14)............................... 385 A.5 Extraction Yields for Supercritical Fluid Extraction of Uinta Basin Bitumens at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03)............................... 386 A .6 Experimental Reproducibility Results for Supercritical Fluid Extraction of Asphalt Ridge and Sunnyside Bitumens at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92)......................................................................387 B.1 Cumulative Carbon Number and Boiling Point Distribution for Uinta Basin Bitumens (in Cumulative Weight Fractions).................................. 389 B.2 Cumulative Carbon Number and Boiling Point Distribution of Solubility Fractions (Cumulative Weight Fractions)................................ 390 B.3 Cumulative Carbon Number and Boiling Point Distribution of Solubility Fractions (Cumulative Weight Fraction).................................. 391 B.4 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Sunnyside Bitumen at 5.6 MPa (Pr=1.2) and 380 K (Tr=1.03) (Cumulative Weight Fraction)........................................392 B.5 Carbon Number and Boiling Point Distribution of Extracts from SFE of Sunnyside Bitumen at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92) (Cumulative Weight Fraction)....................................................................393 B.6 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Sunnyside Bitumen at 10.4 MPa (Pr=2.3) and 380 K (Tr=1.03) (Cumulative Weight Fraction)........................................394 B.7 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Sunnyside Bitumen at 10.4 MPa (Pr=2.3) and 422 K (Tr=1.14) (Cumulative Weight Fraction)........................................395 B.8 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Sunnyside Bitumen at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03) (Cumulative Weight Fraction)........................................396 B.9 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Asphalt Ridge Bitumen at 5.6 MPa (Pr=1.2) and 380 K (Tr=1.03) (Cumulative Weight Fraction)................................ 397 xx B.10 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Asphalt Ridge Bitumen at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92) (Cumulative Weight Fraction)................................ 398 B.11 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Asphalt Ridge Bitumen at 10.4 MPa (Pr=2.3) and 380 K (Tr=1.03) (Cumulative Weight Fraction)................................ 399 B.12 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Asphalt Ridge Bitumen at 10.4 MPa (Pr=2.3) and 422 K (Tr=1.14) (Cumulative Weight Fraction)................................ 400 B.13 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Asphalt Ridge Bitumen at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03)....................................................................................401 B.14 Cumulative Carbon Number and Boiling Point Distribution of Residual Fractions Obtained from SFE of Sunnyside Bitumen (Cumulative Weight Fraction)....................................................................402 B.15 Cumulative Carbon Number and Boiling Point Distribution of Residual Fractions Obtained from SFE of Asphalt Ridge Bitumen (Cumulative Weight Fraction)................................................................... 403 xxi NOMENCLATURE Constants, Equations of State and Beta Distribution Constants, Equations of State and Beta Distribution PHC EOS Parameter Lower Bound of the Independent Variable i Upper Bound of the Independent Variable il Cumulative Distribution Continuous Distribution Function Fugacity of Component i in the Liquid Phase Fugacity of Component i in the Vapor Phase Characterizing Variable in Continuous Thermodynamics Initial Boiling Point, K Boltzmann Constant, erg/K Watson Characterization Factor Equilibrium Ratio(Yj/Xj) of the Component i Total Number of Moles in Liquid Phase, mol Fugacity Coefficient of Component i in Liquid Phase Molecular Weight of the ith Component, g/gmol Molecular Weight, g/gmol 7 (used in Whitson Correlation) Residual Helmholtz Energy, j Number of Moles in Phase i, mol Number of Components in a Hydrocarbon Mixture Nq Number of Pseudocomponents Nt Total Number of Moles in Overall Composition, mol P Pressure, MPa P° Reference State Pressure, 1 bar Pc Critical Pressure, MPa Pr Reduced Pressure, P/Pc Q External Molecular Surface Area, cm2/Mol T Reduced Temperature used in Perturbed Hard Chain Equation of State, T/T T* Characteristic Temperature, K T Temperature, K Tc Critical Temperature, K Tr Reduced Temperature, T/Tc Vc Critical Molar Volume, cm3/mol v Reduced Molar Volume used in Perturbed Hard Chain Equation of State, VA/' V Soft Core Volume, cm3/mol V* Hard Core Volume, cm3/mol V Total Number of Moles in Vapor Phase, mol Vi Fugacity Coefficient of Component i in Vapor Phase Xj Mole Fraction of Component i in Liquid Phase Yj Mole Fraction of Component i in Vapor Phase Z Compressibility Factor Zj Overall Mole Fraction of Component i co Acentric Factor Hi Chemical Potential of Phase i xxiii Interaction Parameter, Gamma and Beta Distribution Function Constants Gamma and Beta Distribution Function Constants Mean used in Gamma Distribution Function Variance in Gamma Distribution Function Dirac Delta Function Mole Fraction of the Continuous Ensemble Fraction Vaporized Potential Energy Per Unit Area, erg/cm2 ACKNOWLEDGMENTS I am deeply indebted to my wife Santhi who assisted me in the course of my studies. My special love and appreciation are extended to my son, Satish. I thank my parents for their constant encouragement and dedicate this dissertation to them. I would like to express my deep appreciation to Professor Francis V. Hanson, as the chair of my supervisory committee for his assistance, encouragement and advice throughout the course of this work in making this project a success. My special gratitude goes to Professor Milind D. Deo and Professor John. V. Fletcher for their assistance. The other supervisory research committee members, Dr. James G. Speight of Western Research Institute and Dr. Dennett Netterville of Syncrude Canada Limited, are also thanked for their efforts. The author would like to thank Dr. Daniel C. Longstaff and Dr. Deepak A. Deshpande for reviewing this dissertation. I would also like to acknowledge all my colleagues in the University of Utah Oil Sands Research and Development Group for many beneficial discussions. The financial support of the U.S Department of Energy through the Laramie Project Office of the Morgantown Energy Technology Center and of the Mineral Leasing Fund of the State of Utah is gratefully acknowledged. CHAPTER 1 INTRODUCTION World crude oil reserves are diminishing due to increased consumption of petroleum products. New crude discoveries have not kept pace with the increased consumption. Petroleum products have been replaced by natural gas in process heating, fertilizer manufacture and to some extent as a motor vehicle fuel. Low crude oil and natural gas prices have limited the possibilities for producing petroleum products from oil sands bitumens without substantial government subsides. The hydrocarbon resources present in the oil sands deposits throughout the world have been estimated to be in excess of 3.0 trillion barrels of petroleum equivalent^ ,2], These numbers are significant compared to the conventional petroleum resources which are estimated to be 1.5 trillion barrels[2]. Oil sands deposits will eventually be developed once the technical and economic barriers have been reduced. Worldwide and USA Oil Sands Resources Bitumen or ultra heavy oil is present in sandstone reservoirs in a number of countries worldwide. Major oil sands deposits have been reported in Canada, Colombia, the former Soviet Union, the United States and Venezuela[1]. Small to medium oil sands deposits have been identified in Albania, Italy, Madagascar, 2 Romania and Trinidad-Tobago[1], Canada is the only country producing hydrocarbon liquids from oil sands reservoirs on a commercial scale. The hydrocarbon liquid production from oil sands reservoirs was reported to account for 16 % of the total Canadian oil production^]. The estimated worldwide oil sands resources[2] are presented in Table 1.1. The United States ranks fifth in the amount of bitumen in-place[2-6]. Major oil sands deposits have been reported in Alabama, California, Kentucky, Texas and Utah. Small deposits have also been identified in 17 other states[2]. The estimated domestic oil sands resources are presented in Table 1.2. Most of the studies[7-12] carried out at the University of Utah on oil sands have focused on the deposits in the State of Utah. The estimated bitumen in place[13] for the various Utah deposits is reported in Table 1.3. Major oil sands deposits have been identified in the Uinta Basin and in the central-southeast region of Utah. The physical and chemical properties of oil sands bitumens are significantly different from those of conventional crude oils[2,4,5]. The American Petroleum Institute (API) gravity of conventional crude oil is usually greater than 20°API with a viscosity less than 1000 cP at reservoir conditions[2]. The API gravity of heavy crude oil ranges from 10 to 20° API with a maximum viscosity of 10,000 cP[2]. The oil sands bitumens are much heavier than conventional and heavy crude oils with API gravities normally less than 10° API and viscosities higher than 10,000 cP. The asphaltene (i.e., pentane insolubles) and resin contents of bitumens are higher than crude oils and consequently 3 Table 1.1 Estimated Oil Sands Resources Worldwide[2] (109 Barrels) Africa Canada Europe Venezuela USSR USA Resource 1.75 2000 0.4 1000 168 36.2 Reserves 0.175 333 0.06 100 16 2.9 Total 3206.35 452.135 4 Table 1.2 Oil Sands Resources of the United States[2] Deposits Measured (10s bbl) Major Deposits (>100 Billion Barrels) Alabama 1.8 Alaska California 1.9 1.7 Kentucky 0.1 New Mexico S. Oklahoma Texas 3.9 Tri state3 0.2 11.9 Utah 0.1 Wyoming Minor Deposits (<100 Million Barrels) Alabama California Utah a Kansas, Missouri and Oklahoma - - - - - Speculative (109 bbl) Total (109 bbl) 4.6 10.0 2.6 1.7 0.2 0.8 0.9 2.7 7.5 0.1 6.4 10.0 4.5 3.4 0.3 0.8 4.8 2.9 19.4 0.2 0.1 0.2 0.7 0.1 0.2 0.7 5 Table 1.3 Estimate of Oil Sands Resources of the State of Utah[13] Deposits (10s bbl) 438 3 480 1200 2200 40 60 175 95-120 330 550-1100 1200-1850 35-50 112-202 1048 100-125 1160 4250 5200-5850 125-140 227-317 430 55 3600 6 170 355-455 3400-7900 7-10 1307 455-545 9500-14000 23-26 Indicated (106 bbl) (106 bbl) Uinta Basin & Northeast Reqion: Asphalt Ridge 435 Asphalt Ridge, NW 2 Hill Creek 350 2500 PR Spring Sunnyside 1800 50 Whiterocks Other 20 Deposits 50 Central & Southeast Reqion: Circle Cliffs 707 San Rafael Swell 35 Tar Sand Triangle 2500 Other 10 Deposits 10 Total Inferred/ Conjectured (106 bbl) Measured 6 the hydrogen/carbon (H/C) atomic ratios are less. The nitrogen and sometimes metals (i.e., vanadium, nickel and arsenic) contents of oil sands bitumens are higher than those reported for conventional crude oil[2]. The high viscosities, asphaltene contents, Conradson carbon residues, heteroatom and metals contents of the oil sands bitumens make it difficult to integrate bitumens into conventional refinery operations. Hence, bitumens extracted from oil sands require upgrading before being accepted as refinery feedstocks. Utah's oil sands resources and bitumen characteristics have been reported in the Iiterature[14,15], A comparison of the physical and chemical properties of oil sands bitumens from the Uinta basin and the central-southeast region of the State of Utah and a conventional crude oil blend from the North Sea Brent field[16] is presented in Table 1.4. Nature of Bitumen The oil sands from the Athabasca deposits of Canada are water wet with the bitumen forming the continuous phase. The oil sands from the Uinta Basin are thought to be bitumen wet[2] with water forming a discontinuous phase in the sand matrix. Hence, the conventional hot water separation technique[17] used to extract Athabasca bitumen could not be directly applied for the extraction of Uinta Basin bitumens[8]. The oil sands of the Uinta Basin are of fresh water origin whereas the oil sands in the central-southeast region of Utah and Athabasca, Canada are of marine origin. The prominent difference between the Athabasca and Uinta Basin bitumens is that the former contain an order of Table 1.4 Physical and Chemical Properties of Uinta Basin Bitumens Properties Whiterocks Bitumen Specific Gravity (288/288 K) API Gravity, “API Conradson Carbon, wt % Pour Point, K Viscosity, cP @ 343 K 12.9 9.5 327 4825 Asphaltenesd), wt % Saturates, wt % Aromatics, wt % Resins, wt % Molecular Weight, g/gmol 2.9 35.7 7.0 54.5 653 Elemental Analvsis8) C, wt % H, wt % N, wt % S, wt % H/C Atomic Ratio 0.98 87.0 11.2 1.4 0.4 1.56 Asphalt Ridge Bitumen 0.985 12.1 13.9 320 5050 6.88 39.2 9.0 44.1 426 86.9 11.6 1.7 0.4 1.60 PR Spring Bitumen Sunnyside Bitumen 1.005 1.015 Tar Sand North Sea Triangle Brent Blend Bitumen Crude Oil[16] 1.01 9.3 14.2 319 47000 7.9 15.0 348 173000 8.6 16.7 306 42638a) 19.3 33.4 3.6 43.8 670 23.6 20.0 15.1 36.8 593 34.1 13.3 21.4 31.3 571 87.0 11.3 1.3 0.4 1.56 86.8 10.8 1.1 0.7 1.49 85.2 10.4 0.4 4.0 1.47 0.83: 38.3 2.13“ 231 1223c) 0.45 N.A. N.A. N.A. N.A. N.A. N.A. N.A. 0.4 N.A. Table 1.4 (continued) Properties Simulated Distillation Volatility, wt % <477 K 477-617 K 617-811 K >811 K a) b) c) <i) a) N.A. Whiterocks Bitumen Asphalt Ridge Bitumen 46.6 0.5 7.4 38.7 53.4 53.5 1.3 11.8 40.4 46.5 PR Spring Bitumen Sunnyside Bitumen 45.4 0.4 8.2 36.8 54.6 40.9 0.6 7.8 32.5 59.1 Tar Sand North Sea Triangle Brent Blend Bitumen C rude0il[16] 34.4 0.7 7.6 26.2 65.6 84.0 33.0 25.0 26.0 16.0 at 318 K Micro Carbon Residue at 303 K Pentane Insolubles C,H,N,S Analysis normalized to 100% Not reported 03 magnitude more sulfur than the latter and the nitrogen content of the former is lower than the latter. The nickel/vanadium ratios are greater than unity for the Uinta basin bitumens whereas the ratios are reversed for the marine origin bitumens. This has been attributed to the difference in their origins. Recovery Processes The bitumen recovery options include both in-situ and mining-surface recovery processes. Both the in-situ and mining-surface recovery processes are currently used in Canada for commercial bitumen production by Syncrude[18], Suncor[19], and ESSC)[20] as well as others[21]. A schematic identifying various in-situ and mining-surface recovery processes available for bitumen extraction is presented in Figure 1.1. In-situ processes for bitumen recovery include both thermal and nonthermal methods. The API gravities of the bitumens are generally less than 10°API and the viscosities are greater than 10,000 cP at reservoir conditions[2]. The high bitumen viscosity makes it immobile at the formation temperature and pressure. Hence, the in-situ processes concentrate on decreasing the viscosity of bitumen either by injection of thermal energy[2] or by dissolving the lighter hydrocarbons in carbon dioxide[22] or light hydrocarbon solvents[23]. The Cold Lake (Canada) oil sands deposit has been put on commercial bitumen production using a cyclic steam stimulation process[20]. The dense gases and light petroleum liquids along with steam injections have been evaluated for the recovery of light hydrocarbons from Canadian oil sands reservoirs[23], A 10 Figure 1.1 Schematic of Various Processes for Bitumen Extraction from Oil Sands Mintng & Surface Recovery Processes Aqueous Displacement Hot Water Processing Solvent Extraction Thermal Processing Subcritical Fluidized Bed Solvent Extraction Pyrolysis Modified Hot Supercritical Fluid Water Processing Extraction Rotary Kiln Pyrolysis Oil Sand Bitumen In-situ Processes Thermal Steam Non-Thermal Combustion Miscible Solvent Flooding Chemical Flooding 12 number of reviews of the thermal and nonthermal (solvent assisted) in-situ bitumen recovery processes have appeared in the literature[23-26], Mining-surface recovery processes are considered when the overburden-to-pay zone ratio is less than unity[27]. The mined oil sands are transported to processing locations where the bitumen is extracted from oil sands by various processes as indicated in Figure 1.1. The various surface processes include: • hot water process [17]; • solvent assisted aqueous method[28]; • solvent extraction[29-32]; • pyrolysis[10,11 ]; • supercritical fluid extraction[33,34]; and; • other thermal methods[35,36]. The hot water and solvent assisted aqueous processes raise environmental concerns due to the need for an aqueous tailings pond. A tailings pond is necessary since several years are required for the settling and separation of fines[37]. The solvent extraction process used for bitumen extraction from oil sands ores leaves behind residual solvent in the extracted sand and also poses potential environmental problems. Supercritical fluid extraction (SFE)[38] has been developed as an alternative to the conventional solvent extraction process. This process has several advantages over conventional processes: • solvent and solute can be separated by depressurization; 13 • high selectivity in separation of solubility class compounds; and; • selectivity can be varied by adjusting the pressure and temperature of the extraction process. An attempt has been in this study to produce upgraded liquid products from oil sands bitumens using a supercritical fluid extraction technique. Research Objectives A semicontinuous supercritical extraction system was built and experiments were conducted by Hwang[39] using the bitumen from the Whiterocks oil sands deposit of the Uinta Basin. Subsequently, modifications were made to the system to measure the density of the extract phase and the volume of the initial charge of solvent to the extractor. Experiments were carried out by Subramanian[40] using the bitumen from the PR Spring oil sands deposit with commercial propane as the solvent. The extraction process was modeled using the Peng-Robinson[41] cubic equation of state with a pseudocomponent lumping scheme[42]. The critical properties were estimated using Lee-Kesler correlations[43]. The present study focused on SFE of bitumens from the Asphalt Ridge and Sunnyside oil sands deposits using commercial propane as solvent. The extractions were carried out at the same operating conditions as those used for the Whiterocks[39] and PR Spring bitumens[40], so that meaningful comparisons could be made for the different bitumen feedstocks. The SFE system was also upgraded with the addition of a data acquisition system. 14 The bitumen feedstocks, extract samples, residual fractions and the solubility fractions obtained from adsorption chromatography[44] of the bitumens were analyzed using an extended American Standards for Testing Material (ASTM) simulated distillation technique developed during this study. This technique[45] permitted the determination of the boiling point distribution up to 911 K. The bitumens, two SFE extract phase samples and the residual SFE fractions were subjected to adsorption chromatography to separate each into solubility fractions: asphaltenes, resins, aromatics and saturates to investigate the influence of the nature of the feedstocks on the SFE process. Elemental analyses were conducted on the bitumens and on the residual fractions. The Peng-Robinson cubic equation of state[41] and a pseudocomponent lumping scheme[42] were used by Hwang[39] and Subramanian[40] to model the extraction process. The lumping scheme[42] introduced uncertainties vis-a-vis the representation of continuous mixtures like bitumen by a finite number of discrete lumps. In this study, continuous thermodynamics[46,47] was used to model the extraction process. Numerous equations of state (EOS) could have been chosen to represent the system, such as the Perturbed Hard Chain (PHC) EOS[48]; however, the Peng-Robinson EOS[36] was selected since only a limited number of parameters were required for its application to the description of the bitumens and bitumen-derived heavy oils. CHAPTER 2 LITERATURE REVIEW Supercritical Fluid Extraction Fundamentals Supercritical fluid extraction (SFE) utilizes the solvent power of supercritical fluids in the supercritical region. The supercritical fluid region for a pure component is defined[38] as the region on a P-T diagram at or above the critical pressure, Pc, corresponding to reduced pressures, Pr, > 1.0, and the critical temperature, Tc, corresponding to reduced temperatures, Tr, > 1.0)[49], The critical pressure and temperature are defined as the highest pressure and temperature at which a pure material exists in vapor-liquid equilibrium. Above the critical pressure and temperature the vapor and liquid phases disappear. The other distinctive characteristics of the critical state are the divergence in the compressibility and the phenomenon of critical opalescence[50] which leads to the visible scattering of light. In practice, the supercritical region has been taken to include Tr from 0.9 to 1,2 and Pr above 1 0[49], An alternate term “dense-gas extraction'’[51] emphasizes the use of high pressures and solvent densities where the solvent has enhanced extraction power. The term “destraction” derived from Latin words “destillare" and “extrahere” has been used in Germany[52] to infer that SFE is related to distillation and extraction which utilizes liquid solvents. However, SFE involves both an increase in the vapor pressure of the solvent and phase separation. The solvation power of a SCF is directly related to its density[38] which may be changed by varying the temperature and/or pressure of the system. The solvent density can be varied from liquid-like densities where the supercritical fluid is an effective solvent to vapor like densities where the supercritical fluid is a poor solvent. As an example for carbon dioxide[49] at Tr of 1.0, increasing the reduced pressure from 0.9 to 1.1, increases the density from gas-like values, 0.09 g/cm3, to liquid-like densities, 0.64 g/cm3. However, at high reduced temperatures, T, of 1.55, the reduced pressure must be increased from 0.9 to 5.7 to produce an equivalent increase in density. This sets the operational limit for the supercritical fluid extraction. At reduced pressure of 1.5, decreasing the reduced temperature from 1.2 to 1.0 increases the density of carbon dioxide from vapor-like values, 0.21 g/cm3, to liquid-like values, 0.77 g/cm3. At very high reduced pressures, the temperature has little effect on the density. The solvating power of the supercritical solvent is directly related to its density. Experimental Observations Ewald et al.[53] conducted solubility studies using a binary system consisting of ethylene and p-iodochlorobenzene to study the solubility of piodochlorobenzene in supercritical ethylene. The results enhancement in solubility near supercritical conditions. indicated an Extractions were conducted at 298 K (Tr=1.06) and at pressures varying from 0 to 8 MPa ( 0.0 < Pr 17 > 1.6) using ethylene as the solvent. The results indicated that at low pressure 0.5 MPa (Pr=0.1) and 298 K (Tr=1.06), the gas phase concentration of piodochlorobenzene was 0.007 g/liter[54]. At 6.0 MPa (Pr=1.2) and 298 K (Tr=1.06), the gas phase composition of p-iodochlorobenzene increased to 8.0 g/liter. The importance of conducting SFE near the critical temperature of the solvent was illustrated by Eisenbiss[55] with phenanthrene at 313 K and 40 MPa using four solvents. The vapor phase concentrations of phenanthrene for the various solvents are compared in Table 2.1. The solvents whose critical temperatures were well below the extraction temperature, i.e., nitrogen (Tc=126 K), methane (Tc=191 K) and carbon tetrafluride (Tc=222 K) did not extract phenanthrene. However, the solvents whose critical temperatures were closer to the extraction temperature, i.e., ethylene (Tc =283 K), ethane (Tc =305 K) and carbon dioxide (Tc =304 K) proved to be effective solvents. Theory of Supercritical Fluid Extraction A better understanding of SFE can be obtained from an examination of the thermodynamics of supercritical fluids. Consider a simple binary solid-gas system. The solubility of a solid in a gas is given[56] by [2.1] where y2 is the mole fraction of solute in the gas phase; E is the enhancement factor; i.e., the enhancement in the solubility over 18 Table 2.1 Comparison of the Phenanthrene Concentration in Vapor Phase for Various Solvents Solvent Critical Reduced Reduced Critical Temperature Pressure Temperature Pressure K MPa Gas Phase Concentration mg/liter Ethane 305.3 4.88 1.0 8.2 39 Ethylene 282.3 5.03 1.1 7.9 44 Carbon Dioxide 304 7.4 1.03 5.4 18 Nitrogen 126.2 12.62 2.5 3.2 <1.0 Methane 190.4 19.04 1.6 2.1 <1.0 Carbon Tetrafluride 227.6 22.76 1.4 1.8 <1.0 that of an ideal gas mixture; P2 is the vapor pressure of the solute, MPa; and; P is the system pressure, MPa. Enhancement factors of 104 are common for solid SCF systems and values over 1010 are known for squalane in supercritical carbon dioxide[56] and solid oxygen in supercritical hydrogen[58]. Rowlinson and Richardson[59] have derived an equation for E using the volume expansion virial equation of state: " RT V V ....... (2 2] where V is the molar volume, cm3/mol R is the gas constant; 8.314 cm3 MPa mol'1 1C1; T is the system temperature; K, and; B and C are the second and third virial coefficients, respectively. For a mixture of two gases Equation 2.2 can be modified as follows: PV — , =\ + j j V + J J V ' ^ .............. [2.3] where the virial coefficients J2 and J3, .. .are given by: J2 — Xj | + 2x,x2 Bn + x2 B22 Jy = x i3(’m + 3x1i x:C,12 +3X|X22Clr, + X,aCm [2-4] [2.5] 20 The virial coefficients Bn, B22, Cm and C222 in Equations 2.4 and 2.5 are the virial coefficients for pure species 1 and 2. B12 is the cross virial coefficient which accounts for the interaction between molecules 1 and 2. Considering the gas solution to be infinitely dilute, such that the mole fractions of X1 and x2 are made equal to 1 and 0; respectively, Equations 2.4 and 2.5 can be approximated as: J2=Bn [2.6] J3=C111 [2.7] Similarly, the chemical potential, n, for the component 2 can be written as: M: (g) / RT = M°2 / RT+\n(y2R D + 2J12 / V + 3J23 / I V 2 [2.8] where y2is the mole fraction of component 2; ^ 22 _ J 23 — ^ 1^12 ^ ^ 2 Ci)2 ^12 * ^12 > 2^]^2^122 * 2 ^222 ~ ^112 ' 2 n d , H°2 is a standard chemical potential of component 2. The chemical potential for pure solid component 2 is given by: M 2(S) = ^ 2 ( S ) + Pv 2 where //+2is the standard chemical potential; and; [2.9] 21 v, is the molar volume of solid, cm3/mol When pure solid is in equilibrium with its vapor (as for vapor pressure determinations in the absence of component 1): p 2 = M°2 (s) + RTln(C2°RT) [2 .10] where C 2° is the concentration of component 2 in the vapor phase. The terms Pv2 and virial coefficients are ignored in Equation 2.10 due to low pressure. At equilibrium between solid 2 and gas 1 at high pressure, the chemical potential of component 2 in both phases are equal: H2(gas) = |i2(solid) [2 . 11] Equating we obtain: in E = Pvs I R T - 2 j ^ I V - 3 J n l2V2 [2 . 12] where E is a measure of enhanced solubility, defined as: E = C2/C2° [2.13] By substituting for P/RT in Equation 2.12, using Equation 2.3, a relationship between E, V and Bi2 could be approximated as: ln £ = v, -_2.fi,12 V [2.14] 22 where B12 is the cross-virial coefficient between gas and solute (cm3/mol); and; v2 is the specific volume of the solute at T and P. The cross-virial coefficient is a measure of the extent of interaction between the solute and solvent molecules. The cross-virial coefficient is given B12= ~ ^ ( B ° + a > uB') [2.15] ' cl2 where B°= 0.083 - 0.422/(Tr1216); and; B1= 0.139 - 0.172/(Trl242). The greater the interaction between the solute and solvent, the greater the negative value of Bi 2. This is illustrated in Figure 2.1. The second order virial coefficients were calculated over a temperature range of 200 to 1000 K and are presented in Figure 2.1 for propane, n-butane, propane-decane and nbutane-decane systems. At a reduced temperature of 1.0, the propane-decane mixture has Bi2 of -500 cm3/mol compared to -830 cm3/mol for butane-decane binary mixture. Thus butane will be a better solvent than propane for the extraction of decane based on Bi2. Also from Equation 2.14, since vs and V are positive, the greater the enhancement factor E. Some general conclusions can be drawn from Equations 2.1 through 2.14 as follows: 23 Figure 2.1 Effect of Temperature on Enhancement Factor for Propane and n-Butane and its Binary Mixtures with n-Decane Virial Coefficient, cm3/mol 24 Reduced Temperature 25 1. Decreasing the extraction temperature (decreasing the reduced temperature) will decrease the V and increase the negativity of B12 (refer to Figure 2.1 for propane or propane-decane systems) and consequently increase E according to equation 2.3 for a given SFE solvent. Moreover, reducing the temperature of the extraction will reduce the solute vapor pressure P2. As a result of these two effects, the solubility will pass through a maximum at an optimum temperature T. 2. It has been graphically shown by Williams[60] and in Figure 2.1 that at a given extraction temperature, the solvents with higher critical temperatures will have larger negative values of B12 and consequently higher solvation capabilities than solvents with lower critical temperatures. For example, butane will have a higher solvating capability than propane since the former has a higher critical temperature than the latter. This is illustrated in Figure 2.2 where the enhancement factors are higher for the butane-decane system compared to the propane-decane system at 40.0 MPa in the temperature range from 600 to 1000 K. 3. V decreases with an increase in SFE pressure which leads to an increase in E and hence the solubility, since Bi2 is independent of pressure (as per Equation 2.14). For SFE systems with high extraction temperature and low solute volatility, the v,-2B12 and P2 terms would be smaller and hence high pressure would be required to obtain a reasonable enhancement factor. It should be noted that these conclusions can be used only as a qualitative guide. 26 Figure 2.2 Effect of Temperature on the Enhancement Factor for Propane-Decane and Butane-Decane Binary Systems Enhancement Factor 27 Temperature, K 28 The density controls the solvation power and the selectivity of the SCF. The solvent and depressurization. solute in SFE can be separated easily by simple The SCF possesses liquid-like densities but gas-like viscosities and diffusion coefficients that enhance its solvating capabilities relative to the liquid state[60], The fundamentals of SFE, properties of SCF and their applications have been reviewed in the literature[49, 54. 60-64], Applications of Suoercitical Fluid Extraction Supercritical fluid extraction has found numerous applications in the chemical and petroleum processing industries. The current areas of research and its applications are listed in Table 2.2[65-104], Supercritical carbon dioxide is used extensively in the food processing industry because it introduces no health hazards, is nonflammable, inexpensive and has a low critical temperature. Examples for such applications include decatenation of coffee[52,65,66], extraction of hops and spices[67,68], oil extraction from seeds and foods[69,70] and many other applications as listed in Table 2.2. Recently, carbon dioxide has received wide publicity for its ability to extract pollutants from water and soil[7175]. Supercritical fluids have been used extensively in chemical separation and purification applications such as desalination of sea water with isopropylbenzene, particulate separation, isomer and isotopes separation[76], extraction of monomers, oligomers and fractionation of polymers[77,78], regeneration of activated carbon[79] and many more[61-65]. Supercritical fluids have been used extensively in chromatographic applications[80] such as 29 Table 2.2 Applications of Supercritical Extraction Food Processing • • • • • • • • • • • • • Decaffeination of coffee and tea Deodorization of oils and fats Processing low vapor pressure oils Extractions Nicotine from tobacco Vegetable oil and fats from seeds Food coloring from plant material Hops and spices Fruit juices Flavor and fragrances from natural products Tall oil from wood Potato chip degreasing Acetone from antibiotics Chemical Separations and Purifications • • • • • • • • • • • • • • • • • Organic acids from water Extraction of pollutants from soil and water Alcohol-water separation Desalination of water using isopropylbenzene Activated carbon regeneration Polymer fractionation Separation of aromatic and paraffinic hydrocarbons Separation of nonpolar from polar compound Aromatic isomer separation Isotope separation Purification of organometallic compounds Redistribution of particle size by SFE nucleation Hydrothermal breeding of synthetic quartz crystal Supercritical fluid chromatography Drying and aerogel formation Cleaning, e.g., quartz rods for light guide fibers Removal of monomers, oligomers and solvent from polymers 30 Table 2.2 (Continued) Heavy Hydrocarbon Processing • • • • • • Deasphalting petroleum fractions Enhanced in situ oil and gas recovery Recovery and purification of oils and lubricants Coal liquefaction Shale, oil sand, and lignite extraction Extracting ozokerite from ore Reactive Separations • • Extraction of sec-butanol from iso-butane Hydrothermal oxidation of organic wastes in water 31 extraction, separation and analyses of pharmaceuticals[81], polymers[82], foods[83] and environmental pollutants[84], Supercritical fluids are also used in the hydrocarbon processing industry to deasphalt petroleum fractions[52], in coal processing[85], in the purification of oils and lubricants[49], in the upgrading of heavy oils, bitumens atmospheric and vacuum residues[33, 86], and enhanced oil recovery of crude oil and oil sands bitumens[87,23]. Phase Behavior Studies at Supercritical Conditions Supercritical fluid phase behavior calculations have been reported in the literature for binary[88-92], ternary[93-97] multicomponent systems[98-102] involving hydrocarbon-carbon dioxide and hydrocarbon-propane[52,103,104] systems at high pressures and moderate temperatures. Supercritical Fluid Extraction of Oil Sands Supercritical solvents have been tested for their capability to extract bitumen derived light oils from oil sands[105-109], Martin and William[105] patented a SFE process to extract bitumen from Athabasca oil sands using pentane and toluene as solvents in the temperature range from 503 to 673 K and a pressure of 10.3 MPa. They extracted 75 and 89 wt% of the total organic matter from the Athabasca oil sands using pentane (Pr=3.06, Tr=1.07) and toluene (Pr=2.51, T r=1.14) respectively, as solvents. Pang and McLaughlin[106] conducted experiments on SFE of Athabasca oil sands using carbon dioxide, methanol, acetone, isopropyl alcohol, n-heptane and cyclohexane as solvents. The reduced temperatures ranged from 1.0 to 1.9 32 and the reduced pressures ranged from 1.2 to 3.6 depending upon the solvent employed. Supercritical fluid extractions were performed in a 1-L autocalve for 30 min duration with 0.12 mole of solvent per gram of Athabasca oil sands. The results indicated that the extraction yields varied from 2.3 to 11.5 wt% of the oil sands as against a maximum of 12.1 wt% (total organics plus water) obtained from soxhlet extraction using benzene as the solvent. The extraction yields increased with increase in the density of the solvent and system pressure at constant extraction time and temperature. Panzner et al.[107] performed extractions of the solute adsorbed on solid particles under supercritical conditions. Pentane in the subcritical and the supercritical state was used to extract Athabasca oil sands to recover pentane deasphalted maltenes. The extraction experiments were carried out at high temperatures and pressures (T = 413 to 513 K; Tr = 0.88 to 1.09 and P = 1.9 to 8.2 MPa; Pr = 0.56 to 2.43). They were able to extract 10.6 wt% of the oil sands as deasphalted oil with low vanadium, nickel and iron content against 14.8 wt% of total bitumen content. Compton[108] carried out studies using Utah oil sands with supercritical solvent mixtures. The mixtures of two or more solvents were intended to reduce the energy requirement of the process. The experiment was performed at 422 K and 17.3 MPa with toluene and methane as solvents in the ratio of 3:1 by weight. After four hours of extraction, 95 wt% of the organic matter present in the oil sands was extracted. 33 Jocoby[109] conducted studies on the extraction of Athabasca oil sands using solvents such as ethane, propane, butane and pentane. The extractions were performed using these solvents in the supercritical and liquid states. He has also conducted experiments on Athabasca oil sands using propane as solvent at 383 K (Tr=1.03) and 20 MPa (Pr=4.71). Approximately, 9.5 wt% of the bitumen present in the sand matrix was extracted as a viscous, red liquid. Hwang[39] studied the SFE of hexadecane, a paraffinic crude oil, a bitumen derived liquid produced during pyrolysis of the Whiterocks (WR) bitumen and the WR bitumen with solvents such as carbon dioxide and propane. The extractions were performed using the original version of SFE extraction apparatus described in Chapter 3. SFE experiments were conducted on the bitumen-derived liquid and the WR bitumen using propane as the solvent. The extraction experiments were conducted at pressures of 5.6 (Pr=1.2), 10.4 (Pr=2.3) and 17.3 MPa (Pr=3.8) and temperatures of 339 (Tr=0.92), 380 (Tr=0.1.03) and 422 K (Tr=1.14): a combination of five extraction conditions to study the effect of pressure, temperature and solvent density on extraction yields. Hwang[39] concluded that that the extraction yields were controlled by complex interplay between extraction efficiency (linked to solvent density) and feedstock volatility. The feedstocks and extract phases were subjected simulated distillation to obtain boiling point distributions up to 818 K. The simulated distillation results indicated that the extract phases were significantly upgraded liquids compared to the WR bitumen. potential of SFE for upgrading bitumens. This observation also indicated the Hwang[39] modeled the extraction 34 process using pseudocomponent lumping scheme[42] and the Peng-Robinson equation of state[41], modeling SFE system. Hwang[39] assumed the initial overall composition for The extract phase densities for the modeling were estimated using the Peng-Robinson equation of state in the absence of experimental data. Despite these shortcomings, the predicted extract phase compositions matched well with the experimental compositions. Subramanian[40] modified the extraction system used by Hwang by the addition of a positive displacement pump and a densitometer to measure the initial overall compositions and extract phase densities, respectively. The extraction experiments were conducted at the same five operating conditions [39] so that a meaningful comparison of the extraction yields could be made. The SFE experiments were performed using the PR Spring (PRS) bitumen as feedstock and propane as the solvent. The extraction results indicated an increase in cumulative extraction yields with an increase in pressure at constant temperature. The cumulative extraction yields decreased as the extraction temperature increased at constant pressure. The feedstock, extract phases and residual fractions were subjected to simulated distillation to obtain boiling point distributions up to 973 K. The simulated distillation results indicated that the extract phases were significantly upgraded liquids compared to PRS bitumen. A nonvolatile residual fraction remained in the extraction vessel. This confirmed the observation reported by Hwang[39] for SFE of the WR bitumen using propane as the solvent. The residual fractions obtained from SFE of WR and PRS bitumens were subjected to solvent fractionation[44] to obtain asphaltenes, 35 saturates and aromatics and resin fractions. The results indicated that the asphaltene content of the residual fractions were higher (higher than the asphaltene content estimated on prorated basis) than the bitumens indicating removal of the cosolubilizing agents during SFE. The decrease in the saturates and aromatics contents of the residual fractions with propane density indicated preferential removal of the saturated compounds compared to resins and asphaltenes. This was confirmed by the reduction in the H/C ratio of the residual fractions compared to the bitumens. The SFE of the PRS bitumen was modeled by Subramanian[40] using pseudocomponent lumping scheme[42] and the Peng-Robinson equation of state[41]. The initial overall composition was estimated using the bitumen and solvent (measured using the positive displacement pump) charged to the system. Significant volume change during mixing was observed during initial charging of the solvent. The measured extract phase density was used in the material balance calculations instead of an estimated density. match was observed between predicted and measured A reasonable extract phase compositions. Commercial Supercritical Fluid Extraction Processes ROSE Process The Residuum Oil Supercritical Extraction (ROSE) process was developed by Kerr-McGee Corporation and commercialized in 1979 to upgrade vacuum resids[110]. The results obtained from ROSE pilot plant studies using 36 Lagunillas vacuum resid as feed and propane, butane and pentane as SFE solvents are presented in Table 2.3. The results presented in Table 2.3 indicate that as the carbon number of the solvent increased, the extraction yields increased; however, there was a concomitant increase in the specific gravity, viscosity, Conradson carbon residue, and nickel and vanadium contents of the deasphalted oil. Nonetheless, the values reported for all three solvents were lower than the values reported for the feed. Feedstocks, such as coal-derived liquids (Table 2.4) and oil sands bitumens (Table 2.5), have also been processed[110,111]. The product streams produced from the ROSE process when the Athabasca bitumen resids obtained from distillation and visbreaking operations were used as feedstocks and butane and pentane were used as solvents were the asphaltenes, the resins and a deasphalted oil. The product yields and the qualities of products produced from the Athabasca bitumen 613 K plus resid are presented in Table 2.6. The results from Table 2.6 indicate that as the carbon number of the solvent increased from butane to pentane, the deasphalted oil yields increased; however, the quality of the product, as reflected by the API gravity, Conradson carbon residue, nitrogen, sulfur, viscosity and metal contents, declined. Kerr-McGee also conducted pilot studies recently to upgrade heavy oils and resids from the Zauta and Boscan crudes from Venuzeuela using propane, butane and pentane as solvents[33]. The results obtained from pilot plant studies with the Zuata crude using propane as solvent with solvent-to-feed ratio of 8:1 are presented in Table 2.7. The extraction results indicate that a significantly upgraded deasphalted heavy oil 37 Table 2.3 Quality of Yield Obtained from ROSE Unit for Lagunillas Vacuum Resid[112] Feed Solvent Yield, wt% Propane 100 API Gravity, °API 4.8 Conradson Carbon Residue wt% 23.5 Viscosities, cst @ 372 K 11,600 @ 408 K 900 Nitrogen, wt% Sulfur, wt% Nickel, ppm Vanadium, ppm a) Deasphalted Oil 0.63 4.3 95 650 ............ DAOa)................. i-Butane n-Butane n-Pentane 13.3 30.8 58.2 70.3 17.0 14.2 11.3 10 .1 3.3 5.9 10.4 13.7 82 25 0 .2 0 2.7 1 4 98 28 0.31 3.2 7 34 312 65 0.36 3.8 20 130 424 82 0.45 4.0 31 230 38 Table 2.4 Quality of Yield Obtained from ROSE-SR Unit for Coal Liquefaction Resid from Bituminous Coal [112] Feed Yield, wt% Ash Heavy Deashed Concentrate Resid Light Deashed Resid 18.0 32.8 49.2 Yield, wt% on ash and unconverted coal free Basis 100 6.0 37.6 56.4 Sulfur, wt% Ash, wt% Ring & Ball Softening Point, K 10 0 1.4 7.5 373 5.0 41.7 - 0.4 0.08 457 0.7 0 .0 1 339 39 Table 2.5 Quality of Yield Obtained from ROSE Unit for Resid from Retorting Oil Bearing Diatomacious Earth Using Iso-Butane as Solvent[112] Feedstock Yield, wt% API Gravity, °API Conradson Carbon Residue wt% Viscosities, cst @ 372 K @ 408 K Nitrogen, wt% Sulfur, wt% Nickel, ppm Vanadium, ppm a) Deasphalted Oil Asphaltene DAOa) 43.4 56.6 -1 . 2 -18.7 12.3 19.8 42 3.1 10 0 49 14 0.79 1.0 20 20 2400 320 1.08 1.4 42 43 11 5 0.57 0.7 3 4 40 Table 2.6 Quality of Product and Yield Obtained from ROSE Unit for n-Butane and n-Pentane Extractions of Athabasca Oil Sands Bitumen (613 K plus Resid)[112] Feed Solvent Yield, wt% 10 0 5.7 API Gravity, °API Conradson Carbon Residue, wt% 15.6 Viscosities, cst @ 372 K 2000 @408 K 265 Nitrogen, wt% Sulfur, wt% Nickel, ppm Vanadium, ppm a) Deasphalted Oil 0.52 5.4 115 287 Asphaltene DAOa) — n-Butane------ Asphaltene DAOa) -------n-Pentane------ 35 65 -8 . 0 13.2 -11.9 4.5 48 36 - - 0.97 8 .6 290 700 144 37 0.28 3.7 20 67 20 - - 1.12 9.4 400 930 80 10 .1 7.6 254 55 0.37 4.4 43 125 41 Table 2.7 Quality of Product and Yield Obtained from ROSE Unit Using Zuata Crude as Feed and Propane as Solvent[33] Crude Yield, wt% 100 API Gravity, °API 9.2 Conradson Carbon Residue, wt% 13.4 Sulfur, wt% Nickel, ppm Vanadium, ppm 3.4 81 384 n Ao a) Crude _ _ _ _ _ _ _ _ _ _ _ L/i r s c u Atmospheric Resid 10 0 57 a) rv___________i_ _ _ _ __ _ _ _ _ _ _ _i /-kii i UuUOi ——— — 45 9.1 17.8 15.4 1.4 0 .8 2 .6 2.4 0.4 3.8 93 448 - — Atmospheric Resid 7.8 2 ' a) Deasphalted Oil i 42 with high API gravity and low metal content was produced from the atmospheric resid. HSC-ROSE Process The HSC-ROSE process[113] combines the High Conversion Soaker Cracker (HSC) process from Toyo Engineering Corp. and Mitsui Coke Co., Ltd. The ROSE process developed by Kerr-McGee Corporation to upgrade variety of feedstocks like vacuum residues, bitumen, shale oil and coal derived liquids. Pilot plant studies[113] were carried out with the Cold Lake bitumen and its vacuum residue. Selected physical and chemical properties of the deasphalted oil obtained from the HSC-ROSE process are presented in Table 2.8. The results from the above studies proved the capability of supercritical fluids to extract bitumen from oil sands and to produce upgraded bitumen-derived liquids from extracted bitumen. Demex Process The Demex process[114] was developed by UOP Inc., Des Plaines, Illinois. This process utilizes a novel supercritical fluid separation technique to upgrade heavy oils. The Demex process is an extension of the commercial propane solvent deasphalting process which separates high metal vacuum residue into a demetalized oil (DMO) relatively low in metal content and an asphaltene fraction with high a metal content. The DMO is a desirable feed stock for hydrodesulfurization, fluid catalytic cracker and hydrocracker units. There are many commercial units currently on stream. 43 Table 2.8 Quality of Syncrude Produced by the HSC-ROSE Process[113] Low Conversion HSC Distillates plus ROSE DAOa) Yield on Bitumen, wt% 82.0 High Conversion HSC Distillates plus ROSE DAOa) 82.0 Specific Gravity, (288 K/288 K) API Gravity, °API 0.9330 2 0 .1 0.9245 2 1.6 Conradson Carbon Residue, wt% 3.3 4.4 Viscosity, cst @294 K 130 65 @311 K 48 29 Nitrogen, wt% 0.24 0.26 Sulfur, wt% 3.6 3.7 Nickel, ppm 7 12 13 22 Vanadium, ppm a) Deasphalted oil 44 A schematic of the Demex process is presented in Figure 2.3. The vacuum residuum charge is mixed with a solvent such as propane along with recycled solvent from the second stage and is fed into the first stage extractor. The pressure and temperature in the first stage extractor permit it to operate as a liquid-liquid extractor. The asphaltenes are rejected in the first stage. The first stage overhead is heated in a heat exchanger with hot solvent. The temperature increase decreases the solubility of the resins and high molecular weight aromatics which are separated in the second stage extractor. The bottom stream from the second stage is recycled to the first stage extractor and a portion of the recycled resins is withdrawn as product. The overhead from the second stage is heated in a heat exchanger with hot solvent and a heater so that the temperature of the solvent-DMO mixture is raised above the critical temperature of the solvent. At this high temperature, DMO is insoluble in the solvent and is separated as product DMO in the DMO supercritical separator. The solvent vapors are condensed, dewatered and pumped back to the first stage extractor. The properties of the DMO and the asphaltene fraction obtained from the DEMEX process using Arabian light vacuum resid at two extraction levels are presented in Table 2.9. It is observed from Table 2.9 that the quality of the DMO decreases with increase in DMO yield with propane as solvent. The quality of the DMO obtained from the Demex process is significantly upgraded compared to the vacuum resid feed leaving behind a residue high in metals, Conradson carbon residue, nitrogen and sulfur. 45 Figure 2.3 Schematic of the Demex Process[114] /T Heater DMO Supercritical Separator ~i HI III—i Fin Cooler Asphalt Separator and Stripper DMO Stripper r c ) ISteam DMO Solvent Drum Condensate Makeup Solvent T 1 sSteam t Asphalt cn 47 Table 2.9 Products Yield and Quality from the Demex Process[114] Properties Asphaltene DMO Vacuum Conditionl Condition 2 Condition 1 Condition 2 Resid Yield, vol% 10 0 56.0 78.0 Specific Gravity (288 K 1277 K) 1.0 2 0.959 0.986 Nitrogen, wt% 0.31 0.14 0 .2 1 Sulfur, wt% 4.0 2.74 3.25 Conradson Carbon Residue, wt% 2 0 .8 Metals, Vanadium + Nickel, ppm 98 Softening Point, K - 5.6 10.7 6 19 - - 44.0 1.10 - 5.4 - 2 2 .0 1.16 - 6.3 - 201 341 389 450 48 SOLVAHL Process The SOLVAHL process[115] is a deasphalting process developed by Asvahl, an association of Elf, Total and IFP and licensed by IFP, using liquidliquid extraction and supercritical fluid separation technique to upgrade heavy oils to produce deasphalted oil suitable for catalytic cracking, hydrocracking and other refinery downstream processes. The SOLVAHL process could be used on vacuum residues using C3, C4 and C5 solvent to produce DAO with reduced metals, Conradson carbon, sulfur and nitrogen levels (Table 2.10). This process has not been licensed by commercial oil refining companies to process heavy oil. However, Asvahl validated the SOLVAHL process in a 32,000 tons/year developmental plant located in Feyzin, France. The process flow diagram is presented in Figure 2.4. Typical yield and quality for the Arabian Light Vacuum Resid using the SOLVAHL Process[115] are presented in Table 2.10. The three solvent deasphalting processes described above are similar in nature, but different in many important design details. These three processes are an extension of the propane deasphalting process used for a long time for preparation of deasphalted oils for lube oil manufacture or other refinery downstream processes. The unique feature of these process are recovery of solvent from DAO under supercritical conditions. This reduces the energy requirement by 30-50 % compared to conventional deasphalting process, where the solvent is recondensation. recovered from deasphalted oil by evaporation and 49 Figure 2.4 Schematic of the SOLVAHL Process[115] Heat Heater Exchanger Heater Separator DAO Strippers 51 Table 2.10 Yield and Quality for the Arabian Light Vacuum Resid using the SOLVAHL Process[115] Properties Feed Yield, vol% _ Specific Gravity 1.003 0.933 Nitrogen, wt% 0.29 0.192 Sulfur, wt% 4.05 2.55 DAO from C3 Solvent 45.5 Conradson Carbon Residue, wt% 16.4 1.65 Vanadium, ppm 61 1.0 Nickel, ppm 19 1.4 345 34.9 Viscosity, cSt @373 K 52 Modeling the Bitumen Extraction Process Experimental data obtained from laboratory scale SFE systems can be used to fit to suitable thermodynamic models. The chemical process industry requires accurate thermodynamic property data for mixtures and phase equilibrium data to design supercritical processes to upgrade bitumen feedstocks. It is difficult to obtain experimental data at all conditions of interest. Thermodynamic models are useful in process design since it is possible to proceed with a minimum of experimental data. For processes involving complex mixtures like bitumen few SFE models using propane as solvent are available. Complex mixtures like crude oils, bitumens, shale oils and coal-derived liquids contain many components and it is not possible to specify the composition or to identify the complete slate of components in these mixtures. The development of thermodynamic models of the extraction processes involving these complex multicomponent mixtures requires a method for describing the mixtures. Two methods are commonly used: pseudocomponent lumping[42] and continuous thermodynamics[46,47]. Pseudocomponent lumping is used to represent the complex mixture by a finite number of discrete lumps or “components’ to which pseudocomponent properties are assigned. The grouping of the pseudocomponents is rather arbitrary and the phase equilibrium calculation results are sensitive to the number of components and the grouping technique used. Continuous thermodynamics represents a complex mixture by means of a continuous distribution function of appropriate characterization 53 properties such as molecular weight, boiling point or specific gravity. Continuous thermodynamics uses a statistical procedure to describe complex hydrocarbon mixtures and the grouping is performed using mathematical principles. Pseudocomponent Lumping Scheme Complex mixtures like bitumen contain a variety of components ranging in carbon numbers from C6 to C90 and higher as indicted by simulated distillation[45] data. Moreover, each carbon number can be subdivided into paraffinic, naphthenic, aromatic, resinous and asphaltenic compounds. The number of theoretical isomers for paraffins from C1 through C2o has been calculated by Alberty and Gehrig[116] and are shown in Table 2.11. There are 136 different isomers for a paraffin with carbon number 10. When the carbon number is doubled to 20, the number of possible isomers increase to 3,396,844. If this analysis is extended to carbon number 90 and naphthenic and aromatic compounds were included, a complex mixture like crude oil or bitumen would consist of millions of components[116]. Since it is not practical to use a very large number of components in phase equilibrium calculations, a pseudocomponent lumping scheme was developed to represent mixtures such as bitumens by a manageable number of pseudocomponents. Whitson[42] proposed a lumping scheme to compute the number of pseudocomponents required to represent the complex mixture as: 54 Table 2.11 Theoretical Number of Isomers for Paraffins Carbon Number Number of Isomers 1 1 2 1 3 4 5 2 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 1 3 5 11 24 55 136 345 900 2412 6563 18217 50699 143255 408429 1173770 3396844 55 N a = Ini I + 3 .3 lo g 10(iV - «)] [2.16] where: NGis the number of pseudocomponents; N is the total number of components in the mixture; and; n is the first component in a C7+ fraction, 7. The molecular weight separating each pseudocomponents is given by: [2-17] where: l is 1 , ...., NG; Mi is the molecular weight of pseudocomponent I, g/gmol; Mn is the molecular weight of the last component, g/gmol; and; Mn is the molecular weight of heptane. Several authors have adopted different guidelines for grouping complex hydrocarbon mixtures. Lee et al.[117] plotted all properties related to the crude oil components as the dependent variables against boiling point as the independent variable. The fractions with similar slopes were grouped together as a pseudocomponent. The phase equilibrium predictions obtained using the Peng-Robinson (PR)[41] equation of state for a vapor/crude oil/water system at 17.3 MPa and 588 K were in close agreement with the experimental data. This grouping technique required elaborate phase equilibrium calculations and graphical procedures to obtain the optimum number of pseudocomponents and grouping patterns. Pederson et al.[118] grouped the components together in 56 equal weight fractions as determined from gas chromatographic and true boiling point distributions. This grouping technique was simple and yet able to provide reasonably good phase equilibrium predictions using the Soave-Redlich-Kwong (SRK)[119) equation of state for North Sea oils and gas condensates without adjusting the interaction parameters. Li et al.[120] grouped pseudocomponents for every decade of Henry’s constant, k. This lumping method did not require cumbersome calculations but does not have a strong theoretical basis. Mehra et al. [ 1 2 1 ] adopted a complex statistical method for grouping fractions that minimizes the error in the phase-saturation calculations. This method requires elaborate phase equilibrium calculations for each consecutive fraction to be grouped together. Colonomous et al.[122] used a linear programming technique to examine all possible combinations and minimize the difference between the properties of the lumped distribution and the original distribution. This method provided accurate phase equilibrium data but involved extensive calculations to define the number of pseudocomponents and grouping patterns. Montel and Gouel[123] used similarities in specific properties of the feedstock to lump 150 identifiable monomers in the Ci to Ci0 range. The Cn+ fraction was not treated. This method requires complete analysis of the hydrocarbon mixture under investigation; however, it is not possible to obtain a full range of data for all the components. The molecular weights, specific gravities and boiling points for the pseudocomponents were obtained either by analytical procedures or were computed from boiling point distributions estimated from true boiling point 57 distributions or simulated distillation techniques[45]. These pseudocomponent properties were used to estimate the critical temperatures, critical pressures and acentric factors using correlations Daubert[124] or Twu[125], proposed by Lee-Kesler[43], Riazi- However, the pseudocomponent selection was arbitrary and the calculated results were sensitive to the number of pseudocomponents and to the grouping of components. Continuous Thermodynamics According to Gibbs[126], physical and/or chemical equilibrium is established when the chemical potentials of all species in all phases are identical: 1 2 H s - fd = fl 3 = fl 4 ................ = f l m [2 18] where \x,mis chemical potential of species j in phase m. Continuous thermodynamics is governed by an extension of Gibbsian thermodynamics to mixtures whose compositions are represented by statistical functions of one or more of the macroscopic properties of the mixtures such as molecular weight, boiling point, specific gravity, degree of aromaticity, number of carbon numbers per molecule, or Lewis basicity. Bowman[127], Edmister[128], Aris and Gavalas[129], and Hoffman et al.[130] have suggested phase equilibria calculation procedures for complex mixtures represented by continuous distributions. However, these procedures are restricted to physicochemical models using Raoult’s law solutions, relative 58 volatilities and ideal gas mixtures. Their approach was not concerned with the general problem associated; with continuous thermodynamics: solving continuous distributions and material balance equations simultaneously which is independent of physicochemical models. Fundamental studies on continuous thermodynamics Vrij[131 ], were reported by Blum and Stell[132], Dickinston[133], Smith and Rowlinson[134], Salacuse and Stell[135], Briano and Glandt[136] and Gualtieri et al.[137] in the late 1970s and early 1980s. Vrij[131 ] discussed the interparticle multicomponent system of hard spheres. interaction applied to a He also developed expressions for light scattering from a multicomponent system of colloidal particles dispersed in a low molecular weight solvent. The reliability of the Percus-Yevick[138] approximation for hard spheres was also discussed. Blum and Stell[132] extended the Percus-Yevick[138] approximation for a polydisperse fluid of hard spheres to a simple permeable-sphere model in order to diversify its applicability to colloid and polymer systems. Dickinson[133] described the fluid phase equilibrium for a polydisperse system by an extended theory of conformal solutions. He has also proved that the distribution of particles between the coexisting phases is more sensitive to interparticle forces than van der wall type forces. Smith and Rowlinson[134] studied a system of molecules containing a large number of species (polydisperse) and concluded that such a system is unstable at high densities. Salacuse and Stell[135] investigated statistical thermodynamics of polydisperse systems. Equilibrium conditions and a Gibbs- 59 Duhem[126] relation for a two phase system were also derived by them. Briano and Glandt[136] proved that statistical concepts related to random variables could be applied to model continuous (polydisperse) mixtures. Gualtieri et al.[137] developed a mathematical framework to provide a method for generalizing the thermodynamics of a finite system to that of a polydisperse system. They have also developed two new functions, the mole fraction distribution and mole fraction density function to describe polydisperse systems of the van der Walls model. They also solved the equations for three phase equilibria problems. These studies focused on mathematical relationships to represent complex mixtures and little or no attention was given to the application of continuous thermodynamics to chemical industry systems. Major contributions were made by Ratzsch and Kehlen[139-145] and Cotterman and Prauznitz[146-148] involving application of continuous thermodynamics to the chemical process industries. Ratzsch and Kehlen[139145] developed a consistent approach for the application of continuous thermodynamics to all mixtures containing a large number of similar chemical species. Each study[139-145] stressed the significance and advantages of using continuous thermodynamics over pseudocomponents in thermodynamic treatments. They also generalized a method to separately treat two or more ensembles (e.g., paraffins and aromatics) of similar chemical species in a complex hydrocarbon mixture along with some individual (discrete) species. Cotterman and Prausnitz[146-148] also illustrated the difficulty involved in solving the material balance and phase equilibrium equations in flash 60 calculations and suggested two numerical procedures for solving them. Additional examples were presented on application of continuous thermodynamic methods in enhanced oil recovery and upgrading of heavy crude residuum using supercritical solvents[47]. The discrete and continuous distribution multicomponent mixture are presented in Figure 2.5. representations for a The discrete mixture is shown on the left side as Figure 2.5a and b. The independent variable T such as molecular weight used to characterize the mixture is plotted against the mole fractions, Xj, in Figure 2.5a. The cumulative mole fraction for the total number of components ‘I’ is presented in Figure 2.5b. A continuous distribution used to represent the mixture is shown on the right side as Figure 2.5c and d. The continuous mixture does not have discrete identifiable components but instead the mixture is characterized using a single distribution variable, I, which has distribution function F(l) such that the fraction of molecules characterized by the range I to l+AI is F(I)AI where Al is a small increment in I. The normalization for the lumped system is given by summation over the entire range of I, whereas integrals over I are used for the continuous distribution. At equilibrium, the chemical potential of each component (j) is equal in each of the m equilibrated phases for the lumped or discrete case. The equilibrium condition is given by: f t ' r M r f t ' , * n ) ................ • m " l2 1 9 l 61 Figure 2.5 Representation of Discrete and Continuous Distributions for Complex Hydrocarbon Mixtures 62 Finite Component Mixture Component Number j Figure 2.5a Continuous Mixture Distribution Variable I Figure 2.5c Component Number j Figure 2.5b Distribution Variable I Figure 2.5d 63 where is the chemical potential of species j and the superscript represents the phase index. For the continuous distribution function, F(l), with one distribution variable, I, the chemical potential is given by: [2.20] where (ij(l) is the chemical potential for distributed variable I for phase index m. The chemical potential, [Xim , for lumped or discrete mixtures can be calculated from an equation of state relating pressure P, total volume V and number of moles ni, n j,.......... as[47]: r ■ a : 4 -J RT d V -R T In PqV ’ V n mRT SP ^ m cn, [2 .21] T,U.nj" where P0 is the reference state pressure, potential at 1 1 bar, and |i (0 is the ideal-gas chemical bar. A similar equation[45] can be derived for continuous mixtures. m f | 1 y" 1 P V ' i ' RT dV-RTln----- £— — + u\T ,I) [2.22] — ^ ■11\ F ( n ' ~ V n"Fm(I)RT * V 7 1 J ' T.VJ'=1 where Fm(l) is the distribution function in phase m; nc is the total number of moles in the continuous mixture; d denotes a variational or functional derivative; 64 F(l+) is the molar distribution function for characterizing variable l+=l; and n°(T,l) is the ideal gas chemical potential at 1 bar. Equation 2.22 provides the key relationship for introducing molecular thermodynamics into the framework of continuous thermodynamics. The above discussion was based on the assumption that a single distribution variable is sufficient for describing the mixture of interest. However, two or more distribution variables can be used to describe a continuous mixture which can not be represented by single distribution adequately. Cotterman[47] and William and Teja[149] derived chemical potential equations for two distribution variables I and J (bivariate) and used a distribution function F(I,J) instead of F(l). Any number of distribution variables can be used, but, the solution of equations involving more than one distribution variable becomes complex. The choice of appropriate variable (l,J,..) and the appropriate function F, is important to describe the continuous mixture represented by F(l,J,..). The choice of the statistical function, F, depends on the ability of the function to represent the mixture of interest with sufficient accuracy. The function choice is also dependent on the ease with which it can be solved to obtain various thermodynamic properties such as the dew point, the bubble point, phase compositions, etc. Behrans and Sandler [150] successfully used the log of the mole fraction vs carbon number to represent crude oil and condensate from various oil and gas fields for estimating dew point, bubble point and saturation pressures. 65 Kehlen and Ratzsch[139] and Hoffman[130] used a normal or Gaussian distribution function successfully to represent polymers and reservoir fluids. The Schulz[151] distribution function was used by Gualtieri et al.[137] in a mathematical description of continuous thermodynamics for dilute polydisperse systems. Cotterman[47] used gamma[152] and beta[153] distribution functions to represent petroleum reservoir fluids and polymers. The gamma distribution function was used successfully in dew point calculations for a natural gas, solvent loss in a high pressure absorber, flash calculations for fractionation of a polydisperse polymer solution, a C 02-paraffins system, and oil and resin separation in a supercritical fluid extraction system using C 0 2 and propane as solvents. William and Teja[149] used a bivariate log normal (Gaussian) distribution function proposed by Hahn and Shapiro[154] and single variable distribution functions to estimate dew points for natural gas condensate, a naphthenic absorber oil, a synthetic aromatic oil and crude oils. They used the following options for phase equilibrium calculations: 1. Semicontinuous thermodynamics using the Patel-Teja[155] equation of state and a bivariate distribution function with true boiling point and specific gravity as the characterizing variables (CPT2) to describe natural gas condensate, a naphthenic absorber oil, a synthetic aromatic oil and crude oils. 2. Semicontinuous thermodynamics using the Patel-Teja[155] equation of state and a single distribution function with true boiling point as the characterizing 66 variable (CPU ) to describe natural gas condensate, a naphthenic absorber oil, a synthetic aromatic oil and crude oils. 3. Semicontinuous thermodynamics using the (SRK)[119] equation of state and a single distribution function with molecular weight as the characterizing variable (CSRK) to describe natural gas condensate, a naphthenic absorber oil, and crude oils. 4. Pseudocomponent approach using the Patel-Teja[155] equation of state (DPT) to describe natural gas condensate, a naphthenic absorber oil, a synthetic aromatic oil and crude oils. 5. Pseudocomponent approach using the (SRK)[119] equation of state (DSRK) to describe natural gas condensate, a naphthenic absorber oil, a synthetic aromatic oil and crude oils. It was concluded by William and Teja[149] that the two semicontinuous methods involving the Patel and Teja[155] equation of state to predict dew point temperatures and liquid densities for a variety of hydrocarbon mixtures were superior to the two discrete methods and semicontinuous method using the (SRK)[119] equation of state. An attempt was made by Peng et al.[156] to extend the applicability of the PR equation of state to oil reservoir phase behavior predictions using continuous thermodynamics. Gamma and beta distribution functions were used to represent the reservoir fluids under investigation. The results demonstrated that the gamma distribution may not be a suitable model in representing reservoir fluids because it is open ended on the uncharacterized portion and the dew point 67 calculations are very uncharacterized portion. sensitive to approximations made to represent They suggested using a truncated form of the beta distribution function for dew point calculations. Haynes and Matthews[157-160] used the true boiling point distribution to describe a hydrocarbon mixture consisting of 20 paraffin components. The phase envelopes (P-T diagram) constructed using a 20 discrete components scheme and 8 quadrature points were in good agreement with each other. The (SRK)[119] equation of state was used for the phase envelope computation. Angelos et al.[158] obtained P-T-x-y data for two synthetic mixtures containing various solubility fractions and compared the experimental vapor and liquid phase compositions with predicted values. The phase equilibrium calculations were performed using a single variable continuous distribution function based on the true boiling point curve. The fit obtained between experimental and predicted values was reasonable and validated the use of true boiling point curves in continuous thermodynamics applications. Mani et al.[159] also used true boiling point curves to represent a naphtha/kerosene blendstock in bubble point and equilibrium flash vaporization calculations using the PR[41] equation of state, continuous thermodynamic principles and the quadrature technique. They also converted true boiling point curves into mole fractions using weighting factors and estimated critical parameters at the quadrature points. The predicted bubble point pressures were in good agreement with the measured values. Ying et al.[161 ] used a cubic spline curve to fit the boiling point distribution of a solvent dewaxed lubricating oil. The vapor composition 68 distribution curves obtained from bubble point calculations matched well with experimental values. They also indicated that a cubic spline curve fit can be used effectively to represent complex hydrocarbon mixtures. There are other distribution functions such as exponential decay[153], the Weibull function[153] and the Tung function[162] which can be used to represent gas condensates, polymers and colloids. Several single variable distribution functions are available and their applications to continuous thermodynamics are presented in Table 2.12. Cotterman [47] successfully used gamma and beta distribution functions to model supercritical fluid extraction of complex mixtures of saturates, aromatic, oils and resins using propane as the solvent. Since gamma and beta distributions are suitable[47,146,160] for representing heavy oil and its solubility fractions, respectively, they are discussed in detail in the following sections. Gamma Distribution The gamma distribution function[47] has the form: f ( / ) = (/ Y) exp(~———) p aT(a) p ' [2.23] 1 J where r is the gamma function and a, p are constants greater than zero. The distribution function satisfies the normalization equation: f F(I)dI = 1 y [2.24] Table 2.12 Various Single Variable Distribution Functions Distribution Formula Normal[134] | 77]1 ^ 2 exp Mean Variance Application Petroleum[134]& ~ ^ °f ^2^] Polymers[135] " exp[ Gamma[45] - lo) l (’>)\ “ 1 U2/ ti Heavy oils[143] & Polymers[45] expU-lrfil/o] Schultz[146] UV* Heavy Oils & Polymers[132] Exponential 48] ^W^exp [ - ( / - /o) /(^)] Weibull[153] arj(I-Io )n~' e x p [-a (/ -Io )n] Tung[162] tjI Io(I / Io)”-1exp[-(I / Io f] Gas Condensates lo+T| a !/T (1 + 1/ tj) a 2ln{Y{\ + 2lTf)-\T{\ + M rf)f) Petroleum & Polymers[157] IoT{\ + 1/tj) Petroleum & Colloids[157] Io7T(\ + 2 /tj)-[T (\ + \/r j)f cn CO 70 The shift parameter y fixes the origin where F(l) is zero. The mean and variance are given by: [2.25] Mean, i=l Variance a , 2 = JT A',/,2 - 0 , i=i [2.26] where X( is the normalized mole fraction of the ith component; and ; I is the independent variable used to characterize the oil such as molecular weight or boiling point. a, p are given by the following expressions[47]: 2 p =— 9 ,- 7 ; and: [2.27] [2.28] The gamma distribution function is normally used to represent heavy oils using an independent variable such as molecular weight or boiling point in the continuous thermodynamics applications. The variable ranges from zero and infinity[47]. Beta Distribution The beta distribution function[156] is used for gas condensate systems and is defined as: [2.29] where the parameters a and b are greater than unity and the continuous independent variable I varies between 0 and 1. F(l) also satisfies the normalization equation: [2.30] 0 A proper scaling of the independent variable is required to represent the continuous independent variable I such as molecular weight, boiling point, etc. using the beta distribution function. Let CL and Cu represent the lower and upper bound of the independent variable I. If mean and variance are estimated graphically, then the relationships for the mean and variance are as foliows[156]: Mean 9, = Variance a 2 = ' where 0mw or bp and [2.31] [2.32] ( Q ,- Q ) c2mw or sp are calculated using equations 2.31 and 2.32, respectively. The parameters a and b can be evaluated as: 72 and b= % -a 0/ [2.34] The beta distribution function is suitable for hydrocarbon mixtures for which the lower and upper limits are well defined: such as distilled fractions like gasoline, diesel fuel, and/or gas oil fractions. The beta distribution function is also used for crude oils that can be distilled completely (10 0 % overhead). Typical gamma and beta distributions are shown in Figure 2.6. SemiContinuous Mixtures Many hydrocarbon mixtures contain components that cannot be easily fit with a continuous distribution. Such components can be represented as discrete components. Examples of systems to which semicontinuous thermodynamics have been applied are solvent-hydrocarbon mixtures in SFE processes[40], light hydrocarbons in gas condensate mixtures[47] and solvents in polymer solutions[47]. The accurate description of semicontinuous mixtures requires the representation of both the continuous and discrete components. Hydrocarbon mixtures contain many discrete components. These discrete components can be described as weighted Dirac delta functions[163] using a limiting procedure. Let p(l) be the distribution function representing a mixture containing many discrete and continuous components. Normalization equations can be written as: 73 Figure 2.6 Typical Gamma and Beta Distribution Functions G am m a Distribution Beta Distribution Distribution Function, F(l) \ Independent Variable, I 75 \p (l)d l = / 1.0 [2.35] and the discrete components can be represented as foilows[45]: [2.36] where * is the weighting factor of the Dirac delta function for each discrete component I. The integral of the Dirac delta function is unity and Equation 2.36 can be transformed as follows for k discrete components: Z * . +77 = l [2.37] where Xi is the mole fraction for discrete component i in the overall mixture and -n is the mole fraction of the continuous fraction. The continuous distribution can be described as: jF (I)d I= 1 i [2.38] The above procedure can be applied conveniently to molecular models such as equations of state (EOS) or expressions for the Gibbs free energy for semicontinuous mixtures. This model also can be easily extended to represent many continuous mixtures with each weighted by an overall mole fraction, n,, and each continuous distribution being represented by Fj(l). This procedure is useful when attempting 76 to describe systems with multimodal continuous represented by a sum of single modal distributions. distributions that are An example for such an application is the representation of solubility fractions such as saturates, aromatics, resins and asphaltenes as multimodal continuous distributions. The sum of these fractions represents an overall heavy oil distribution with each solubility class being represented by the same distribution variable. The multimodal continuous distribution can be conveniently fit to most suitable distribution functions. The normalization equation can be written as follows[47] [2.39] j ] and each continuous fraction can be represented as[45]: [2.40] Flash Calculations It is necessary to establish procedures to perform flash calculations using a suitable equation of state to apply phase equilibrium thermodynamics to chemical engineering process design. A schematic for isothermal vapor and liquid flash for a continuous mixture is shown in Figure 2.7. Isothermal flash calculations provide liquid and vapor phase compositions and relative amounts of the liquid and vapor phases from information such as feed composition, pressure and temperature. 77 Figure 2.7 Schematic for Flash Calculation of Continuous Mixtures Feed Vapor 79 For a continuous mixture with I as the independent variable, the normalization equation[47] can be written as: \F (l)d ] = \ 1 [2.41] For semicontinuous mixtures, the normalization equation[47] can be written as: j^ x i +r1\F {I)d I = \ l i [2.42] where k is the number of discrete components. The feed stream is related to the liquid and vapor outlet streams through the material balance equation. For every discrete component i, the material balance equation[47] can be written as: * . = & , + 0 -5 )* , [2.43] where £ is the fraction vaporized, and zh x and y, are mole fractions of the discrete component i in the feed, liquid and vapor, respectively The material balance equation is written as follows for a continuous fraction present in the semicontinuous mixture: Vf F f (I) = £rjvF y(I ) + ( 1 - £)tjlF l(I) [2.44] where superscripts F, V and L designate feed, vapor and liquid phases, respectively. 80 The liquid and vapor streams are assumed to be at thermodynamic equilibrium. The equilibrium condition in terms of the chemical potential is written as: for discrete component i[47]: p v = n,L\ [2.45] for continuous fraction with I as the independent variable //( /) = //(/); [2.46] The objective of the flash calculation is to simultaneously solve the material balance (Equations 2.43 and 2.44) and the phase equilibrium conditions (Equations 2.45 compositions. and 2.46) equations to estimate equilibrium phase The equilibrium condition for continuous and semicontinuous mixtures can be obtained from the equation of state most suitable for the pressure, temperature and type of the mixture (polar or nonpolar) under investigation. A relationship can be developed between the vapor and liquid phases so as to give sets of nonlinear equations. However, extension of this to flash calculations is more difficult because of the introduction of a third distribution function to represent the feed. The solution for continuous and semicontinuous mixtures can be solved only approximately, because there is no universal distribution function available to represent all three streams so as to solve the material balance and phase equilibria equations. This shortcoming was overcome by the introduction of quadrature technique[151]. 81 There are two methods available for solving the material balance and phase equilibria equations: 1. method of moments[164] and 2. quadrature method [154] The method of moments (MOM)[47] is normally used for bubble point and dew point calculations. Cotterman used MOM to demonstrate its applicability for flash calculations involving a 40:60 mole % C 02-paraffin mixture. This method assumes the same distribution function is adequate for the feed, vapor and liquid streams to satisfy the material balance equation while performing the flash calculation. A universal distribution function can not be assumed to represent the feed, vapor and liquid streams for a system involving bitumen-propane. The estimated vapor and liquid compositions obtained from flash calculations did not sum up to give the feed composition. The quadrature method utilizes a collection of quadrature points to represent feed, vapor and liquid streams. The quadrature points are determined from a class of orthogonal polynomials. The quadrature method has been used successfully for phase equilibrium calculations using many different distribution functions. Quadrature Method The quadrature method[47] provides a solution for flash calculations of continuous and semicontinuous mixtures by introducing numerical integration into the integral algebraic equations, Equations 2.44 and 2.46. 82 The Gaussian quadrature method provides an efficient means of integrating the continuous distribution function by summing a finite number of weighted functions evaluated at specified values of the integration variables. These values are called quadrature points. The material balance and phase equilibria equations must be satisfied exactly at the quadrature points. For s quadrature points, the normalization equation can be written as: [2.47] P where w(lp) is the weighting function; and F(lp) is the function to be integrated at the quadrature point, lp. Quadrature integration is discussed in detail in the numerical methods literature[165,166]. The distribution of the mixture represented by the independent variable I for each stream (feed, vapor and liquid) is described by a collection of quadrature points instead of continuous distribution functions. The continuous distribution equations 2.31 and 2.33 can be replaced by s sets of the following equations: r1FF F(Ip) = £ if F v(I p) + ( \ - g)tjLF L(I f )\ and; [2.48] [2.49] where s represents the number of quadrature points. 83 Thus, the quadrature method is analogous to the pseudocomponent procedure. The quadrature points and weighting factors in the quadrature method are not arbitrary and are determined from a class of orthogonal polynomials for the s quadrature points. Tabulated values of quadrature points and weighting factors have been prepared and reported by authors Abramowitz and Stegun[167], The s number of quadrature points chosen is equivalent to 2s1 randomly chosen pseudocomponents. The required number of quadrature points is determined where the predicted phase equilibrium compositions remains constant with variation in number of quadrature points. A typical procedure for optimizing the number of quadrature points required to represent complex mixture like oil sands bitumen is presented in Chapter 4. Normally 8 to 10 quadrature points are sufficient to represent heavy oils[47]. Flash calculations were performed using s quadrature points to represent the hydrocarbon mixture to provide phase compositions for the liquid and vapor streams, at the corresponding quadrature points. Equations of State The equations of state (EOS) used in the flash calculations are selected based on the system under investigation. There are many equations of state available and their usefulness for different applications depends on many factors: the pressure and temperature of the system, the molecular model of the hydrocarbon mixture under investigation, the number of parameters in the EOS, and the prediction capabilities. 84 Four equations of state have been used predominantly for phase equilibrium calculations involving bitumen-solvent systems: 1. SRK cubic equation of state[119] 2. PR cubic equation of state[41] 3. Perturbed Hard Chain (PHC) equation of state[48] 4. Statistical Associating Fluid Theory (SAFT)[168, 169] The SRK[119] and PR[41] equations of state have been widely used in phase equilibria calculations involving gas condensate, crude oil, heavy oils, and bitumens and when combined with a pseudocomponent lumping scheme their applicability has been well established. The SRK cubic equation of state[119], the PR cubic equation of state[41] and the PHC[48] equation of state have been used to describe the Cold Lake bitumen-C0 2 system by Radosz[170] using twelve pseudocomponents. The SAFT[168] equation of state has been used to study bitumen-C0 2 phase behavior at high pressure and high temperatures[171]; however, the SAFT EOS will not be used in this work as it requires the evaluation of too many parameters before it can be efficiently used. Perturbed Hard Chain Equation of State The original PHC equation of state proposed by Beret and Prausnitz[172] was valid for simple and complex molecules and for the entire array of fluid densities, ranging from ideal gases to liquids and highly compressed gases. The PHC EOS was based on a hard core reference fluid with a square well attractive potential. This equation of state's ability to predict the second virial 85 coefficient was poor at moderate densities. This necessciated empirical corrections to improve second virial coefficient predictions. Cotterman[48] has upgraded the PHC equation of state for both the high and low density regions. Cotterman[48] retained the hard core reference fluid; however, he chose a Lennord-Jones model for the attractive potential which introduced a temperature dependent soft core volume as proposed by Barker and Henderson[173]. This modification improved the prediction of pure fluid properties and provided flexibility in the selection of mixing rules for mixtures. The phase equilibria criteria for continuous and discrete mixtures are written as: for a discrete component i[47]: ^ = HjL\ [2-45] for a continuous fraction with I as the independent variable ^ ( / ) = / i t (7); [2.46] The Helmholtz free energy is used in the PHC theory as a starting point because of the specific volume term in the equation. The PHC EOS was developed for application to low, medium and high density range fluids compared to the cubic equations of state (RK, SRK, PR) which uses pressure and temperature limits for selection of the appropriate EOS. The molar Helmholtz free energy, Ar, represents the effect of intermolecular forces, and is defined as the difference between the total Helmholtz free energy for the mixture and an ideal gas mixture at the same temperature, volume and composition as: 86 Ar(T y,ni,n2....) = A(Ty,nlln2....)-A 1G(TtV,n],n2....) [2.50] The residual chemical potential for a discrete component i can be found by differentiating the residual Helmholtz energy, nAr, with respect to the moles of species i, n, holding the temperature, volume and the number of moles, nk, k*i, constant. n ,= 3iAr cki. [2.51] Tyynt ,k*i Following Salacuse and Stell[135], the residual chemical potential for a continuous mixture is the derivative of the residual Helmholtz free energy, nAr, with respect to the extensive distribution function nF(l) at a fixed value of T, V and characterizing variable, l=l+: chAr 3iF(I+) [2.52] TVJ* =1 The molar residual Helmholtz free energy for pure components and mixtures can be written as a sum of repulsive and attractive contributions[48]: Ar = Ar(repulsive) + Ar(attractive) The repulsive contribution [2.53] (generalized Carnahan-Starling[174] expression) can be written as a function of reduced molar volume v and Prigogine’s parameter[174], the external degrees of freedom c, as: 87 — (repulsive) = f rep(c,v) [2.54] The Rrigogine’ assumption at high densities states that the density dependent degrees of freedom can be treated as equivalent transitional degrees of freedom. The parameter c reflects the extent to which a molecule can exercise rotational and vibrational motions which are affected by its neighbors. Larger molecules have higher external rotational and vibrational motions than smaller molecules. Thus, they are more flexible and possess greater external degrees of freedom c. The attractive contribution is determined from pure fluid data and computer simulation as a function of reduced temperature, T, reduced molar volume, v, and the parameter c as: A' — where (attractive) = f an(c,v J ) v is the reduced molar volume; T is the reduced temperature (T = TIT * ) ; v* is the molar soft-core volume (v = v/v*), cm3/mol; and; T is the characteristic temperature (e^/cA:B), K; e is the potential energy per unit surface area; [2.55] 88 q is external molecular surface area; and; kB is the Boltzmann’s constant The ratio of the hard core (temperature dependent) diameter, d, to the soft core (temperature independent) diameter a as a function of reduced temperature, T (T = T / T C ) is given by: a 1+ T (0.52915 + 0.0031817 T) Equation 2.56 is valid for reduced temperatures T in the range from 0 to 5.0. Two characteristic volumes: the soft core volume v* defined by Barker and Henderson[173] and hard core volume v+ are related through the ratio of their diameters as: v*=v+(d/a)3 [2.57] where d is the hard core diameter; and; o is the soft core diameter. The hard volume is defined as the volume occupied by a large molecule with r equal segments as: [2.58] 89 and the reduced volume is defined as: v= — ^ [2.59] N„rv where Na is the number of molecules. The PHC equation of state uses three parameters for each pure component: v*, eq/kB(cT*) and c which are characteristic of the molar soft core volume, the molecular potential energy and the number of external degrees of freedom of the molecules, respectively. The pressure explicit equation of state is obtained by differentiating the residual Helmholtz energy with respect to volume: p _nR T r aiar V [2.60] r.a/ta The three pure component parameters v*, sq/k0 (cT*) and c have been fit to vapor pressure and liquid density data for a wide range of pure compounds by Cotterman[48], He also obtained correlations for the three parameters as a function of molecular weight for various classes of compounds such as alkanes, alkylbenzenes and alkylcyclohexanes. Cotterman[48] showed that v* and cT* are proportional to the molecular weight and T* approaches a constant value at high molecular weights. Thus, c, cT* and v* can be extrapolated to high molecular weight ranges where experimental data on vapor pressure and liquid densities are scarce. 90 Cotterman[47] considered a semicontinuous propane (91 wt%) and two continuous mixtures of oil (6 mixture consisting of wt%) and resin (3 wt%) fractions, respectively, for phase equilibrium calculations using continuous thermodynamics and the PHC EOS[48], The oil and resin mixtures (as shown in Figure 2.8) were assumed for modeling purposes and no phase equilibrium composition measurements were made. The continuous distribution used for representing the feed oil and resin mixture is presented in Figure 2.8. The molecular weight range of the oil was 200 to 500 g/gmol and that of the resin was 450 to 850 g/gmol. The PHC EOS parameters were estimated using alkane based correlations. The liquid and vapor phase compositions were obtained for pressures ranging from 5 to 10 MPa and temperatures from 375 to 450 K. and the corresponding oil and resin yields in the vapor and liquid streams, respectively, were estimated. A typical vapor liquid composition distribution obtained for the oil and resin mixture at presented in Figure 2.9. 8 MPa (Pr=1.9) and 398 K (Tr=1.08) is Suitable pressure and temperature conditions were identified for maximizing the concentration of the oil in the vapor stream and of the resin in the liquid stream. Cotterman[48] also measured phase equilibrium compositions for two systems containing propane and oil mixtures with a flow cell apparatus at temperatures near 400 K (Tr=1.08) and pressures up to 5.5 MPa (Pr=1.3). Two different oils (synthetic) were used for this study. The first oil was rich in aromatics and the other was rich in saturates. The properties of the two oils are presented in Table 2.13. The phase equilibrium measurements were performed 91 Figure 2.8 Feed Distribution for Oil and Resin used by Cotterman[47] Molar Distribution (x10s) 92 Molecular Weight, g/gmol 93 Figure 2.9 Molecular Weight Distributions for Oil and Resin Fractions in Vapor and Liquid as Predicted by Cotterman Mixture at 8 MPa (Pr=1.9) and 398 K (Tr=1.08)[47] Molecular Weight, g/gmol 200 400 600 Molecular Weight, g/gmol 800 95 Table 2.13 Properties of Saturates and Aromatics-Rich Oils used by Cotterman[48] Oil Property Saturates-rich Specific Gravity, @333K C/H Ratio Saturates, wt% Aromatics, wt% Molecular Weight, g/gmol 0.8355 6.23 8 8 .6 11.4 340.0 GC Simulated Distillation. Normal Boilina Point. K 575.5 Initial 1 0 wt% off 622.4 670.6 50 wt% off 90 wt% off 711.7 Final 744.2 Aromatics-rich 0.9237 7.56 37.9 62.1 310.0 581.5 624.0 668.4 707.7 746.9 96 for the propane and saturates-rich oil in the temperature range from 374.4 K (Tr=1.01) to 413.5 K (Tr=1.12) and in the pressure range from 3.1 MPa (Pr=0.73) to 5.5 MPa (Pr=1.3). The propane-to-oil weight ratio ranged from 3.5 to 4.1. The phase equilibrium measurements were performed for the propane and aromatics-rich oil in the temperature range from 392.7 K (Tr=1.06) to 413.5 K (Tr= 1.12) and in the pressure range from 3.1 MPa (Pr=0.73) to 5.5 MPa (Pr=1.3). i The propane-to-oil weight ratio ranged from 3.4 to 3.8. Measured solubilities were correlated using the PHC EOS[48] using continuous thermodynamics. A plot showing the measured and predicted vapor phase composition of the saturates-rich and aromatics-rich oil are presented in Figure 2.10 and 2.11. Calculated and experimental results were in good agreement when the interaction parameters were fine tuned. However, the prediction failed near the critical region of the mixture (5.5 MPa and 400 K, estimated) where most of the equations of state fail. Phase Behavior Studies on Bitumen Systems Phase equilibrium calculations have been reported for oil sands bitumens and various solvents[176-178], A pseudocomponent lumping scheme was used to the represent bitumens in the phase equilibrium calculations. Mehrotra et al.[176] performed density and gas solubility predictions using the PR EOS[41] using carbon dioxide and ethane as solvents and the Athabasca and Peace River bitumens as feedstocks. Temperatures ranged from 293 to 373 K and the pressures were varied from 1 to 6 MPa. The Athabasca and Peace 97 Figure 2.10 Measured and Predicted Vapor Phase Composition of a Saturates-rich Oil[48] 98 Heavies in vapor Phase, wt% Experimental Data • 392.5 K A 413.5 K 0.10 --— 0.01 30 Calculated Results all kjj =0 k|> adjusted to mixture Data 40 Pressure, bar 50 60 99 Figure 2.11 Measured and Predicted Vapor Phase Composition of an Aromatics-rich Oil[48] Heavies in vapor Phase, 100 Pressure, bar 101 River bitumens were represented by 5 and 4 pseudocomponent lumps, respectively. The estimated properties for the two bitumen pseudocomponents are presented in Table 2.14. The critical pressures, critical temperatures and the acentric factors for the two bitumens were estimated using four different sets of correlations proposed by Lee-Kesler[43], Whitson[42], Huang-Daubert[179] and Bergman-Cavett[180]. The critical properties estimated from the four sets of correlations are presented in Table 2.15 and 2.16. The estimated critical pressures, critical temperatures and acentric factors for pseudocomponents 1 and 2 (HYP 1 and HYP2) were close to each other. However, for the heavier pseudocomponents, the differences in the estimated properties were much larger. The estimated properties were used with the PR EOS[41] to predict carbon dioxide solubilities in bitumen and densities to determine which critical property correlation was superior. The predicted carbon dioxide solubilities in the Athabasca and Peace River bitumens were compared to the measured solubilities. The results indicated that Lee-Kesler correlations[43] gave the best overall predictions of carbon dioxide solubilities. The solubilities of carbon dioxide in Athabasca and Peace River bitumens and of ethane in the Peace River bitumen were predicted using critical properties estimated by the Lee-Kesler correlations[43] and the PR EOS[41). Temperature and pressure combinations were chosen to match the conditions of the measured solubility data[181]. Comparisons have been made between the predicted and experimentally measured solubilities and are presented in Figures 2.12 and 2.13. The solubility and density predictions made with a 102 Table 2.14 Pseudocomponents for Athabasca and Peace River Bitumen and their Estimated Properties[176 ] Pseudo Components NBPa) (K) Specific Gravity Molecular Weight (g/gmol) 0.7939 0.8291 0.8955 1.0599 1.1580 142.6 192.8 290.1 508.5 1092.8 0.7994 0.8990 0.9679 1.0927 145.0 205.0 335.0 925.0 wt% Athabasca Bitumen HYP1 HYP2 HYP3 HYP4 HYP5 453.0 523.0 623.0 773.0 1011.0 1.8 5.0 1.8 71.4 2 0 .0 Peace River Bitumen HYP1 HYP2 HYP3 HYP4 "ij 453.0 553.0 693.0 973.0 Normal Boiling Point 4.3 12 .8 25.7 57.3 Table 2.15 Estimated Critical Properties of Athabasca Bitumen Pseudocomponents[176] Method Property HYP1 HYP2 HYP3 HYP4 HYP5 Lee-Kesler[43] Tc, K Pc, MPa CO 637.02 2.46 0.46 703.86 1.95 0.61 801.2 1.54 0.81 971.87 1.43 0.92 1176.68 0.79 1.50 Whitson[42] TCl K Pc, MPa G> 641.17 2.33 0.40 708.74 1.85 0.52 807.70 1.48 974.40 1.50 0.90 1178.14 0.73 1.28 Tc, K Pc, MPa (D 633.09 2.43 0.46 694.91 1.94 0.65 780.66 1.44 896.80 1030.86 0.54 2.31 Tc, K Pc, MPa 0) 633.36 2.57 0.46 694.91 2.51 0.61 808.31 1.51 0.69 Huang-Daubert[179] Bergman-Cavett[180] 0 .6 8 1.0 1 1.0 2 1.52 961.60 1.39 1.0 0 1121.77 1.6 6 2.40 103 Table 2.16 Estimated Critical Properties of Peace River Bitumen Pseudocomponents[176] Property HYP1 HYP2 HYP3 HYP4 Lee-Kesler[43] T c, K Pc, MPa 0) 638.72 2.49 0.46 750.24 2.13 0.61 878.83 1.46 0.91 1121.49 0.72 1.54 Whitson[42] T c, K Pc> MPa 0 642.64 2.36 0.40 754.04 1.96 0.52 884.60 1.38 0.76 1128.26 0.70 1.25 Huang-Daubert[179] To, K Pc, MPa © 648.17 2.41 0.37 740.83 0 .6 6 855.48 1.40 1.17 1013.06 0.60 2.16 Tc, K Pc, MPa (D 642.57 2.43 0.44 753.80 2.15 0.57 880.56 1.35 0.78 1152.70 1.07 1.26 Method Bergman-Cavett[180] 2 .0 2 104 105 Figure 2.12 Solubility of Carbon Dioxide in Athabasca Bitumen[176] Carbon Dioxide Solubility, Weight% 106 273 293 313 333 Temperature, K 353 373 107 Figure 2.13 Solubility of Carbon Dioxide in Peace River Bitumen[176] Carbon Dioxide Solubility, Weight% 108 Pressure, MPa 109 pseudocomponent lumping scheme, the Lee-Kesler correlations[43] for critical property predictions and the PR EOS[41] provided good agreement with experimentally measured data. Lu et al.[177] compiled measured solubility data for pure gases such as methane, carbon dioxide, nitrogen, and ethane in Athabasca bitumen. The selected data were correlated using a pseudocomponent lumping scheme and PR EOS[41]. The Athabasca bitumen was treated as a single pseudocomponent in the gas solubility predictions. The critical properties for the Athabasca bitumen were obtained from various sources as listed below: • Set I From Fu et al.[182] • Set II. The critical properties were obtained by extrapolating data complied by Reid et al.[126] at a molecular weight of 544 g/gmol. • Set III From Lin[183] • Set IV From Lai et al.[184 ] • SetV From Mehrotra et al. [185] The characterization parameters used for carbon dioxide solubility in Athabasca bitumen are presented in Table 2.17. The authors indicated that a satisfactory correlation was obtained between predicted and measured carbon dioxide solubilities in the Athabasca bitumen for the five sets of critical properties. However, Set III gave the more acceptable deviations from the experimental data relative to the other four sets of critical parameters. Jamaluddin et al.[178] presented a method to predict carbon dioxide solubilities and carbon dioxide saturated solute densities in heavy oils and 110 Table 2. 17 Critical Parameters used by Lu et al. [177] to Predict Solubility of C 0 2 in Athabasca Bitumen Set Critical Temperature (K) Critical Pressure (MPa) Acentric Factor I 911 0.692 1.659 II 810 0.963 1.080 III 824 1.267 1.231 IV 893 0.770 1.485 V 930 1.503 0.792 various Canadian bitumens. The authors used the heavy oil or bitumen as a single component and the modified Martin cubic equation of state[186] to predict phase equilibrium compositions. Typical properties of the C 02l heavy oils and the bitumen used for phase equilibrium calculations are presented in Table 2.18. The predicted carbon dioxide solubilities and carbon dioxide saturated heavy oils densities for four bitumens were compared with experimentally obtained values[187] along with predicted values by Kokal and Sayegh[188] and Mehrotra and Svrcek[187] using the PR EOS[41], Carbon dioxide solubilities in Cold Lake bitumen at two different temperatures and at pressure ranging from 2 .0 to 12 .0 MPa are presented in Figure 2.14. The following conclusions were reported: • The modified Martin EOS[186] predicted gas solubilities in bitumens that matched well with the experimentally determined values. • The Martin[186] and PR EOS[42] were both able to predict C 0 2 solubilities in bitumen. • The C 0 2 saturated bitumen densities predicted by the modified Martin EOS were superior to the corresponding densities predicted by the PR EOS[41], Yu et al.[170] and Huang and Radosz[189] conducted experiments on mutual solubilities of supercritical carbon dioxide and the Cold Lake bitumen up to 523 K and 16 MPa. The PHC EOS[48] along with SRK[119] and PR[41] equations of state have been used successfully by Yu et al.[170] in phase equilibrium calculations involving the Cold Lake bitumen and carbon dioxide using a pseudocomponent lumping scheme. The estimated physical properties 112 Table 2.18 Properties of C02, Heavy Oils and Bitumen used by Jamaluddin et al. [178] Component MW (g/gmol) CM o o 44.01 Tb (K) - Tc (K) Pc (MPa) 304.1 7.38 0.239 CO P (kg/m3) - 951 857 859.0 Oil A Oil B Oil C 350.0 236.0 345.0 690.7 565.8 665.0 870.7 746.5 817.6 1.39 1.76 1.09 0.925 0.692 0.995 Athabasca Bitumen 595.0 935.9 1089.1 0.78 1.363 1074 Cold Lake Bitumen 533.0 870.2 1034.5 1.03 1.184 10 10 Peace River Bitumen 527.5 8 6 8 .0 1044.6 1.09 1.166 1025 Wabasca Bitumen 446.5 787.0 960.0 1.127 1.10 0 1007 113 Figure 2.14 Comparison of Carbon Dioxide Solubility in Cold Lake Bitumen and Saturated Liquid Density by Modified Martin EOS with Measured Values[178] 0.0 20.0 40.0 60.0 8CI.0 Saturated Liquid Density, kg/m3 C 02 Solubility, mole% 1200.0 1100.0 J__ 299 K -s- 350 K 1000.0 900.0 800.0 —!_______ I_______ I_______ !_ 3.0 6.0 9.0 12.0 Pressure, MPa 15.0 of the bitumen pseudocomponents used for the phase equilibrium calculations are presented in Table 2.19. The PHC EOS reasonably predicted C 0 2 solubilities in the bitumen phase (Figure 2.15) except at high pressure (16.0 MPa) when the mixture approached its critical point. The SRK[119] and PR[41] EOS predictions also gave reasonable representation of the experimentally obtained solubilities of C 0 2 in bitumen except near the critical point of the mixture (Figure 2.16). The solubility of bitumen in the C 02-rich phase predicted by PHC[48] EOS was in excellent agreement (Figure 2.17) with experimentally obtained values except at lower temperatures, 373 K, near the critical point. SRK[119] EOS predicted solubility was in better agreement (Figure 2.18) with experimentally obtained values than the PHC[48] predicted values at 323 K in the 4 to 16 MPa pressure range. This was explained by the relative inability of PHC[48] EOS to predict phase equilibrium compositions near the critical point of the mixture. Huang and Radosz[190] also conducted studies on the effect of molecular lumping on the phase equilibrium calculations for C 02-bitumen systems using PHC[48] and SRK[119] EOS. They concluded that a uniform lumping technique was adequate to predict vapor-liquid equilibria for hydrocarbons such as petroleum derived distillates that exhibit a Gaussian distribution. Nonuniform lumping was found to provide a better representation of resids and bitumens for which non-Gaussian distribution functions such as gamma distributions are applicable. The distillable fractions (fractions boiling below 811 K) were approximated by more pseudocomponents than the nondistillable portion. Huang 116 Table 2.19 Estimated Properties of Bitumen Pseudocomponents[170] a) NBP Range (K) wt% Molecular Weight (g/gmol) 1 433 - 463 0.3 131.2 0.852 1.824 2 463 - 483 0.4 146.0 0.862 1.804 3 483 - 503 0.9 158.8 0.869 1.788 4 503 - 523 1.6 172.4 0.877 1.772 5 523 - 548 2 .6 188.7 0.885 1.754 6 548 - 573 3.3 208.3 0.894 1.734 7 573 - 598 3.9 229.3 0.904 1.715 8 598 - 623 4.1 251.8 0.913 1.695 9 623 - 648 4.2 275.9 0.922 1.675 10 648 - 723 25.8 393.3 0.960 1.595 — 14.8 813.0 1.0 0 0 1.551 — 38.1 2580.0 1.033 1.394 No. 11 12 Normal boiling point Density (g/cc) H/C Ratio 117 Figure 2.15 Measured and PHC Predicted Weight Fraction of C 0 2 in Bitumen[170] Weight Fraction of C 02 in Bitumen 118 Pressure, MPa 119 Figure 2.16 Measured, Soave and PR Predicted Weight Fraction of C 0 2 in Bitumen[170] 120 c 0) E 2 m CM O (J o c o 9 o ca O) 3 5 Pressure, MPa ------ RKS Predicted ------ PR Predicted 121 Figure 2.17 Measured and PHC Predicted Weight Fraction of Bitumen in C 02[170] 122 CM O O c c 0) E 3 Si OQ «*O c o o 2 O) o 5 Pressure, MPa 123 Figure 2.18 Measured, Soave and PR Predicted Weight Fraction of Bitumen in C 02[170] Weight Fraction of Bitumen in CO 124 CM Pressure, MPa 125 and Radosz[190] has suggested that six to eight lumps are sufficient to describe bitumen systems with a minimum of two components required to represent overhead fractions. Hwang[39] used five pseudocomponents lumps of equal mole fractions to represent the WR bitumen in phase equilibrium calculations. Lee-Kesler[43] correlations were used to estimate the critical properties of the lumped bitumen components. Hwang[40] predicted extracted phase composition of the bitumen propane mixture for the first, middle and last extraction windows for one SFE condition (10.4 MPa and 380 K) out of a total of five conditions. The agreement between the predicted extract phase (Table 2.20) and the experimentally measured extract phase was good. Subramanian[40] modeled the SFE of PRS bitumen using propane as the solvent. Seven pseudocomponent lumps were used: five pseudocomponents of equal mole fraction in the volatile region and two pseudocomponents in nonvolatile region. The critical properties were estimated using the Lee-Kesler [43] correlations. EOS[41] Extract phase compositions were predicted using the PR and compared with measured compositions (using simulated distillation). A typical fit obtained at 10.4 MPa (Pr=1.2 ) and 380 K (Tr=1.03 ) is presented in Figure 2.19. It is observed from the plot that the match obtained between predicted and experimental values is reasonable considering the uncertainties involved in predicting equilibrium phase compositions near or above the critical pressure and temperature of the solvent for solvent-bitumen mixtures. 126 Table 2.20 Comparison Between Predicted and Experimentally Obtained Extract Phase Compositions for Whiterocks Bitumen Using Propane as Solvent at 10.4 MPa (Pr=2.3) and 380 K (Tr=1.03)[39] Components First Extraction Window E C 98.9 98.4 PS 1 0 .6 PS 2 Propane Middle Extraction Window E C 99.2 99.3 1.2 0.4 0.4 0.3 PS 3 0 .1 PS 4 PS 5 Last Extraction Window E C 99.6 99.7 0.4 0 .1 0 .1 0.3 0 .2 0 .2 0 .1 0 .1 0 .1 0 .1 0 .1 0 .1 0 .0 0 .0 0 .0 0 .0 0 .0 0 .0 0 .0 0 .0 0 .0 0 .0 0 .0 0 .0 E Experimental data C Calculated data All values in Mole % 127 Figure 2.19 Comparison Between Experimental and Predicted Extracted Phase Composition for PR Spring Bitumen-Propane System at 10.4 MPa (Pr=1.2) and 380 K (Tr=1.03) o • o • Experimental Values, mole % o • o • o • • o • o • o • o 128 129 The uncertainties in the modeling of Hwang[39] and Subramanian[40] are related to the following: • The extraction system operates in a semicontinuous mode: batch and continuous with respect to solute and solvent, respectively. The extract phase composition varied as the overall system composition changed with extraction time. The extract phase samples were collected from the separator for each 25 liters (@STP) propane vented through the system This first window extract phase composition (average composition for 25 liters of solvent vented) was compared to the phase composition predicted from the initial overall composition. The overall composition of the second window in the extractor was obtained by material balance using the weight of extract, the amount of propane passed through the system, the extract phase density at the end of first extraction window and the initial overall compositions. The inherent limitations of this method were related to the volume change due to mixing of the feed remaining in the extractor and the fresh solvent coming into the system. These considerations represent the inherent disadvantage associated with semicontinuous extraction systems. • Traditional modeling is done using the extract and residual phase compositions obtained in an equilibrium cell where the solute and solvent is charged to PVT cell in a batch or continuous mode. The phases are allowed to equilibrate for several hours before steady state samples are withdrawn for analysis. The phase compositions are predicted using pseudocomponent lumping schemes and a suitable cubic equation of state. If the predicted 130 phase compositions do not match the experimental values, the interaction parameters are adjusted until the agreement between the model and the experimentally obtained values is attained. The semicontinuous system used by Hwang[39] and Subramanian[40] does not have these capabilities. • Bitumens are comprised of light, moderate and heavy hydrocarbons of different solubility classes. The properties and phase behavior of these medium-heavy hydrocarbons are not well established. Moreover, each bitumen is different in its composition and success or failure obtained with a particular bitumen-solvent system may be specific to that system. The modeling was performed in this dissertation to confirm the experimental trends using continuous thermodynamic principles[47] and the PR EOS[41] instead of trying to predict the extract phase compositions. Modeling Approach The continuous thermodynamics principle was successfully used by Cotterman[47], Peng et al.[156], Mani et al.[159], Matthews et al.[158] and PatelTeja[155] to model crude oil and gas condensate systems using C 0 2 as solvent. The pseudocomponent lumping scheme used an arbitrary grouping pattern and number of components. The phase compositions predicted using the pseudocomponent lumping scheme were very sensitive to the phase equilibrium calculations. Extensive optimization must be done to optimize the number of lumps and the grouping pattern. Continuous thermodynamics uses statistical distributions to represent the hydrocarbon systems of interest. In the quadrature 131 method, the quadrature points and weighting factors are not arbitrary and are determined from a class of orthogonal polynomials. Hence, an attempt was made in this dissertation to use continuous thermodynamics to group the bitumen components. Cotterman[47] and Yu et al.[170] used PHC EOS[48] to predict phase equilibrium compositions for the saturates and aromatics-rich oils and Cold Lake bitumen[170] using propane and carbon dioxide; respectively, as solvents. The PHC EOS[48] predicted phase equilibrium compositions that matched with experimentally determined phase compositions except near the critical conditions of the solvent. Yu et al.[170] indicated that the SRK[119] EOS prediction capability was better than the PHC EOS near the critical conditions. Mehrotra et al.[176] and Lu et al.[177] used PR EOS [41] to predict the solubility of C 0 2 and other hydrocarbon gases in Canadian bitumens. The solubility predicted by PR EOS [41] matched very well with the experimentally measured compositions in the temperature range from 296 to 378 K and for pressures from 2 to 6.0 MPa. Jamalludin et al.[178] predicted solubility of carbon dioxide in four Canadian bitumens using a modified Martin EOS[186] and compared the predicted values with the experimental and predicted values using the PR EOS[41], They have concluded that the PR[41] EOS prediction capabilities are similar to modified Martin[186] EOS for pressures from 2 to 12 MPa and temperatures in the range 296 to 371 K. These observations led to the selection of the PR EOS[41] for phase equilibrium calculations. CHAPTER 3 EXPERIMENTAL APPARATUS AND PROCEDURES The supercritical fluid extraction (SFE) system was built by Autoclave Engineers Inc., Erie, Pennsylvania. This system was designed for semi- continuous extraction of complex hydrocarbon mixtures such as crude oil, heavy oils and coal liquids using carbon dioxide as the supercritical solvent. Detailed descriptions of the original experimental system and its components have been reported by Hwang[39] and Subramanian[40], The system was modified in the course of this study by the addition of a data acquisition system from Keithley Metrabyte to monitor flow rates and extract phase densities on a continuous basis. The pressure control valve originally located upstream of the extractor was moved to the downstream side of the densitometer to stabilize the system. The system can be conveniently divided into four subsystems: a) the supercritical fluid supply system; b) the extractor and densitometer assembly; c) the data acquisition system; and d) the separator assembly. A schematic of the supercritical fluid extraction system used in this study is presented in Figure 3.1. Supercritical Fluid Supply System The supercritical fluid supply system consisted of a high pressure, positive displacement pump and a dynamic, high pressure fluid circulation pump. 133 Figure 3.1 Schematic of the Supercritical Fluid Extraction System Temperature and Pressure Readout Back Pressure Vah/B — k5c— Sepirator F tow M ste^ eH Row n * \ Totalizer^ Liquid Sample Propane to Flare 135 The positive displacement system was acquired from D.B. Robinson Associates, Edmonton, Canada and consisted of a cylinder, piston, and power pack to pump the fluids at high pressures. The pump was designed to handle liquids at flow rates ranging from 0 to 1800 cm3/h at a maximum pressure of 135 MPa and temperature of 1033 K. The volumetric capacity of the pump cylinder was 500 cm3. The fluid was transferred from the storage tank into the pump cylinder, pressurized and pumped into the supercritical fluid extractor. This pump was capable of pumping fluid at a rate of one hundredth of a cm3/min and the volume discharged by the pump could be measured to within 0 .1 cm3 accuracy. The fluid pumping system consisted of a storage cylinder (C0 2 or propane), a nonreturn valve and line filter, a high pressure liquid circulation pump, and a circulating refrigerant bath to cool the pump head. Liquid was withdrawn from the storage cylinder (by means of a siphon tube for C 02) and was introduced into the extractor by means of the high pressure, positive displacement liquid metering pump (LDC Analytical, model 396-89). The maximum working pressure of the pump was 41 MPa, and the flow rate could be varied from 46 to 460 cm3/hr. The head of the circulation pump was maintained at 273 K by means of a circulating cooling bath (Fisher Scientific, Model 900) to prevent vapor lock. The constant temperature circulating bath was capable of pumping coolants at temperatures ranging from 258 K to 373 K. temperature controller installed on the circulating temperature to within ± 0.02°C of the set point. bath The maintained the 136 Supercritical Fluid Extractor and Densitometer Assembly The 300 cm3 extractor was rated for a working pressure of 37 MPa at 616 K. The effective volume of the extractor with the stirring device assembly in place was 280 cm3. The extractor temperature was controlled by a three-mode proportional integral temperature controller/indicator installed on the main control module. The extractor assembly consisted of the extraction vessel, a flanged stainless steel closure, a magnetic-drive, packerless stirring device and a pressure gauge mounted on the flange-head. High speed agitation was achieved by rotation of external magnets mounted on the top closure which in turn actuated magnets fastened to the stirrer shaft. A 1/4 to 2 hp adjustable speed DC motor (Electric Motors, Inc.) rotated the stirrer shaft at the desired speed. The extractor assembly was constructed of 316 stainless steel. A metal gasket was used for high temperature (311 K < T < 380 K) and high pressure (>7.0 MPa) applications. An "O" ring Buna-N gasket was used for low temperature (T <311 K) and low pressure (< 0.69 MPa) applications. An electric furnace (Autoclave Engineers, Inc.) mounted on the outside of the extractor was used to heat the extractor. A three-mode temperature controller provided with a type-K thermocouple was used to control the temperature inside the extractor. The control thermocouple was located in the thermowell adjacent to the stirrer blades. An eight-point digital temperature indicator mounted on the main control 137 panel monitored the temperatures of fluid in the extractor, densitometer and separator assemblies. The densitometer consisted of two parts: 1) an external measuring cell and 2) a display unit. The external measuring cell (Anton Paar K.G, model DMA 512) was placed on the downstream side of the extractor assembly to measure the extract phase densities at high pressure-high temperature conditions; that is, at supercritical conditions. The recommended operating temperature range of the measuring cell was 253 K to 425 K and its maximum working pressure was 40 MPa. The 0.3 cm OD stainless steel sample tube was excited by two magnetic converters in conjunction with electronic control and amplifier circuits. The frequency of oscillation varied with the variation in the density of the fluid inside the tube. The frequency of oscillation was converted into a period (sec1) of oscillation and displayed on the display unit. The display unit (Anton Paar, model mPDS 2000) mounted on the main control panel indicated the period of oscillation of the sample tube in the measuring unit. The readings were converted to densities using densitometer constants determined in the calibration experiments. The output from the display unit was transmitted to the data acquisition system on a continuous basis where the signals were converted to the period of oscillation and hence to a density. The densitometer was connected to the extractor assembly by 3.2 mm SS tubing and Swagelok fittings. A bypass line was provided to isolate the densitometer when the system was being cleaned. 138 Data Acquisition System The data acquisition system consisted of a power supply, analog/digital (A/D) board, a diagnostic board, a driver interface board, an IBM PC compatible computer and data acquisition software. The data acquisition boards and software program were obtained from Keithley Metrabyte. The power supply board, model PWR 100, provided the voltage (±15 V) required for the diagnostic, A/D and driver boards. The primary function of the diagnostic board, model MDG-1, was to test the system for proper operation. The A/D board, model MAI-16, could receive inputs from 16 input points simultaneously and could accept four standard input voltages (±10, ±5, ±2.5 and ±1.5 V). Special voltage ranges could be accepted based on user installed resistors. The driver board, model MDB-64, was installed in an expansion slot of the IBM PC and interfaced directly with the data acquisition software. Labtech Notebook XE for DOS operating system with universal drivers was used to acquire data from the SFE system. Labtech Notebook was capable of receiving digital inputs through the A/D boards, sending control output signals to the controllers and communicating through RS 232 C serial ports. The power supply, diagnostic board, A/D board and driver boards were connected in series by means of a 52 pin connector cable. The Anton Paar densitometer and the EG&G flow meter were connected to the A/D board and sent analog signals (0-5 V) which were digitized at the A/D board and transferred to the PC for storage. Data were collected from the two measuring devices every second, and these 139 digital inputs were converted into density and flow rates and were graphically displayed on the monitor during the course of the experiment. SFE Separator Assembly The SFE separator assembly consisted of the fluid transfer line between the densitometer and the separator, a back pressure regulator valve, a separator vessel, the liquid sample withdrawal valve and a gas outlet line with a flow totalizer. The separator vessel had a capacity of 500 cm3 and was constructed of 316 stainless steel. The maximum working pressure of the vessel was 13.7 MPa at 505 K. The cover of the vessel was fixed to the vessel with quick disconnect couplings and was sealed with Buna-N O-ring seals. A type-J thermocouple was used to monitor the temperature inside the separator. The ball valve mounted at the bottom outlet was used to withdraw the liquid phase samples from the separator. The extraction system pressure was maintained by a manually operated back pressure regulator (Haskel Engineering & Supply Co.) installed between the extractor-densitometer system and the separator. It also served as a pressure reduction valve through which the supercritical extract phase flowed to the separator where the solute was recovered when the extract phase flashed. The fluid transfer lines between the extractor, the densitometer and the separator were traced with heat tape to maintain the transfer line temperatures the same as that of the extractor. The line temperatures were monitored by a type-J thermocouple located at the outlet of the densitometer. A 140 proportional controller (Briskheat Corporation) supplied the power necessary to the heating tape for temperature maintenance. The solvent separated from the solute in the separator was vented after passing through a flow meter and a flow totalizer. When propane was used as solvent, the vent line was connected to a flare in the evacuation hood where the propane was burned. The separator and the vent line were protected by a pressure relief valve. An electronic flow meter (EG&G Flow Technology, model FTO-4NIYA-GHC-5) was used to measure the flow rate and cumulative volume of the solvent vented through the system. The flow rate data were continuously monitored and stored by the data acquisition system. A mechanical flow totalizer connected in series with the electronic flow meter also measured the cumulative volume of the gas vented through the system. Calibration of the Densitometer The densitometer was calibrated before being used to monitor the densities of the supercritical extract phases. The fluid density was calculated from the recorded period of oscillation according to the following equation. P = a [ t 3] - B where T is the period of oscillation (ms); and; A and B are instrument constants. [3.1] 141 The instrument constants were determined during the calibration procedure. Air and distilled water were used to calibrate the densitometer. The calibration procedure was as follows: 1. The densitometer U-tube was cleaned with toluene, dichloromethane and acetone. The U-tube was purged with air to remove residual solvent. Purging was continued until the densitometer reading stabilized. The densitometer reading was recorded with air in the vibrating U-tube. 2. The U-tube was completely filled with distilled water using a microsyringe, and the densitometer reading was recorded. Care was taken to insure that no air bubbles were trapped inside the U-tube. 3. The densities of air and water were corrected for temperature (measured for air and water) and pressure of the atmosphere and used to calculate the densitometer constants A and B in Equation 3.1. O// Sands Bitumen Preparation Oil sands bitumen is a complex hydrocarbon mixture which can vary in color from reddish brown to black. Bitumens are semisolid in appearance and viscous to brittle in character. Bitumens have high concentrations of heteroatoms such as nitrogen and sulfur, and they have high concentration of metals such as nickel, vanadium and arsenic. The nickel-to-vanadium ratio is a function of the origin of the bitumen. The bitumens used in the SFE studies were recovered from crushed, screened and well-mixed ore samples by conventional Dean-Stark extraction 142 apparatus using toluene as the solvent. The ores were mined from the Whiterocks, PR Spring, Asphalt Ridge and Sunnyside oil sands deposits of the Uinta Basin (Utah). The extractions were carried out in a Dean-Stark apparatus with Whatman single thickness cellulose thimbles. Sharkskin filter paper was wrapped around the outside of the thimble to retain any fine mineral particles that penetrated the thimble wall. Crushed oil sands (800 - 1000 g) were placed in the thimble and 2000 cm3 of fresh toluene was placed in the extractor reservoir. Extraction was carried out for 12-24 hrs at total reflux depending on the compaction (12 and 24 hrs respectively, for the Asphalt Ridge and Sunnyside oil sands ores) and bitumen content of the oil sands ore. The solvent reservoir temperature was maintained between 385 and 389 K. A water trap between the reservoir and the condenser removed water from the system. Approximately 2000 cm3 of a bitumen/toluene solution (= 1 wt% bitumen) was obtained after each extraction. This procedure was repeated for 10 extractions with fresh oil sands ore being added to the extractor each time. The same toluene/bitumen solution in the reservoir was used for each extraction to concentrate the bitumen and to reduce subsequent toluene evaporation time. Make-up toluene was added every time a fresh sample of oil sands was placed in the extractor. A rotary evaporator (Buchi, Model EL-131S) was used in conjunction with a vacuum pump (Precision Scientific, Model S-35) to evaporate the toluene from the concentrated toluene/bitumen solution. The condenser cooling water and 143 the water bath temperatures were maintained at 288 K and 358 K, respectively. A cold trap, using dry ice, was placed between the rotary evaporator and the vacuum pump to avoid solvent carry over to the vacuum pump and backstreaming of the pump oil into the evaporator. Approximately 1 liter of toluene/bitumen solution was placed in a 2 -liter round bottom boiling flask for evaporation. The flask rotated at 100 RPM. The bulk of the toluene evaporated at 358 K and 6.62 kPa. A sample was withdrawn from the bitumen-toluene solution for simulated distillation (SIMDIS) analysis to determine the residual toluene concentration. Solvent evaporation was continued for 20 more hrs at 5.3 kPa. Additional samples were withdrawn for residual toluene analysis after 4 and 20 hr intervals. The residual toluene content of each bitumen sample was determined by SIMDIS. Bitumen samples (0.2 nl) diluted with carbon disulfide were injected into an HP model 5890 Series II gas chromatograph which used on-column injection. A capillary column model Petrocol EX 2887 (5m x 0.53 mm) with a 0.10 mm thick film was used. Helium was used as carrier gas at a flow rate 13 cm3/min. The temperature programs used for the residual toluene content determination are presented in Table 3.1. A residual toluene concentration of less than 0.1 wt% was achieved for the PR Spring[40], Asphalt Ridge and Sunnyside bitumens after rotoevaporating for 20 hrs at 5.3 kPa with the water bath temperature held at 358 K. 144 Table 3.1 Temperature Programs for Determination of the Residual Toluene Content in Bitumen by Gas Chromatography Initial Temperature (K) Initial Time (min) Ramp (K/min) Oven 243.0 6.0 12.0 653 Injection 313.0 2.0 10.0 673 Column Used: Petrocol EX2887 Final Temperature (K) Helium Flow Rate : 13 cm3/min Final Time (min) 8.75 10.0 145 Experimental Procedures The operating procedures employed for the supercritical fluid extraction experiments are described below, 1. The main electric power switches located on the extractor module were turned on to supply power to the system. 2. A 50 g sample of bitumen was weighed and placed in the extraction vessel. 3 The vacuum was turned on for 10 min to evacuate the inlet lines to the extractor, positive displacement pump cylinder and all the flowlines from the gas storage system to the pump. 4. The inlet valves to the extractor and the isolation valve between the extractor and the positive displacement pump were closed. The positive displacement pump cylinder was filled with C 0 2 or propane from the reservoir until the overflow line from the pump to the flare was filled with liquid solvent. This ensured complete fillup of the pump cylinder with liquid solvent. The positive displacement pump was turned on to pressurize the solvent in the pump cylinder to the desired operating level. 5. The heater on the extraction vessel was activated. The extraction temperature was set at the desired level and maintained by the three­ mode temperature controller. 6 . The solvent in the positive displacement pump was transferred into the extractor until the operating pressure and temperature were attained with the extractor outlet valve closed. The initial and final positive displacement pump readings were recorded. The difference was used to compute the initial charge to the extractor, the volume change upon mixing, and the initial overall composition. The outlet valve from the extractor was opened slowly to fill the downstream lines with system fluid to the back pressure control valve. The pump continued to operate as the lines were being filled. 7. The IBM PC was turned on and the Labtech Notebook software was loaded and set up in the data acquisition mode. 8 . The positive displacement pump was used to feed the solvent to the system during the continuous extraction experiments. The flow rate through the system was maintained at 3.8 cm3/min (at operating pressure) using the flow rate adjustment valves on the positive displacement pump. 9. The proportional temperature controllers were turned on to provide heat to the flowlines between the extractor and separator. 10. When the desired system pressure was reached, the back pressure valve was opened and the extract phase was transferred from the extractor to the separator to maintain the system pressure constant. The system pressure was maintained at the desired level by manually adjusting the back pressure control valve. The pressure was monitored from the pressure gauge mounted on top of the extractor. 11. The supercritical fluid was admitted to the extractor while the system was stirred. The extract phase containing solvent and solute was transferred to the separator through the densitometer. The pressure reduction occurred at the back pressure control valve to facilitate separation of solvent and solute in the separator. The gas (propane) separated from the extract phase was flared in specially designed burners after passing through a flowmeter and flow-totalizer. 12. The densitometer and flow rate readings were monitored and stored every second by the data acquisition system. 13. The product accumulated in the separator was collected by opening the liquid phase sampling valve after 25 liters @STP of propane had passed through the flow-totalizer. The extracted solute sample was weighed and stored for analyses. In each extraction experiment, six samples were collected and identified as extract windows #1 through # 6 . The extraction time, period of oscillation from the densitometer and cumulative volume of solvent that had passed through the system were recorded during sample collection from the separator. Each sample corresponded to an extraction window of 25 liters (@STP) of solvent vented through the flow-totalizer. 14. The extraction was terminated after collecting six samples. The system was allowed to cool to the ambient temperature and the extractor was depressurized. The residue in the extractor was collected, weighed and stored for subsequent analysis. The residue was a fragile solid mass 148 distributed between the stirrer blades and the extractor liner bottom in all experiments. 15. After completion of the experiment, the entire system was cleaned. The cleaning solvent sequence was toluene, dichloromethane and acetone. Finally the system was purged with air overnight to remove any residual solvent and to dry the system. Product Analysis Liquid Product Analysis The Asphalt Ridge and Sunnyside bitumens, all extracts and the residues from the extraction experiments were analyzed using a simulated distillation technique to obtain the carbon number distribution up to C90. The SIMDIS analyses of the samples were carried out using a Hewlett Packard (HP) 5890 Series II gas chromatography (GC). The GC was equipped with on-column injection, a flame ionization detector (FID) and an automatic sampler. A schematic of the GC setup is presented in Figure 3.2. Helium was used as the carrier gas at a flowrate of 20 cc/min. Air, hydrogen and nitrogen were used to sustain the FID flame. Injections were performed using a Hewlett Packard Model A 7673 automatic sampler. The signals from the detector were sent to an IBMPC through a Hewlett Packard 3396 Series II integrator using a Hewlett Packard file server program. The signals were integrated using a slicing program at the integrator and were stored as report files for further computations. A Microsoft 149 Figure 3.2 Schematic of the Gas Chromatograph System Automatic Sampler Sample Tra FID Detector T Gases H2IN„ He &Air \J Capillary V 1 Column I Oven Gas Chromatograph RS-232-C Integrator IBM-PC cn O Windows based SIMDIS program was developed to read the data and obtain the carbon number distributions for totally and partially eluted samples[45], A Petrocol capillary column, Model EX 2887 (5 m length, 0.53 mm OD and 0.10mm film thickness), was used for analyses of the SFE extract samples, the residual fractions, and the saturates, aromatics, and resins fractions obtained from adsorption chromatography of the bitumen. A Petrocol capillary column, model 2887 (5 m length, 0.53 mm OD and 0.50mm film thickness), was used for analyses of the Asphalt Ridge and Sunnyside bitumens. The recommended maximum operating temperature for these columns was 653 K. The temperature programs used for the columns are presented in Table 3.2. Initially a Model 2887 column with 0.5 mm film thickness was used. While using this column, it was observed that the resolution of the chromatogram was good but the amount of sample eluted during a chromatographic run was low. Hence, a EX2887 column was used with thinner film thickness (0.10 mm) at the cost of lower resolution but a greater amount of bitumen elution during the run. The samples were assigned carbon numbers based on the elution patterns of standard samples containing normal alkanes ranging from C10 to Cgo. Three standard samples, namely, Polywax 655 (C2o to Cn0), ASTM PS-12-60N (C12 to C6o) and ASTM PS-18-44N 60N (C, 8 to C44) from Supelco, Inc. were used for calibration purposes. The carbon numbers corresponding to the retention time of the standard sample are presented in Figure 3.3. The standard SIMDIS assumption was used in analyzing the data: the normal alkane eluted 152 Table 3.2 Temperature Programs Used to Obtain Carbon Number Distributions of Bitumen and Bitumen Products by Gas Chromatography Petrocol EX 2887 & 2887 Columns Initial Temperature (K) Initial Time (min) Ramp (K/min) Final Temperature (K) Oven 308.0 4.5 12 .0 653.0 Injection 473.0 2 .0 10 .0 673.0 Helium Flow Rate: 20 cm3/min Final Time (min) 8.75 2 0 .0 153 Figure 3.3 Carbon Number (C 5 to C90) versus Retention Time Calibration Curve for Simulated Distillation Analysis Retention Time, minutes 155 last for a particular carbon number homologous series. Carbon number C 90 eluted last under the oven temperature programs, helium flow rates and standard samples used. The uneluted portion was determined using an internal standard (C i4 to C 17) mixture. A Visual Basic computer program was developed to estimate the carbon number distributions and is presented in Appendix E. The Conradson carbon residues and the pour points of the feedstocks were determined according to the procedures outlined in the ASTM D 189-65 and ASTM D97-66 methods, respectively. The viscosities of the Whiterocks, Asphalt Ridge, PR Spring and Sunnyside bitumens were determined using a Brookfield cup-cone digital viscometer (Model DVT-II+). The viscosities were measured at four different temperatures to establish the relationships between temperature and viscosity for each bitumen. The densities of the bitumens were measured using the procedure developed by AOSTRA[191] for semisolids. The saturate, aromatic, resin and asphaltene (SARA) contents of the feedstocks, extracts and residues were determined by an analytical method developed by Bukka et al.[44,192]. The most commonly used solvents to make the initial separation of bitumen into maltenes and asphaltenes are n-pentane and n-heptane. The solvating power of the alkane increases gradually with increase in the carbon chain length, thus pentane yields quantitatively more asphaltenes than heptane for a given hydrocarbon sample[2]. Moreover, removal of pentane from the maltenes fraction by rotoevaporation can be accomplished at lower atmospheric equivalent temperature relative to heptane; 156 hence, it has gained popularity as the standard solvent for asphaltene separation. The procedure for the SARA fractionation of the bitumens is presented at Appendix C. The elemental (C,H,N,S) analyses of the feedstocks and residual fractions and the molecular weight determinations for the feedstocks were carried out at Galbraith Laboratories, Inc., Knoxville, Tennessee. Galbraith used a Perkin Elmer 240 elemental analyzer. Molecular weights were determined by a three point vapor pressure osmometry method using toluene as the solvent. Gas Analysis The carbon dioxide and propane used for extraction studies were commercial grade gases supplied by Liquid Air Corporation and Wasatch Propane, Inc., respectively. These gas samples were analyzed using a gas chromatograph (Carle, Model Series Sx). presented in Table 3.3. The analyses of the feed gases are 157 Table 3.3 Analyses of the Gases Used as SFE Solvents Carbon Dioxide Propane Supplier Liquid Air Corporation Wasatch Propane, Inc. Grade Commercial Commercial Cylinder Content 23 Kgs 14 Kgs Saturation Pressure (@ STP) 5723 kPa 758 kPa Composition (vol%) Carbon dioxide 99.5 lmpuritiesa) 0.5 Methane: Ethane: Propane: B u ta n e : 0 .2 5.4 93.9 0.5 Air, W ater and Light Hydrocarbons (Methane, Ethane, Propane, etc.) CHAPTER 4 RESULTS AND DISCUSSION Supercritical fluid extraction (SFE) was carried out using the system described in Chapter 3. The bitumens from the Asphalt Ridge (AR) and Sunnyside (SS) oil sands deposits of the Uinta Basin, Utah were used in the SFE experiments using propane as solvent. The SFE of the bitumens from the Whiterocks (W R ) and PR Spring (PRS) oil sands deposits was conducted by Hwang[39] and Subramanian[40], respectively. The four bitumens differ significantly in their physical and chemical characteristics. Thus, it was possible to compare the effect of pressure, temperature, solvent density and composition on the SFE extraction yields, the quality of the extract phases and the nature of the residual fractions for four oil sands bitumens from the Uinta Basin, Utah. Feedstock Characterization Four different bitumens from the WR, AR, PRS and SS oil sands deposits of the Uinta Basin (Utah), were used for the SFE experiments using propane as solvent. The feedstocks were prepared as discussed in Chapter 3 for both characterization and the SFE experiments. The physical and chemical properties of the bitumens are presented in Table 4.1. 159 Table 4.1 Physical and Chemical Properties of Uinta Basin Bitumens Whiterocks Properties Bitumen 0.980 Specific Gravity Asphalt Ridge PR Spring Sunnyside Bitumen Bitumen Bitumen 0.985 1.005 1.015 12 .1 9.3 7.9 13.9 14.17 15.0 (288K/288 K) 12.9 API Gravity, °A P I Conradson Carbon, 9.5 Residue, wt % 327 320 319 348 4,825 5,050 47,000 173,000 Pour Point, K Viscosity, cp @ 343 K a) Asphaltenes , wt % 2.9 6 .8 19.3 23.6 Saturates, wt % 35.7 39.2 33.4 2 0 .0 Aromatics, wt % 7.0 9.0 3.6 15.1 54.5 44.1 43.8 36.8 Resins, wt % Molecular Weight, g/gmol Elemental Analysis 426 653 670 593 b) C, wt % 87.0 86.9 87.0 8 6 .8 H, wt % 11.2 11.6 11.3 10 .8 N, wt % 1.4 1.3 1.1 S, wt % 0.4 1.7 0.4 0.4 H/C Atomic Ratio 1.56 1.60 1.56 0.7 1.49 Simulated Distillation 46.6 53.5 45.4 40.9 1.3 0.4 0 .6 4 7 7 -6 1 7 K 0.5 7.4 11.8 8 .2 7.8 6 1 7 -8 1 1 K 38.7 40.4 36.8 32.5 53.4 46.5 54.6 59.1 Volatility(<811 K), wt% <477 K >81 1 K Pentane Insolubles C,H,N,S analyses normalized to 100% 160 Physical Properties The specific gravities of the bitumens were measured using the procedure proposed by Syncrude Limited for semisolid hydrocarbons[179]. The W R bitumen had a specific gravity (288 K/288 K) of 0.98, compared to 0.985 for the AR bitumen, 1.005 for the PRS bitumen and 1.015 for the SS bitumen. The W R and SS bitumen had the lowest and highest specific gravities, respectively. The AR and PRS bitumen specific gravities fell in-between with the AR density lower than that of the PRS bitumen. The viscosities of the bitumens were measured at 343 K to provide a direct comparison. The viscosity of the W R bitumen was lowest at 4825 cP compared to the SS bitumen at 173,000 cP and the viscosities of the AR (5050 cP) and PRS bitumens (47,000 cP) fell in between. The viscosities for these bitumens were measured using a Brookfield Cup and Cone viscometer, Model DVT-II+, in the temperature range from 318 to 353 K. In general, the viscosities of the four bitumens decreased with increase in temperature and the rate of decrease in viscosities were different for the different bitumens. The relationship between viscosity and inverse temperature is presented in Figure 4.1. It is observed from Figure 4.1 that the decrease in the viscosity of the SS bitumen was greatest with increase in temperature, the PRS and W R bitumen have similar slopes and the AR bitumen had the lowest decrease for a given increase in temperature. The activation energies, Ead for viscous flow for the four bitumens are presented in Table 4.2. The pour points, a measure of the fluidity of the bitumens, of the AR and PRS bitumens were similar at around 320 K, the W R bitumen was 327 K and the 161 Figure 4.1 Relationship Between Temperature and Viscosity for the Bitumens 162 Tem perature, K 320 325 330 335 340 1000/T(K) 345 350 355 163 Table 4.2 The Eactfor Viscous Flow for Four Bitumens from Uinta Basin (Utah) Bitumen ko Eact Whiterocks Bitumen 5.2187x10‘13 -12.574 Asphalt Ridge Bitumen 2.8487x1 O’11 -11.281 PR Spring Bitumen 2.6708x1 O'12 -12.843 Sunnyside Bitumen 4.3948x1 O'02 -5.207 vis cos ity, c P - kQg -E T Temperature “T ” in Kelvin 164 SS bitumen was highest at 348 K. Chemical Properties The bitumen feedstocks were mixed with an excess of normal pentane (40 cc of pentane/gram of sample) to precipitate the asphaltenes fraction and to dissolve the maltene fraction. The procedure adopted by Bukka et al.[44,192] was used to fractionate the bitumens, and the extract phases and residual fractions obtained from SFE. The maltenes were subjected to adsorption chromatography using Fuller's earth as the adsorbent and were sequentially eluted using solvents of increasing polarity to isolate saturates and aromatics, resins I and resins II. These compound classes were separated based on their solubility using solvents such as pentane, tetrahydrofuran and methanol. The saturates and aromatics were further adsorbed on neutral alumina and eluted using the same sequence of solvents mentioned above to isolate saturates, aromatics I and aromatics II. The aromatics I and II fractions and the resins I and II fractions were combined and were reported as aromatics and resins, respectively. The fractionation results obtained for the four different bitumen feedstocks are presented in Figure 4.2. It is seen from Figure 4.2 that the asphaltene content of the W R bitumen was lowest at 2.9 wt%, and the asphaltene contents of AR (6 . 8 wt%), PRS (19.3 w%) and SS bitumens (23.6 wt%) were progressively greater. The saturates content of the SS bitumen was lowest at 20.0 wt% compared to the high value for the AR bitumen of 39.2 wt%. The PRS and W R bitumens saturates contents 165 Figure 4.2 Comparison of the Solubility Fractions of the Bitumens 60 □ Whiterocks Bitumen 50 M Asphalt Ridge Bitumen ■ PR Spring Bitumen □ Sunnyside Bitumen O) S S As pha Ite nes Satu rates Aromatics Resins 166 167 fell in the intermediate range at 33.4 and 35.7 wt%, respectively. The aromatics content of the SS bitumen was highest at 16 wt% and the aromatics content of the other three bitumens was less than 10 wt%. The resin content of the W R bitumen was highest at 54.5 wt%, the SS bitumen lowest at 36.8 wt%. The AR (44.1 wt%) and PRS bitumens (43.8 wt%) had intermediate resins contents with the AR resin content slightly higher than that of the PRS bitumen. It is also observed from the chemical compositions of the bitumens that as the asphaltene content of the bitumens increased, the specific gravity, viscosity at 343 K and Conradson carbon of the four bitumens increased. Thus, the asphaltene content appears to have affected these properties of the bitumens under investigation. The molecular weights of these bitumens were determined by Galbraith Laboratories, Inc., Knoxville, TN using a three-point vapor pressure osmometry method with toluene as the solvent. The molecular weight of the AR bitumen was lowest at 426 g/gmol, the W R and PRS bitumens had similar molecular weights around 660 g/gmol and the SS bitumen had an intermediate molecular weight of 593 g/gmol. The elemental analyses for carbon, hydrogen, nitrogen and sulfur were also determined by Galbraith Laboratories, Inc., using a Perkin-Elmer elemental analyzer. The carbon, hydrogen, nitrogen and sulfur contents of the bitumen were normalized and are reported in Table 4.1. The AR bitumen was more saturated than the other bitumens with a hydrogen/carbon (H/C) atomic ratio of 1.6. The W R and PRS bitumens had H/C ratios of 1.56. The SS bitumen was 168 the least saturated with a H/C atomic ratio of 1.48. The H/C atomic ratio of these bitumens correlated with their saturates contents: the higher the saturates content, the higher the H/C ratio of the bitumen. The AR bitumen had the highest saturates content and the highest H/C atomic ratio. The nitrogen contents of the four bitumens were higher than their sulfur contents which is typical for fresh water origin hydrocarbon resources. The quality ranking for the four bitumens as indicated by their physical and chemical properties is as follows: Whiterocks < Asphalt Ridge < PR Spring < Sunnyside Simulated Distillation The boiling point or carbon number distributions of hydrocarbon mixtures is required information for chemical process design. The boiling point distributions of crude oils and heavy oils can be determined by conventional distillation techniques using the American Standard for Testing Material (ASTM) D2892-90[193] and D5236-92[194] procedures, respectively. These ASTM techniques propose the use of vacuum distillation operated at a 5:1 reflux ratio with a distillation column. ASTM methods such as ASTM D2887[195] and D5307[196] were developed using a gas chromatography based simulated distillation (SIM DIS) technique to reduce the time required for boiling point range analysis. The ASTM D2887 simulated distillation procedure provided the means to determine boiling point distributions of oils which were totally elutable during a chromatographic run and was capable of estimating boiling point distributions 169 through 811 K (538°C ). ASTM D5307 was subsequently developed to account for the uneluted portion of the oil, based on the mathematical procedure originally proposed by Worman and Green[185], Neer and Deo[198] established the mathematical equivalence between this procedure and the more intuitive lever arm rule to quantify the uneluted fraction. These two ASTM procedures suggest use of packed bed columns to determine the boiling point distributions of oils up to 811 K (538°C). It should be possible to elute heavier fractions of the oils and obtain boiling point (or equivalent carbon number) distributions at temperature above 811 K (538°C ) with capillary columns using high phase ratios (-5 0 0 ) or phase ratios equivalent to packed columns. The other significant advantage in using the capillary column temperatures. is negligible bleeding at high The development of this technique would be particularly useful for analysis of heavy oils and bitumens which typically contain greater than 50 wt% material boiling above 811 K (538°C). A technique was developed using short, high-phase-ratio capillary columns for the characterization of ultra heavy oils and bitumens in the course of this study. The calculation methodologies suggested in the ASTM D2887 and D5307 methods have been extended to higher boiling point components, and thus, to higher carbon numbers. The modified technique was used to analyze the bitumens from the W R, AR, PRS and SS oil sands deposits of the Uinta Basin (Utah), as well as the saturates, aromatics and resins solubility fractions of the four bitumens and the extract and residual fractions obtained from SFE of the bitumens. Furthermore, a comparison has been made between the boiling point distributions obtained from the ASTM and the proposed technique for the four bitumens, the four extract phases and the four residual fractions generated during SFE of the bitumens. A Hewlett Packard Model 5890 Series II gas chromatograph was used to analyze all the samples. A detailed description of the experimental setup, temperature program, carrier gas flow rate and types of columns used for the analyses was presented in Chapter 3. Injections were performed using the Model A 7673 automatic sampler from Hewlett Packard. The signals from the detector were sent to an IBM-PC through a Hewlett Packard 3396 Series II integrator using a Hewlett Packard file server program. The signals were integrated using a slicing program at the integrator and were stored as report files for further computations. A Microsoft Windows based SIM DIS program was developed to read the sliced and calibration data and to obtain the boiling point and carbon distributions for totally and partially eluted samples. The program was coded and compiled in Visual Basic. The complete listing of the code is presented in Appendix E. The bitumen samples were highly viscous at room temperature and hence were diluted with carbon disulfide (1:1) to facilitate injection. Polywax 655, a calibration mixture from Supelco, Inc., was used to generate a relationship between carbon number and retention time. Polywax 655 is a blend of polyethylene oligomers with a carbon number range of C 10 to C 110 in two carbon number increments. The peaks for carbon numbers C20, C^, C40, C50, C 60,C70, C 80 171 and were observed at approximately 13.5, 19.0, 23.5, 27.5, 29.5, 31.0, 33.0 and 35.5 min, respectively (Figure 3.3). The retention times for earlier peaks (< carbon number 20) were obtained using different calibration mixtures obtained from Supelco, Inc. The relationship between carbon number and retention time was obtained from the calibration mixtures. The boiling points of the C 10 to Cgo hydrocarbons were obtained from TRC Thermodynamic Tables[199], The chromatographic analyses of bitumen and bitumen derived products were performed in the following sequence: 1. Blank run 2. Sample run 3. Sample + internal standard run. Discussion The procedure outlined below was used to obtain boiling point and/or carbon number distributions to the upper limit of the calibration curve. A relationship between carbon number and retention time was obtained up to 973 K (700°C). The uneluted portion of the sample, fraction boiling above 973 K (700°C ), was calculated using chromatograms of the experimental sample and another sample which contained a known weight of the internal standard usually in a 10-to-1 ratio of internal standard-to-bitumen. The internal standard was obtained from Hewlett Packard and contained 14 C 1b ,c, 1b and C 1.T # normal alkanes. The total area for the totally elutable chromatogram was calculated as the difference in the area under the sample and blank chromatograms. The baseline signal was subtracted from both the sample and sample plus the internal standard chromatograms before the uneluted fractions were calculated according to the following procedure[196]: The total area of the chromatogram was calculated from equation [4.1]: Total Area, T = [{/4ZS x /f} - 5 /5 ] x (1 - ISFRAC) (ISFRAC) [4.1] where T is the total area under the chromatogram; R = [B-BIS] [A - A IS ] [4.2] ISFRAC is the weight of the internal standard/(weight of sample + weight of internal standard) and defined as W in equation 5 of ASTM D5307[196]; A is the area of the internal standard plus sample chromatogram up to 973 K (700°C); B is the area of the sample chromatogram up to 973 K (700°C); AIS is the area of the internal standard plus sample segment of the internal standard chromatogram; and; BIS is the area of the corresponding internal standard segment of the sample chromatogram. 173 The weight fractions of different boiling point or carbon number segments were calculated using the total chromatogram area according to equation [4.3]: = \B- [4-3] where • W n is the weight fraction of the sample eluted between carbon numbers n and n-1; and; • Bn and B„.i are the areas of the sample chromatogram up to carbon number n and n-1, respectively. The sample and sample plus internal standard chromatograms of the W R bitumen are presented in Figure 4.3 along with the blank chromatogram. The blank chromatogram shown in Figure 4.3 indicated that at high temperatures, negligible column bleed occurred. It should be noted that C*,, with a boiling point of 973 K (700°C), elutes at around 35.5 minutes (refer to Figure 3.3). Thus, the analysis procedure essentially divides the sample chromatograms into boiling point fractions up to 35.5 minutes plus a noneluted fraction that boils above 973 K (700°C ). The ASTM D5307 method proposed the use of packed columns with which the boiling point distribution could be obtained up to 811 K (538°C). The differential and cumulative boiling point distributions of the W R bitumen are shown in Figure 4.4. A comparison of the amounts of the 811 K (538°C ) plus fractions of the bitumens as determined by the conventional ASTM D5307 and the extended method of analyses and for the 973 K (700°C ) plus 174 Figure 4.3 Chromatograms for Whiterocks Bitumen 0 5 10 15 20 25 Time, minutes 30 35 40 45 176 Figure 4.4 Boiling Point Distribution for Whiterocks Bitumen Differential Weight Fraction Cumulative Weight Fraction 300 400 500 600 700 800 900 1000 Boiling Point, K -vl -vl 178 fractions is presented in Table 4.3. The reported, 811 K plus weight fractions as determined by the ASTM D5307 (area A and B in equation 4.2 estimated up to 811 K) and the extended method D5307 (area A and B in equation 4.2 estimated up to 973 K) were virtually for the same the four bitumens and the SFE residues. The W R bitumen has an 811 K (538°C ) plus fraction of 53 wt% and a 973 K (700°C ) plus fraction of 20 wt%. It was possible to characterize (within the limits of simulated distillation analysis) 80 wt% of the W R bitumen using the modified procedure; that is, 33 wt% more than would have been possible using ASTM D5307. Similarly, 89 wt% of the AR bitumen (34 wt% more than ASTM D5307), 79 wt% of the PRS bitumen (33 wt% more than ASTM D5307) and 73 wt% of the SS bitumen (32 wt% more than ASTM D 5307) were characterized by the method described here. In general, the extended method made it possible to characterize approximately one-third more of these heavy oil samples than would have been possible by conventional ASTM simulated distillation analyses. The high temperature simulated distillation technique described here is limited to boiling point distributions up to 973 K (700°C ) by the upper oven temperature limit of 653 K (380°C ) above which the thermal degradation of the injected samples occurs. The simulated distillation analyses in terms of boiling range fractions for the four bitumens are presented in Table 4.4. The degradation of the uneluted material takes place inside the capillary columns due to thermal and catalytic action by the metals present in the high molecular weight compounds. This technique could be extended for thermally stable samples 179 Table 4.3 Comparison of the Extended Method Results Weight Fraction Samples 811 Kplusa> (538°C plus) 811 K plusb) (538°C plus) 973 K plus (700°C plus) Whiterocks Bitumen 0.52 0.53 0.20 Asphalt Ridge Bitumen 0.46 0.45 0.11 PR Spring Bitumen 0.53 0.54 0.21 Sunnyside Bitumen 0.61 0.59 0.27 Whiterocks SFE Residue 0.90 0.88 0.58 Asphalt Ridge SFE Residue 0.85 0.87 0.61 PR Spring SFE Residue 0.88 0.86 0.59 Sunnyside SFE Residue 0.86 0.87 0.61 Whiterocks SFE Extract - 0.08 0.0 Asphalt Ridge SFE Extract - 0.16 0.0 PR Spring SFE Extract - 0.12 0.0 Sunnyside SFE Extract - 0.18 0.0 a) ASTM D5307 Method (area A and B in equation 4.2 estimated up to 811 K) b) Modified Extended Method (area A and B in equation 4.2 estimated up to 973 K) 180 Table 4.4 Simulated Distillation Analyses for Bitumens Analyzed Properties Whiterocks Bitumen Asphalt Ridge PR Spring Bitumen Bitumen Sunnyside Bitumen Modified SIMDIS Procedure Volatility (<811 K), wt% 46.6 53.5 45.4 40.9 0.5 1.3 0.4 0.6 7.4 11.8 8.2 7.8 617 K -811 K 38.7 40.4 36.8 325 811 K- 973 K 33.0 34.0 33.0 32.0 20.0 11.0 21.0 27.0 Distillation Cuts. wt% <477 K 477 K - 617 K > 973 K 181 beyond boiling point 973 K (700°C ) by oven programming beyond 653 K (380°C ) and by using a higher molecular weight polywax standard and pressure programming of the carrier gas. If an oil contains a small amount of material boiling above 653 K (380°C ), it should be possible to characterize 100% of the sample using the method proposed here. This is illustrated using the extract phases obtained by the supercritical fluid extraction of the four bitumens with propane. The chromatogram for the extract phase from the W R bitumen is shown in Figure 4.5. Analyses with an internal standard (analogous to those described in the previous paragraph) for these extract samples revealed that these extracts did not contain hydrocarbons heavier than C9o(TnbP=973 K). The 811 K (538°C ) plus fractions of the W R, AR, PRS and SS extracts were 0.08, 0.16, 0.12 and 0.18, respectively. However, all four extract phases eluted totally when the upper limit boiling point was 973 K (700°C). Thus, the procedure proposed here permitted complete characterization of samples which consist of relatively small fractions boiling above 811 K (538°C ). The boiling point distribution of the extract phase from the W R bitumen is shown in Figure 4.6. The method can also be applied to extremely heavy petroleum residua. Residual fractions of the four bitumens recovered after supercritical fluid extraction with propane were used to illustrate this point. Approximately 90 wt% of each of these residual fractions boiled above 811 K (538°C ) as indicated in Table 4.3. The extended SIMDIS technique characterized 42 wt% of the W R , 39 182 Figure 4.5 Chromatograms for the Whiterocks Bitumen Extract 1000000 653 K / a .75 min / 1 2 K/min 313/ 4.5 min Column : Petrocol EX2887 from Supelco, Inc. Carrier Gas : He at 20 cc/min co c 0) £ 500000 To c o> CO Blank “ j— 5 i— t — i— i— i— 10 j— i— r— i— |— 15 i— i— i— r p n — i— 1 j i 20 25 Time, minutes i i | 30 35 40 45 184 Figure 4.6 Boiling Point Distribution for the Whiterocks Bitumen Extract Cumulative Weight Fraction Boiling Point, K cn 00 186 wt% of the AR, 41% of the PRS and 39% of the SS residual fractions (Table 4.3). Thus, the technique permitted the characterization of an additional 30 wt% of the residual fractions. The chromatograms of the W R bitumen residual fraction and its boiling point distribution are presented in Figures 4.7 and 4.8, respectively. The carbon number distributions for the four bitumens along with the saturates, aromatics and resins solubility fractions were determined using the simulated distillation technique outlined above. The carbon number distributions for the asphaltene fractions were not obtained because they were solid at room temperature and it was expected that no more than 20 wt% would elute. Thus 80 wt% of the asphaltenes would be uneluted and remain on the column and decrease the life of the column. The carbon number distributions obtained for the four bitumens and the saturates, aromatics and resins solubility fractions are presented in Figure 4.9 through 4.12. Findings Boiling point distributions up to 973 K (700°C) and an estimate of the 973 K (700°C) plus fraction for ultra-heavy oils can be obtained using capillary columns with high phase ratios. Applicability of this technique was demonstrated using ultra-heavy oils (bitumens) from the Uinta Basin (Utah), which had volatilities, 811 K (538°C) minus fraction, of 55 wt% or less. The method allowed for the characterization of 30-35 wt% more of the bitumen and residual fractions obtained by SFE than would have been possible using the ASTM D5307 187 Figure 4.7 Chromatograms for the Whiterocks Bitumen Residual Fraction 300000 250000 313/ Column 200000 653 K ^ 7 5 min 12 K/mln 4.S min : Petrocol EX2887 from Supelco, IInc. Carrier Gas : He at 20 cc/min = 150000 re c O) c/j 100000 50000 Sample + Internal Standard, Sample 0 TI_ 0 5 10 15 20 25 30 Time Minutes 35 40 45 189 Figure 4.8 Boiling Point Distribution for the Whiterocks Bitumen Residual Fraction Weight Fraction Differential Weight Fraction Cumulative Boiling Point, K CO o 191 Figure 4.9 Carbon Number Distributions for the Whiterocks Bitumen and the Saturates, Aromatics and Resins Solubility Fractions 192 1 • Saturates 0.9H o Aromatics Cumulative Weight Fraction 0.8 ♦ Resins K x Whiterocks Bitumen 0 .7 ­ 0.6 o 0.5 0 .4 ­ • x o 0.3 • X * X o a 0.2 • • x 0.1 • X t il* ! ? 0 10 20 * ° O X O x o o♦ o O ♦* ♦ ° o * * ♦ o ♦ o o♦ ♦ ♦ ♦ ♦ ♦ !! 30 40 50 Carbon Number 60 70 80 90 193 Figure 4.10 Carbon Number Distributions for the Asphalt Ridge Bitumen and the Saturates, Aromatics and Resins Solubility Fractions 194 1. 2 - • Saturates o Aromatics 1- ♦ Resins * Asphalt Ridge Bitumen Cumulative Weight Fraction x ■ 0. 8 - 0 .6 - 0.4 a°° * 0. 2 x x • • * 0 10 Ou r.O o°° 20 o° *♦* 30 n O 0 °♦ » * ♦ 40 50 Carbon Number 60 70 80 90 195 Figure 4.11 Carbon Number Distributions for the PR Spring Bitumen and the Saturates, Aromatics and Resins Solubility Fractions 196 . 0.9- • ........................ Saturates o Aromatics 0.8 - C 0.7- ♦ Resins + PR Spring Bitumen o U © 0.6 cn © 0.5 5 + a> > S 04 re 3 E 3 03 o + + o + o * + o •+ o * * • + o♦ ♦ + o o o ♦ o o + + ♦ ♦ ♦ ♦♦ ♦ ♦ + 0.2 0.1 9 m ♦ - 10 • - --j— ' 20 i i 30 i-------------------- ;----------1 40 t t ) i t ...................................— 50 Carbon Number 60 - I- 70 80 90 197 Figure 4.12 Carbon Number Distributions for the Sunnyside Bitumen and the Saturates, Aromatics and Resins Solubility Fractions 198 • 0.9- Saturates o Aromatics 0. 8 - ♦ Resins Cumulative Weight Fraction + Sunnyside Bitumen 0.7- + +• 0. 6 - . + + + « + ++ o° o o +0 ++o o ° o 0.5 0.4 0.3 + 0.2 • 0.1 + # + + + + o o o 10 20 o ♦ ♦ ♦ o° o ♦ O° . ♦ ♦ • ;« .a » tI * 0 o 30 ♦ ♦ T 40 50 Carbon Number 60 70 80 90 199 method. Characterization of the extract phases with volatilities 90 wt% was also demonstrated. Preliminary Process Experiments SFE experiments were carried out using n-hexadecane and C 0 2 to demonstrate the attainment of thermodynamic equilibrium during SFE experiments. The SFE experiments were carried out at 10.4 MPa (Pr=1.41) and 311 K (Tr=1.02) at four different solvent flow rates to study the effect of flow rate on the attainment of thermodynamic equilibrium. One hundred cubic centimeters (76 g) of hexadecane were initially charged to the extractor in each of the four experiments. The extracted phase samples were collected in the atmospheric separator for every 20 I @ STP of C 0 2 solvent vented through the system. The extraction was terminated after five such samples were collected. A negative volume change upon mixing was observed while charging the extractor with supercritical C 0 2. The extraction experiments were performed at four different solvent flowrates (3.8, 5, 6 and 10 cm3/min) at 10.4 MPa and 311 K using a positive displacement pump (D.B Robinson). It had been established previously[40] that flow rates below 5 cm3/min were required to establish thermodynamic equilibrium for the hexadecane-C02 system[40]. The back pressure valve (spring mounted), originally located upstream of the extractor, was moved downstream of the extractor between the densitometer and separator. This modification facilitated pressure control of the extractor and stabilized the extractor pressure. The 200 results obtained for the SFE of hexadecane and C 0 2 are presented in Table 4.5 and Figure 4.13. It is observed from Figure 4.13 that the extraction yields obtained for C 0 2 flow rates (5, 6 and 10 cm3/min) were different than the yields obtained at conditions known to produce thermodynamic equilibrium. The extraction yields obtained for a C 0 2 flowrate of 3.8 cm3/min were constant for every 20 liters (@ STP) of solvent vented through the system. The hexadecane yield increased for the initial two extraction window and decreased in the subsequent extraction windows for all other solvent flowrates. According to the phase rule, for a twocomponent, two-phase system, the extract phase composition should be invariant with the overall composition of the system meaning that the extract phase composition or the extraction yield for every 20 liters @ STP of solvent flowing through the system should remain constant. The extraction yields were constant (refer Figure 4.13) for a flow rate 3.8 cm3/min for six extraction window and for the other three flowrates the extraction yields varied. This confirms that the system operated under thermodynamic equilibrium for a C 0 2 flowrate of 3.8 cm3/min and that the system is capable of producing reliable thermodynamic data. Supercritical Fluid Extraction of OU Sands Bitumens SFE experiments were conducted using the AR and SS bitumens as a part of this study using commercial propane as the solvent. composition of the propane was presented in Table 3.4. The detailed SFE of the W R and 201 Table 4.5 Results of the Hexadecane-Carbon Dioxide Experiments Performed at 10.4 MPa (Pr=1.41) and 311 K (T r=1.02) Volume of C 0 2 Vented Wt% of Hexadecane Extracted 3.8 cm3/mina) 5 cm3/minb) 6 cm3/minb) 10 cm3/mina) 0 0 0 0 0 20 2.1 2.95 1.08 0.92 40 4.3 4.4 2.23 2.48 60 6.6 6.0 3.4 3.9 80 8.6 7.3 4.4 4.8 10.9 9.3 - 5.8 10.5 - 6.4 100 120 - 76.16 grams of Hexadecane charged initially into the extractor 83.49 grams of Hexadecane charged initially into the extractor 202 Figure 4.13 Effect of Solvent Flowrate on the Attainment of Thermodynamic Equilibrium for H exadecane-C02 system at 10.4 MPa (Pr= 1 .41) and 311 K (Tr=1.02) Cumulative wt% of Hexadecane Extracted 203 Cumulative Volume of Carbon Dioxide Vented, (Liters, @STP) 204 PRS bitumens were conducted by Hwang[39] and Subramanian[40], respectively. The SFE experiments for all four bitumens were carried out at five different operating conditions, a combination of three pressures (5.6, 10.4, and 17.3 MPa) and three temperatures (339, 380 and 422 K) with four conditions above the critical pressure and temperature of the solvent and one condition above the critical pressure but below the critical temperature of the propane. The typical operating conditions used for the SFE experiments are presented in Figure 4.14. The experiments were carried out at five similar conditions so that the influence of the nature of the bitumen on SFE yields and on the quality of the extract phases could be observed. Experiments were conducted to measure the density of the commercial propane using the densitometer at the operating conditions used in this study. The measured densities of the propane are listed in Table 4.6. The propane density increased with increasing in pressure at constant temperature and decreased with increasing in temperature at constant pressure. Starling[200] developed an equation of state which was used to estimate the density of propane and other hydrocarbon gases. The propane densities for pressures and temperatures ranging from 1.97 MPa to 10.9 MPa and 355 to 477K; respectively, are presented in Figure 4.15. These curves were obtained by solving the equation of state developed by Starling[200], The propane densities predicted were less than the measured densities which is typical for cubic equations of state. The cubic equation of state predicts liquid densities lower than the experimentally measured one[178] and the deviation from the measured values 205 Figure 4,14 Operating Conditions for SFE of Bitumens Using Propane as Solvent 206 440 : | ▼ 420 * 400 £ ^ ▼ = 380 — re g.360 . + ▼ ▼ - I I ® 340 ▼ 320 - 300 i I I 2 4 | Pc ' I ‘ ' 1' 1i 6 Propane Pc=4.25 MPa Tc=369.7 K i 1■■i i 8 10 12 14 16 18 Pressure, MPa 207 Table 4.6 Measured Densities of Commercial Propane Density (g/cc) Temperature, K Reduced Temperature Pressure Reduced (MPa) Pressure 339 0.92 380 1.03 5.6 1.3 10.4 2.4 0.566 0.553 17.3 4.1 - 0.569 - 0.533 422 1.14 - 0.545 - 208 Figure 4.15 Propane Density at Various Temperatures and Pressures[188] Density, g/cc 0.45 to 6 Pressure, MPa 8 10 12 209 210 differs for different equations of state[178, 186], hence the propane density was measured. The SFE experiments were conducted with 50 g of bitumen (AR and SS) initially in the extractor. A known quantity of propane was charged to the extractor until the system was brought to the desired operating conditions. The initial solvent charging was performed using a positive displacement pump. The extraction procedure was explained in detail in Chapter 3. Liquid samples were collected from the separator for every 25 liters of propane vented through the system. Six such samples were collected before the extraction was terminated. The extraction was terminated after window #6 because the extraction yield tapered down considerably and terminating extraction after a fixed number of windows (six) permitted meaningful comparisons of the compositions of the residual fractions. The extract phase samples were labeled as extract window #1 through #6. The hydrocarbon remaining in the extractor was collected as the residual fraction for further analysis. Supercritical Fluid Extraction of Asphalt Ridge Bitumen Hwang[39] and Subramanian[40] conducted SFE experiments and studied the SFE behavior of the bitumens from the W R and PRS deposits. This section will examine SFE extraction of the bitumen from the AR deposit whose physical and chemical properties are intermediate to those of the W R and PRS bitumens. The effect of process variables such as pressure, temperature and solvent density on the extraction yields of the AR bitumen has been determined. 211 Pressure Effect The SFE experiments conducted to determine the influence of pressure on the SFE of AR bitumen were carried out at a constant temperature of 380 K (Tr=1.03) and three different pressures of 5.6 MPa (Pr=1.2), 10.4 MPa (Pr=2.3) and 17.3 MPa (Pr=4.1). The maximum cumulative extraction yield, 31.4 wt%, was achieved at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03). The extraction yield decreased with decreasing in pressure at constant temperature 380 K (Tr=1.03) with yields of 24.5 wt% and 13.4 wt% achieved at 10.4 MPa (Pr=4.1) and 5.6 MPa (Pr=4.1), respectively. The amount extracted was quite sensitive to the solvent density: for every 1% change in the solvent density, there was 3% change in the cumulative extraction yield (6 windows). The effect of pressure on the extraction yield of AR bitumen is presented in Figure 4.16. The decrease in the overall extraction yield with decrease in pressure at constant temperature was attributed to the decrease in the solvent density with decrease in pressure (Table 4.6). The extracted phase density was measured for all the SFE experiments for the AR bitumen-propane system using the data acquisition system incorporated into the SFE system. The measured extracted densities for the AR bitumen-propane system at all five operating conditions are plotted versus extraction time in Figure 4.17. The extraction yield increased for the first three windows and decreased in the subsequent extraction windows (Figure 4.17). This trend is attributable to the extraction of lighter hydrocarbons initially and the change in the overall 212 Figure 4.16 Effect of Pressure on SFE Yields with the Asphalt Ridge Bitumen Cumulative Amount Extracted, wt% 213 Cumulative Volume of Propane Vented (Liters, @STP) 214 Figure 4.17 Measured Extract Phase Density During SFE of the Asphalt Ride Bitumen with Propane as Solvent 0.80 10 4 MPa ' 339 K 10 4 MPa * 380 K 0.75 o o o> >> to 0.70 17 3 MPa * 380 K c <u Q a> w ro 0.65 _c CL T3 (U O (D i— 0.60 X LU i 0.55 10 20 30 40 50 60 70 80 Time, Minutes 90 100 110 120 130 140 150 N> On 216 composition of the system with time. Propane has an affinity for the extraction of lighter hydrocarbons relative to heavier hydrocarbons[38], The lighter hydrocarbons were extracted initially and as the extraction proceeded, the system was depleted in lighter hydrocarbons (the SFE system is batch and continuous with respect to bitumen and solvent, respectively) and the remaining bitumen components in the extractor became heavier. As the hydrocarbons were extracted, the overall composition of the system changed continuously and became leaner in bitumen and richer in solvent. The lower affinity of propane for heavier hydrocarbons resulted in a reduction in the transfer of bitumen components from the oleic phase to the extract phase during the later part of the extraction. Furthermore, the lighter hydrocarbons acted as cosolubilizing agents for the heavier species; thus their depletion during the initial extraction periods produced a more refractory residual oleic phase which was less soluble in supercritical propane. Consequently the extraction yields decreased. It is observed from the extract phase density plots (Figure 4.17) that the SFE experiments started with extract phase densities of 0.71, 0.66 and 0.62 g/cm3 at pressures of 17.3 (Pr=4.1), 10.4 (Pr=2.3), and 5.6 MPa (Pr=1.2); respectively, at a constant temperature of 380 K (Tr=1.03). The pure solvent densities measured at these conditions were 0.569, 0.553 and 0.533 g/cm3, respectively. 217 Temperature Effect The extraction experiments were conducted to examine the influence of temperature on SFE of the AR bitumen at a constant pressure of 10.4 MPa (Pr=2.3) and at three different temperatures: 339 K (Tr=0.92), 380 K (Tr=1.03) and 422 K (Tr=1.14). The maximum cumulative extraction yield, 31.3 wt%, was obtained at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92). The cumulative extraction yield decreased with increase in temperature at constant pressure, that is, yields of 24.5 and 18.2 wt%, were achieved at 380 K (Tr=1.03) and 422 K (Tr=1.14), respectively (Figure 4.18). The decrease in extraction yield with increase in temperature at constant pressure was attributed to the decrease in solvent density with increase in temperature (Table 4.6) It is also observed from the extract density plots (Figure 4.17) that the extract phase density was highest at the lowest temperature 339 K, (Tr=0.92) and decreased with increase in temperature at constant pressure. The extraction yield increased for the first four windows and decreased in the subsequent extraction windows (Figure 4.18). As explained in the previous section, this effect was related to the extraction of lighter hydrocarbons initially and the change in the overall composition of the system with time. The initial extract phase densities were 0.680, 0.66 and 0.625 g/cm3 at 339 (Tr=0.92), 380 (Tf=1.03), and 422 K (Tr=1.14), respectively at a constant pressure of 10.4 MPa (Pr=2.3) (Figure 4.17) whereas the pure solvent densities were 0.566, 0.553 and 0.545 g/cm3; respectively. Thus, both the pure solvent density and the extract phase densities decreased with increase in temperature 218 Figure 4.18 Effect of Temperature on SFE Yields with the Asphalt Ridge Bitumen Cumulative Amount Extracted, w t% 219 Cumulative Volume of Propane Vented, (Liters, @STP) 220 resulting in a decrease in the extraction yields. Solvent Density Effect The solvent densities measured at the five operating conditions are plotted (Figure 4.19) against the cumulative extraction yields obtained for SFE with the AR bitumen. It is observed from the plot that the cumulative extraction yields with AR bitumen increased as the pure solvent density increased. However, the maximum extraction yield obtainable using propane as the supercritical fluid would be equal to or less than the deasphalted oil obtained using liquid propane[201]. The pure solvent densities measured at the five operating conditions were 0.533, 0.566, 0.553, 0.545 and 0.569 g/cm3; respectively, at 5.6 MPa and 380 K, 10.4 MPa and 339 K, 10.4 MPa and 380 K, 10.4 MPa and 422 K and 17.3 MPa and 380 K. While performing SFE experiments, the extract phase density was measured on a continuous basis using the data acquisition system. The initial extract phase densities were 0.62, 0.68, 0.66, 0.625 and 0.71 g/cm3; respectively, at 5.6 MPa and 380 K, 10.4 MPa and 339 K, 10.4 MPa and 380 K, 10.4 MPa and 422 K and 17.3 MPa and 380 K. These values indicate that the higher the pure solvent density, the higher the initial extract phase density at the same temperature and pressure. The propane densities were varied by adjusting the pressure and temperature and the extraction capacity of the solvent increased with increased density. The propane densities at 10.4 MPa and 339 K and 17.3 MPa and 380 Kwere similar: 0.566 and 0.569 g/cm3; respectively. The 221 Figure 4.19 Effect of Solvent Density on SFE Yields with the Asphalt Ridge Bitumen Cumulative Extraction Yield, wt% o c n — * o m — * o ro c n ro o c n go Propane Density, g/cc 222 223 cumulative extraction yields for the AR bitumen were likewise nearly the same: 31.3 and 31.4 wt%; respectively. These results indicate that the by maintaining the same extraction density through combinations of pressure and temperature, similar extraction yields could be obtained. Another explanation could be provided for the increase in extraction yields with increase in solvent density. Johnston[201]indicated that the enhancement factor (E) (Chapter 2) for hexamethylbenzene, 2-naphthnol, phthalic anhydride, anthracene and acridine increased exponentially with increase in density of carbon dioxide. The enhancement factor, a measure of the solubility of a solute in the vapor phase is directly governed by vapor pressures. This observation could be applied to SFE of bitumen using propane as solvent and the dependence of extraction yield on solvent density. The propane solvent density can be increased either by increasing the pressure at constant temperature or by decreasing the temperature at constant pressure. The maximum propane solvent density obtainable under supercritical conditions is always less than the propane liquid density and hence the extraction capability will be less than that of liquid propane. Thus, it may be advantageous to operate the extraction in the liquid phase region of the solvent where the temperature and pressure will be less compared to the supercritical conditions and to separate the solvent from the extract phase at supercritical conditions. The main advantages for operating under supercritical conditions are given below: • Extraction provides high selectivity • Selectivity towards extraction of a single compound or group of compounds could be varied by decreasing or increasing the solvent density by adjustment of the operating pressure and temperature • The supercritical solvent-solute separation is achieved by increasing the extract phase temperature when the extracted oil is no longer soluble in the solvent or by simple depressurization. In the case of a liquid solvent, the thermal energy must be added to the solvent to vaporize it from the solventextract phase mixture. Hence supercritical solvent processes are energy efficient. In case of commercial SFE process[111] to upgrade heavy oil, the energy savings are of the order of 30 to 50 % relative to conventional propane deasphalters which utilizes liquid propane. Carbon Number Distribution for AR Bitumen Extract Phases The extract samples obtained from SFE of the AR bitumen were subjected to simulated distillation analyses to determine the boiling point or carbon number distributions. The results obtained at a pressure of 17.3 MPa (Pf=4.1) and a temperature of 380 K (Tr=1.03) are presented in Figure 4.20. It is observed from the boiling point distributions of the extract phases that as the extraction proceeded heavier and heavier components were extracted. This was due to the semi-continuous (batch for solute and continuous for solvent) SFE system used for the extraction experiments where the lighter hydrocarbons were extracted initially. As the extraction proceeded, the extractor was depleted of lighter 225 Figure 4.20 Carbon Number Distributions for the Asphalt Ridge Bitumen, Extracts and Residual Fractions Obtained from SFE at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03) 226 1 0.9 0.8 Cumulative Weight Fraction 0.7 o 1st Window ♦ 3rd Window • 6th Window * Residue □ Bitumen 0.6 0.5 0.4 0.3 0.2 a + . • 0.1 * .0 -.* ..* » * < v ..•* * 0 0 10 20 30 40 50 60 Carbon Number 70 80 90 227 hydrocarbons and hence the heavier hydrocarbons were extracted in the subsequent windows. This was also indicated by the simulated distillation curves shifting towards higher carbon numbers as the extraction window number increased (Figure 4.20). upgraded liquids Furthermore, the extract phases were significantly as indicated by the volatilities, approximately 70 wt%, compared to the original bitumen which was 53.5 wt% volatile. The residual fraction was only 20 wt% volatile. The volatiles are defined as the fraction boiling below 811 K. The effects of temperature and pressure on compositional variation of the extract phases are presented in Figure 4.21. As the system pressure increased at constant temperature heavier extract fractions were obtained in the 2nd extraction window. At constant pressure, as the system temperature is increased, lighter extract liquids were obtained. It is observed from the experimental results that at a constant temperature of 380 K (Tr=1.03), the extraction yield increased from 13.4 wt% (@5.6 MPa) to 31.4 wt% (@17.3 MPa) as the extraction pressure increased. At the same time the volatility (fractions boiling below 811 K) of the 2nd extraction window extract decreased from 85.2 wt% (@ 5.6 MPa) to 65.1 wt% (@ 17.3 MPa). The SFE system was operated on semicontinuous basis (bitumen and propane are batch and continuous respectively) and hence an increase in extraction yield with pressure means decrease in overall volatility (quality) of the extract phase as the solvent extracts heavier hydrocarbons. As the system temperature 228 Figure 4.21 Effect of Pressure and Temperature on the Carbon Number Distributions of the Second Extraction Window Obtained During SFE of the Asphalt Ridge Bitumen 229 Effect of Pressure 1 2nd Extraction Window 380 K(Tr=1.03) .+ 0.9 | 0.8 5 0.7 I 9 os | 0.5 O) <♦> l 9 ■" 9 . S 0.4 1 0.3 O <A c r« I 0.2 o 0 .H ♦ S.6 MPa(Pr=1.2) o 10.4 MPa(Pr=2.3) ■ 17.3 MPa(Pr=3.8) I | ' l—! !"[" !" !' ! ! J \—! ! I | 10 20 30 40 50 60 Carbon Number 70 80 90 Effect of Tem perature 1 0.9 o ro 0.8 0.7 0.6 o> Qi 5 cu > re 3 E 3 o ■ * g * 2nd Extraction Window 10.4 MPa (Pr=2.3) 0.5 0.4 A 0.3 0.2 ■» 0.1 • 339 K(Tr=0.92) ■ 380 K(Tr=1.03) O 0 — 0 10 20 c i ■ I- "j 422 K(Tr=1.14) i i ; i | I 30 40 50 60 Carbon Number ' i i j ! 70 i i i j : 80 90 230 increased at constant pressure (10.4 MPa), the cumulative extraction yield decreased from 31.3 wt% (339 K) to 18.2 wt% (422 K). The volatility (quality) of the 2 nd window extract phase increased from 73.1 wt% (339 K) to 83.9 wt% (422 K). Thus, it could be stated that at constant temperature, as the extraction pressure increased extraction yields increased and produced heavier and less volatile hydrocarbon liquids. In contrast, at constant pressure as the extraction temperature increased extraction yields decreased but produced refined and lighter hydrocarbon liquids. Reproducibility Reproducibility experiments were conducted using the AR bitumen at 10.4 MPa (P,=2.3) and 339 K (Tr=0.92). The results obtained from two experiments are compared in Figure 4.22. It is observed from the experimental results that the difference between the extraction yields (for individual extraction windows) obtained for each of the six windows were within 5 % absolute margin. Hence, it was concluded that the experimental results are reproducible. Supercritical Fluid Extraction of Sunnvside Bitumen The physical and chemical properties such as specific gravity, Conradson carbon, pour point, viscosity at 343 K, pentane insolubles and 811 K plus (non volatile) fractions of the bitumen from the SS oil sands deposit are higher than those for the WR, AR and PRS bitumens. Hence, the SS bitumen was selected as representative of a lower quality bitumen for SFE studies. 231 Figure 4.22 Reproducibility for SFE with the Asphalt Ridge Bitumen at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92) 20 18 16 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92) Asphalt Ridge Bitumen 14 12 10 8 6 4 2 0 1 2 3 4 Window Number 5 6 233 Pressure Effect The experiments intended to determine the influence of pressure on the SFE of bitumen from the SS oil sands deposit were conducted at a constant temperature of 380 K (Tr=1.03) and three different pressures, 5.6 MPa (Pr=1.2), 10.4 MPa (Pr=2.3) and 17.3 MPa (Pr=4.1). The maximum extraction yield, 23.7 wt%, was achieved at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03). The extraction yields decreased with decrease in pressure at constant temperature. Extraction yields of 14.8 and 12.0 wt% were obtained at 10.4 MPa (Pr=2.3) and 5.6 MPa (Pr=1.2), respectively (Figure 4.23). The decrease in the extraction yield with decrease in pressure was attributed to decrease in solvent density with decrease in pressure at constant temperature. The extracted phase density was measured for all the SFE experiments for SS bitumen-propane system. The measured extract phase densities for the SS bitumen-propane system at all five operating conditions are plotted vs time in Figure 4.24. The extraction yield increased for the first three extraction windows and decreased in the subsequent extraction windows (Figure 4.23). As discussed previously, this trend was presumed to be related to the extraction of lighter hydrocarbons initially and to the change in the overall composition of the system with time. Propane has an affinity for the extraction of lighter hydrocarbons compared to heavier hydrocarbons[38]. The initial depletion of the lighter hydrocarbons led to a residual bitumen phase in the extractor which consisted of heavier components. As the extraction process proceeded, the overall 234 Figure 4.23 Effect of Pressure on SFE with the Sunnyside Bitumen Cumulative Amount Extracted, wt% 235 Cumulative Volume of Propane Vented, (Liters, @STP) 236 Figure 4.24 Measured Extract Phase Densities During SFE of the Sunnyside Bitumen with Propane as Solvent Extracted Phase Density, g/cc 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 Time, Minutes N3 CO -vl 238 composition in the extractor changed continuously to leaner in the bitumen component and richer in the solvent. The propane’s lesser affinity for heavier hydrocarbons and the enrichment of solvent in the extractor resulted in decreased extraction of bitumen components during the later part of the extraction. The maximum differential extraction yield was also inferred by the maximum extract phase density (Figure 4.24). The extract phase density plots (Figure 4.24) indicated that the initial extract phase densities were 0.604, 0.58 and 0.571 g/cm3 at pressures of 17.3 (Pr=4.1), 10.4 (Pr=2.3), and 5.6 MPa (Pr=1.2); respectively, at a constant temperature of 380 K (Tr=1.03). The comparable pure solvent densities were 0.569, 0.553 and 0.533 g/cm3, respectively. Thus, once again, it was concluded that higher extraction yields were driven by higher propane densities and in fact can be correlated by the solvent densities. Temperature Effect The influence of temperature on SFE extraction yields with the SS oil sands bitumen was determined at a constant pressure 10.4 MPa (Pr=2.3) and three different temperature of 339 K (Tr=0.92), 380 K (Tr=1.03) and 422 K (Tr=1.14). The effect of temperature on the cumulative extraction yield is presented in Figure 4.25. The maximum extraction yield, 22.4 wt%, was obtained at 339 K (Tr=1.03). The extraction yields decreased with increase in temperature at constant pressure, 14.8 and 11.2 wt% at 380 K and 422 K; respectively. The decrease in the extraction yield was attributed to the 239 Figure 4.25 Effect of Temperature on the SFE Yields with the Sunnyside Bitumen Cumulative Amount Extracted, wt% 240 Cumulative Volume of Propane Vented, (Liters, @STP) 241 decrease in the solvent density with increase in the temperature at constant pressure. The extraction yield increased for the first four windows and decreased in the subsequent extraction windows (Figure 4.25). As explained in the previous section, this effect was attributed to the initial extraction of lighter hydrocarbons and the transition in the overall system composition to heavier components and an enrichment in solvent with time. It is observed from the plots that the extractions were started with an extract phase densities of 0.602, 0.58 and 0.572 g/cm3 at temperatures 339 (Tr=0.92), 380 (Tr=1.03), and 422 K (Tr=1.14); respectively, at a constant pressure of 10.4 MPa (Pr=2.3) (Figure 4.24 ). The pure propane solvent densities at these conditions were 0.566, 0.553 and 0.545 g/cm3, respectively. Thus, it is apparent that as the pure solvent density decreased with increase in temperature, the experimentally measured extract phase density also decreased which led to the decline in the carrying capacity of the solvent. Solvent Density Effect The solvent densities are plotted versus the cumulative extraction yields for the SS bitumen in Figure 4.26. The extraction yield increased with increase in solvent density during SFE of the SS bitumen using propane as solvent. It is observed from the plot that the cumulative extraction yields with the SS bitumen increased as the pure solvent density increased. 242 Figure 4.26 Effect of Solvent Density on Extraction Yields with the Sunnyside Bitumen 243 25- Cumulative Extraction Yield, wt% 20- 15- 10 - 5- 0 0.53 ■ i I ' 0.535 0.54 l f 0.545 1 i 1 1 0.55 i i i 0.555 0.56 0.565 Propane Density, g/cc 0.57 244 The pure solvent densities were measured at the five operating conditions. The extract phase densities were measured on a continuous basis during each extraction. The measured extract phase densities indicate that the starting densities were 0.571, 0.602, 0.58, 0.572 and 0.604 g/cm3 at 5.6 MPa and 380 K, 10.4 MPa and 339 K, 10.4 MPa and 380 K, 10.4 MPa and 422 K and 17.3 MPa and 380 K; respectively. It was observed that the relationship between the pure solvent densities and the starting extract phase densities and operating variables was consistent; thus it was possible correlate the extraction yields by the pure solvent density. The propane density (Figure 2.15) could be varied by adjusting the extraction pressure and temperature. Thus, the propane densities at 10.4 MPa and 339 K and 17.3 MPa and 380 K were similar: 0.566 and 0.569 g/cm3, respectively. Furthermore, the cumulative extraction yields of SS bitumen, 22.4 and 23.7 wt%, corresponding to these densities indicated the significance of solvent density as a correlating parameter. If a combination of pressure and temperature is selected that maintains a fixed solvent density, similar extraction yields should result. An alternate explanation, an increase in E with increase in solvent density, which was described relative to the extraction of the AR bitumen is also valid for the SS bitumen. Carbon Number Distribution for SS Bitumen Extract Phases The extract fractions obtained during SFE of the SS bitumen were analyzed using a modified simulated distillation technique to obtain boiling point 245 and carbon number distributions up to 973 K (C 90). Carbon number distributions for the extract samples obtained from SFE of the SS bitumen at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03) are presented in Figure 4.27. As the extraction proceeded heavier and heavier components were extracted as indicated in Figure 4.27. The extract phases were significantly upgraded (volatilities ~ 80 wt%) compared to the original feedstock (volatility 40.9 wt%). The residual fraction was approximately 20 wt% volatile (fraction boiling below 811 K). The effects of temperature and pressure on compositional variation of the extract phases are compared in Figure 4.28. As the system pressure increased at constant temperature heavier extract fractions were obtained. As the system temperature increased at constant pressure lighter extract liquids were obtained. It is observed from the experimental results that at constant temperature, as the extraction pressure increased, the extraction yield increased from 1 2 . 0 wt% (@5.6 MPa) to 23.7 wt% (@17.3 MPa). At the same time the volatility (fractions boiling below 811 K) of the 2nd extraction window extract decreased from 89.6 wt% (@5.6 MPa) to 63.1 wt% (@17.3 MPa). The SFE system was operated on a semicontinuous basis and hence the increase in extraction yield with pressure led to a decrease in overall volatility (quality) of the extract phase. As the system temperature increased at constant pressure (10.4 MPa), the cumulative extraction yield decreased from 22.4 wt% (339 K) to 11.2 wt% (422 K). The volatility (quality) of the 2nd window extract phase increased from 62.4 wt% (339 K) to 83.4 wt% (422 K). Thus at constant temperature, as the extraction pressure increased the extraction yields increased and produced heavier or poorer quality 246 Figure 4.27 Carbon Number Distributions for the Sunnyside Bitumen and the Extract and Residual Fractions Obtained from SFE at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03) Cumulative Weight Fraction 247 Carbon Number 248 Figure 4.28 Effect of Pressure and Temperature on the Carbon Number Distribution of the Second Extraction Windows Obtained from SFE with the Sunnyside Bitumen 249 Effect of Pressure 1 2nd Extraction Window ^+++ ♦ ♦ ♦ ♦ 380 K(Tr=1.03) Cumulative Weight fraction 0.9 0.8 0.7 0.6 0.5 ♦ 0.4 ♦ O O I 0.3­ 0.2 0.1 0 10 20 30 40 50 60 Carbon Number 70 80 90 Effect o f Tem peratu re 1 2nd Extraction Window 10.4 MPa (Pr=2.3) r Cumulative Weight Fraction 0.9 0.8 » *« ***’ 0.7 0.6 o ■ * o m* 0.5 ■ * O ■ * 0.4 0.3 o ■* 0.2 °m * cm * 0.1 * 339 K(Tr=0.92) ■ 380 K(Tr=1.03) O 422 K(Tr=1.14) 0 0 10 20 30 40 50 60 Carbon Number 70 80 90 250 hydrocarbon liquids. temperature In contrast, at constant pressure as the extraction increased extraction yields decreased but produced refined hydrocarbon liquids. These are similar to the observations made for the AR bitumen-propane system and hence these observations could be generalized for other feedstocks also. Reproducibility SFE experiments were conducted at 10.4 MPa (Pr=2.3) and 339 K (Tr=1.03) using the SS bitumen to demonstrate the reproducibility of the experimental results. The extraction yields obtained in these experiments are presented in Figure 4.29. The difference between the extraction yields is within 5% and hence it was concluded that the experimental results were reproducible. Comparison of SFE of Bitumens from Uinta Basin SFE of four different bitumens from the Uinta Basin of Utah has been conducted at five different operating conditions using propane as the solvent. The extraction yields were determined from the SFE experiments. The bitumen and the residual fractions obtained from SFE were fractionated in to saturates, aromatics, resins and asphaltenes. The boiling point or carbon number distributions was obtained for all the bitumens for the extract and residual fractions, and for the solubility fractions such as saturates, aromatics and resins for the bitumens and residual fractions. Elemental analyses were also obtained for the bitumens and residual fractions as part of the detailed analyses. The effect of pressure, temperature, solvent density and feed chemical 251 Figure 4.29 Reproducibility for SFE with the Sunnyside Bitumen at 10.4 MPa (Pr=2,3) and 339 K (Tr=0.92) Weight % Extracted 252 Window Number 253 composition on the SFE yields of four different bitumens at five operating conditions will be discussed in this section. Pressure Effect Cumulative extraction yields were obtained for the four bitumens at a constant temperature of 380 K (Tr=1.03) and at three pressures 5.6 MPa (Pr=1.2), 10.4 MPa (Pr=2.3) and 17.3 MPa (Pr=4.1). Plots of cumulative extraction yields versus system pressure for the bitumens are presented in Figure 4.30. As the system pressure increased at constant temperature the cumulative extraction yields increased. As discussed before, the increase in extraction yields has been attributed to the increase in the solvent density with increased pressure at constant temperature for a particular bitumen. Moreover, the W R bitumen gave the maximum yield at all three pressures, the AR bitumen gave the second highest yields, and the SS bitumen gave the lowest. The PRS extraction yields were intermediate between the yields for the AR and SS bitumens. The WR bitumen extraction yields were 6 to 10 % (absolute) more than those of the AR Bitumen. The AR bitumen yields were 1 to 5 % more than PRS bitumen. The PRS bitumen extraction yields were 1 to 7 % greater than the yields for the more refractory SS bitumen. The ranking for the four bitumens according to extraction yield as follows: Whiterocks > Asphalt Ridge > PR Spring > Sunnyside This ranking is similar to the ranking observed when the physical (specific gravity, Conradson carbon and viscosity) and chemical (asphaltene and resin 254 Figure 4.30 Effect of Pressure on the Extraction Yields for the Four Bitumens from the Uinta Basin at Constant Temperature 380 K (Tr=1.03) Cumulative Extraction Yield, wt% o cn -o k o- ^i K o J ci s nj co o mw -o U o- ^i c on 255 * contents) properties of the four bitumens are compared. The bitumen (WR) judged to possess superior physical and chemical properties gave greater extraction yields compared to the bitumen (SS) judged to possess inferior physical and chemical properties. Thus, the difference in the cumulative extraction yields amongst four bitumens at a particular temperature could be attributed to their physical and chemical properties. Temperature Effect The effect of temperature on the cumulative extraction yield at a constant pressure of 10.4 MPa (P,=2.3) for the four bitumens is presented in Figure 4.31. As the system temperature increased at constant pressure, the cumulative extraction yield decreased for the four bitumens. The decrease in the extraction yields with increase in temperature was attributed to the decrease in solvent density as the temperature increased. The cumulative extraction yields for the W R bitumen were highest and were 10 to 12 wt% greater than those for the AR bitumen. The extraction yields of SS bitumen were lowest, and were 2 to 6 wt% less than those for the PRS bitumen. The AR bitumen and PRS extraction yields were intermediate with the former yields 2 to 7 wt% higher than that of PRS bitumen. Thus, the order as indicated by the extraction yields was as follows: Whiterocks > Asphalt Ridge > PR Spring > Sunnyside This ranking is similar to the ranking observed with the pressure effects. The cumulative extraction yields for the four bitumens should remain the same if the extraction yields were solely governed by the operating pressure and 257 Figure 4,31 Effect of Temperature on the Extraction Yields for the Four Bitumens from the Uinta Basin at Constant Pressure 10.4 MPa (Pr=2.3) Cumulative Extraction Yield, wt% 325 335 345 355 258 Temperature, K 259 temperature of the SFE. However, it was observed that the four bitumens exhibited different extraction yields at the same operating conditions with the order as shown above. Thus, the extraction yields of the bitumen are not only governed by the operating parameters but also by the feedstock physical and chemical make up. The bitumen with the lowest specific gravity, Conradson carbon, viscosity, asphaltene and highest resin content (WR bitumen) exhibited higher extraction yields than the other three bitumens. The SS bitumen that had the highest specific gravity, Conradson carbon, viscosity, asphaltene and lowest resin content gave lowest extraction yield. The AR bitumen and PRS bitumen extraction yields were in the intermediate range with AR bitumen yield greater than those of the PRS bitumen. This was consistent with the perception that the AR bitumen was of a higher quality than the PRS bitumen. The effect of che'mical make up of the bitumen on the extraction yield is discussed in detail in the subsequent sections. Solvent Density Effect The cumulative extraction yields at all five operating conditions for the four bitumens are plotted against the pure propane solvent density in Figure 4.32. The pure solvent densities were measured in a separate experiment and are reported in Table 4.6. The cumulative extraction yields increased with increase in solvent density. As indicated previously, the W R bitumen gave the highest extraction yields at all five operating conditions whereas the SS bitumen 260 Figure 4.32 Effect of Solvent Density on the Extraction Yields for the Four Bitumens from Uinta Basin o T 1" 0.545 f I- - - - - - - - 1- - - - - - - - 1 i 0.55 ( i j- - - - - - - - I 0.555 Solvent Density, g/cc ' — SS Bitumen --------- 1 | I 0.56 I - - - - - - 1- - - - - - - - 1 1 1 ” ^ .......... l 0.565 I- - - - - - - - - - 1- - - - - - - - 1 - - - - - - ' 0.57 262 gave the lowest yields. Thus, the ranking of the four bitumens according to extraction yields: Whiterocks > Asphalt Ridge > PR Spring > Sunnyside was the same for each of the process variables studied: This would seem to indicate that it may be possible to link process extraction yields to key chemical and physical attributes of the bitumens. The measured extract phase densities for the AR and SS bitumens at all the five operating conditions were plotted against extraction time and are presented in Figures 4.17 and 4.24, respectively. It can be observed from these plots that the starting extract phase densities at 17.3 MPa (Pr=3.2) and 380 K (Tr=1.03) were 0.71 and 0.604 g/cm3 for the AR and SS bitumens, respectively. The corresponding cumulative extraction yield for the AR bitumen was higher at 31.4 wt% compared to SS bitumen at 23.7 wt%. Thus, the higher extraction yield for the AR bitumen relative to the SS bitumen at same solvent density could be attributed to the chemical make up of the bitumens. Similar observations could be made for the AR bitumen at the other four operating conditions, where the starting extract phase density and corresponding extraction yields were higher than for the SS bitumen. Effect of Bitumen Asphaltene Content The extraction yields were different for the four bitumens under investigation at the same temperature, pressure and solvent density. This could be attributed to the difference in the chemical nature of the bitumens (Table 4.1). The asphaltene contents of the four bitumens varied from 2.9 wt% (W R) to 23.6 wt% (SS). The cumulative extraction yields obtained at the five different operating conditions for the four bitumens were plotted against the asphaltene content of the feedstocks and are presented in Figure 4.33. It is observed from the plot that as the asphaltene contents of the bitumens increased from 2.9 wt% for the W R bitumen to 23.6 wt% for the SS bitumen, the cumulative extraction yields decreased. The W R bitumen that had the lowest asphaltene (2.9 wt%) content gave the highest extraction yields at all five operating conditions. The SS bitumen with an asphaltene content of 23.6 wt% gave the lowest extraction yields. The AR (6.7 wt%) and PRS bitumens (19.3 wt%) extraction yields were intermediate with the AR extraction yields higher than the PRS yields at all five conditions. It was established by Speight[2] that for a bitumen sample, the amount of asphaltene precipitated increased exponentially from 18 wt% to 48 wt% when the solvent was switched from pentane to propane. The difference in the propane and pentane soluble fractions is related to the resin molecules present in the bitumens. It should be noted that the highest resin content was 54.5 wt% for the W R bitumen whereas the lowest resin content was 36.8 wt% for the SS bitumen. The AR and PRS bitumens fall in the intermediate range with AR bitumen resin content (44.1 wt%) higher than the PRS bitumen resin content (43.8 wt%). Thus, the bitumen (WR) that had lowest pentane insolubles and highest resin content was expected (assuming equal increase in asphaltene content from pentane to propane insolubles) to have the lowest propane 264 Figure 4.33 Relationship Between Asphaltene Content and Extraction Yield for the Four Uinta Basin Bitumens AR PRS Cumulative Extraction Yield, wt% WR 9— -i----□— ■J----■---- 10 5.6 10.4 10 4 104 17.3 15 Asphaltene Content, wt% MPa * 380 M Pa* 339 M Pa* 380 M P a * 422 M Pa* 380 20 SS K K K K K 266 insolubles compared to the SS bitumen (highest pentane insolubles and lowest resin content). The AR and PRS bitumen propane insolubles were expected to fall in the intermediate range with AR bitumen propane insolubles lower than the PRS bitumen insolubles. As stated in the earlier section, the maximum extraction yield obtainable using supercritical propane will be equal to or less than the maximum deasphalted oil obtainable using liquid propane. Thus, the amount of propane soluble fractions available for extraction is in the order: Whiterocks > Asphalt Ridge > PR Spring > Sunnyside Thus, the WR bitumen that would have highest propane soluble fractions (largest fraction available for extraction since asphaltene was not extracted) gave a higher extraction yield than the other bitumens at all five operating conditions. The SS bitumen was extracted least and the AR and PRS bitumens extraction yields fell in the intermediate range with the AR bitumen yields higher than the PRS bitumen yields. The pentane insoluble test performed on the AR and SS extract samples (2nd window at 10.4 MPa and 339 K) did not yield any precipitate. Thus, the relationship established between the asphaltenes and extraction yields indicates that the asphaltenes (pentane insolubles) played a significant role in decreasing the extraction yield of bitumens and asphaltenes were not transferred to the extract phase. Effect of Bitumen Resin Content The resin contents of the four feedstocks are plotted against the cumulative extraction yields obtained for the SFE of the bitumens at the five 267 different operating conditions in Figure 4.34. It is observed from the plot that as the resin content of the feedstock decreased from 54.5 to 36.8 wt%, the cumulative extraction yields decreased. The WR bitumen had the highest resin content, 54.5 wt%, and the lowest asphaltene content, and gave the highest extraction yields relative to the other bitumens. The SS bitumen that had the lowest resin content (36.8 wt%) and the highest asphaltene content gave the lowest extraction yields at all the five operating conditions. The AR bitumen (44.1 wt%) extraction yields were lower than those of the WR bitumen and marginally higher than those of the PRS bitumen (43.8 wt%) at all five conditions. This trend is similar to the observation made for the asphaltene contents of the bitumens. Effect of Bitumen Saturate and Aromatics Content No clear trends were observed based on the saturates and aromatics contents of the bitumens. It was expected that the AR bitumen (saturates content 39.2 wt%) which had the highest saturates content should have given high extraction yields similar to those of the WR bitumen (saturates content 35.7 wt%); however, the WR bitumen gave higher extraction yields than the AR bitumen at all five operating conditions. The PRS (saturates content 33.4 wt%) and SS bitumens (saturates content 20.0 wt%) again gave intermediate extraction yields with the PRS bitumen yields lower than those of the AR bitumen and higher than those of the SS bitumen. It was also observed that the cumulative extraction yields for the WR, PRS and SS bitumens increased with 268 Figure 4.34 Relationship Between Resin Content and Extraction Yield for the Four Uinta Basin Bitumens 102 8 Cumulative Extraction Yield, wt% 7 6 5 5.6 MPa *380 K 10.4 MPa* 339 K 10.4 MPa* 380 K 10.4 MPa* 422 K 17.3 MPa * 380 K 4 3 2 1.5 101 6 7 6 40 45 50 269 Resin Content, wt% 55 270 increase in the volatility (fraction boiling below 811 K). It was expected that the AR bitumen with the greater volatility and the highest saturates content would exhibit higher extraction yields than the WR bitumen; however, the experimental results indicted that the AR yields were lower than those obtained for the WR bitumen at all five operating conditions. An explanation for this behavior is proposed based on the boiling point distributions of the bitumen solubility fractions. A comparison has been made in Table 4.7 of the contributions from the solubility fractions to the volatility (<811 K) and to fraction boiling below 973 K for the four bitumens. It is observed from Table 4.7 that the WR, AR, PRS and SS bitumens volatilities were 46.6, 53.5, 45.4 and 40.9 wt%, respectively. As explained before, no clear trend was observed when an attempt was made to correlate cumulative extraction yields based on the volatilities of the bitumens. The fractions boiling below 811 K consisted of saturate, aromatics, resin and asphaltene solubility fractions. The contribution to the volatility from the asphaltene class of compounds was small compared to the contributions from other three solubility classes. The trend observed in Figure 4.35 could not be explained based on the estimated contribution to the volatility from the three solubility classes since the AR bitumen had a higher concentration of these classes than the WR bitumen (Table 4.7) which exhibited the apparent anomalous behavior. All the extract phase samples were characterized using a simulated distillation technique with a maximum boiling point of 973 K. The AR and SS 271 Table 4.7 Comparison of Boiling Fractions for Four Bitumens Properties a) Asphaltenes , wt % Saturates, wt % Aromatics, wt % Resins, wt % Simulated Distillation Volatility(<811 K) of Bitumen, wt% Volatility (<811 K) of Saturates, wt% Volatility (<811 K) of Aromatics, wt% Volatility (<811 K) of Resins, wt% Whiterocks Bitumen 2.9 35.7 7.0 54.5 Asphalt Ridge Bitumen 6.8 39.2 9.0 44.1 PR Spring Bitumen 19.3 33.4 3.6 43.8 Sunnyside Bitumen 23.6 20.0 15.1 36.8 46.6 53.5 45.4 40.9 81.2 86.2 78.5 84.1 28.5 23.2 40.9 30.2 21.6 20.4 29.8 19.0 Contribution from Saturates + Aromatics and Resin towards Volatility (<811 K) of Bitumenb) 42.7 44.9 40.7 28.4 Boiling Fraction (< 973 of Bitumen, wt% Boiling Fraction (< 973 of Saturates, wt% Boiling Fraction (< 973 of Aromatics, wt% Boiling Fraction (< 973 of Resins, wt% K) 78.7 90.1 78.1 73.4 99.8 100.0 99.4 99.6 71.4 56.5 80.3 72.9 56.3 54.7 62.4 55.0 K) K) K) Contribution from Saturates + Aromatics and Resin towards Boiling Fraction (< 973 K) of Bitumenb) 71.3 63.4 68.40 51.2 Pentane Insolubles b) Estimated on prorated basis 272 Figure 4.35 Relationship Between Saturates Content and Extraction Yield for the Four Uinta Basin Bitumens Amount of Bitumen Extracted, wt% 100 10 ■ 5.6 MPa*380 K ♦ 10.4 MPa * 339 K □ 10.4 MPa * 380 K + 10.4 MPa* 422 K • 17.3 MPa*380 K ------------r 15 20 35 40 273 25 30 Saturates Content, wt% 274 extract phase samples (2nd extract window samples produced at 10.4 MPa and 339 K) were subjected to pentane insoluble analysis. The test results confirmed that these extract phase samples did not contain asphaltene solubility class compounds. Based on the boiling point distributions of the saturates, aromatics and resins fractions for the four bitumens, the contribution towards the bitumen solubility fractions boiling below 973 K was estimated on a prorated basis relative to the corresponding three solubility classes in the respective bitumens and is presented in Table 4.7. These estimated values indicated that the WR bitumen had the highest extractable (saturates, aromatics and resins) 973 K minus fraction: 71.3 wt% and the SS bitumen had the lowest extractable 973 minus fraction: 51.2 wt%. The AR and PRS bitumens extractable 973 K minus fractions fell in the intermediate range with the AR bitumen 973 K minus fraction (68.4 wt%) higher than that of the PRS bitumen (63.4 wt%). The experimental results also indicated that at all five extraction conditions, the WR bitumen yield was greater than those obtained for the other three bitumens. The refractory SS bitumen gave the lowest extraction yields whereas the AR and PRS bitumen were in the intermediate range with the AR bitumen yields greater than the yields for the PRS bitumen. Thus, the cumulative extraction yield trend obtained from the SFE of the four bitumens using propane as solvent was controlled by the extractable solubility class compounds present in the 811 K plus and 973 K minus range. 275 Compositional Analyses of Residual Fraction The residual fractions in the extractor at the completion of each extraction experiment were fractionated into saturates, aromatics, resins and asphaltenes using adsorption chromatography. The fractionation technique is outlined in Appendix D. The fractionations of the residual fractions were performed to determine the nature of the material left behind in the extractor. The results of the compositional and elemental analyses of the residual fractions obtained from SFE of the four bitumens at five different operating conditions are presented in Tables 4.8 through 4.11 Saturates Content of Residual Fractions A partial SARA analysis was performed on the residual fractions from the SFE of the WR and PRS bitumens by Subramanian[40] without fractionating the saturates and aromatics into saturates, aromatics I and II using neutral alumina. Hence, the results were reported as saturates and aromatics. However, the residual fractions from SFE of the AR and SS bitumens were subjected to complete fractionation into saturates, aromatics, resins and asphaltenes. The saturates and aromatics contents of the WR and PRS bitumen are compared with the saturates and aromatics contents of their respective residual fractions obtained at all five operating conditions in Figure 4.36 and 4.37. The solvent densities are plotted against the saturates and aromatic contents instead of operating variables pressure and temperature, because the extraction yields were proportional to the solvent density, which represents the severity of 276 Table 4.8 Summary of Extraction Yields and Residual Fractions Analyses for the Whiterocks Bitumen Pressure (MPa) Reduced Pressure 5.6 1.2 Temperature (K) 380 Reduced Temperature 0.92 Solvent Density, g/cc 0.533 10.4 2.3 10.4 2.3 10.4 2.3 17.3 4.1 339 1.03 380 1.03 422 1.03 380 1.14 0.566 0.553 0.545 0.569 Product Yields (wt%) Extract Phase Residual Phase 40.0 59.0 39.0 58.0 24.0 73.0 48.0 50.0 24.9 63.7 11.4 15.5 74.5 10.0 20.1 64.5 15.4 27.0 62.2 10.8 0.0 76.5 23.5 3.7 4.9 5.0 4.0 5.8 86.7 10.8 2.0 0.5 1.49 86.7 11.1 1.8 0.5 1.53 86.8 10.6 2.1 0.5 1.46 20.0 79.0 Residual Phase SARA Analyses Saturates & Aromatics, wt% Resins, wt% Asphaltenes, wt% Asphaltenes, wt% (Expected) Residual Phase Elemental Analysis3) 86.8 86.7 C, wt% H, wt% 10.8 11.0 1.8 N, wt% 2.0 0.5 0.5 S, wt% H/C Atomic Ratio 1.52 1.50 a) Normalized to 100% 277 Table 4.9 Summary of Extraction Yields and Residual Fractions Analyses for the Asphalt Ridge Bitumen Pressure (MPa) Reduced Pressure Temperature (K) Reduced Temperature Solvent Density, g/cc 10.4 2.3 10.4 2.3 10.4 2.3 17.3 4.1 339 1.03 380 1.03 422 1.03 380 1.14 5.6 1.2 380 0.92 0.533 0.566 0.553 0.545 0.569 Product Yields (% ) Extract Phase Residue Phase 31.3 68.7 24.5 75.5 18.2 81.8 31.4 68.6 22.6 3.5 58.9 15.0 5.5 1.6 60.0 33.0 13.3 2.3 61.9 22.5 18.8 4.9 59.5 16.8 4.1 1.4 56.5 38.0 7.8 9.9 9.0 8.3 9.9 87.29 10.50 1.73 0.49 1.44 87.00 10.66 1.85 0.49 1.47 87.11 11.00 1.42 0.47 1.51 87.18 10.91 1.43 0.48 1.50 13.4 86.6 Residual Phase SARA Analvses Saturates, wt% Aromatics, wt% Resins, wt% Asphaltene, wt% Asphaltene, wt% (Expected) Residual Phase Elemental Analysis3) C, wt% H, wt% N, wt% S, wt% H/C Atomic Ratio a) 87.38 10.90 1.21 0.50 1.50 Normalized to 100.0 % 278 Table 4.10 Summary of Extraction Yields and Residual Fractions Analyses for the PR Spring Bitumen Pressure (MPa) Reduced Pressure 5.6 1.2 Temperature (K) 380 Reduced Temperature 0.92 Solvent Density, g/cc 0.533 10.4 2.3 10.4 2.3 10.4 2.3 17.3 4.1 339 1.03 380 1.03 422 1.03 380 1.14 0.566 0.553 0.545 0.569 Product Yields (% ) Extract Phase Residual Phase 23.0 77.0 20.8 79.2 15.7 84.3 31.7 68.3 29.3 50.3 20.4 22.7 42.0 35.3 10.0 53.8 36.2 21.0 52.6 26.4 10.8 37.1 52.0 21.2 25.1 24.4 22.9 28.3 87.3 10.8 1.3 0.6 1.47 87.3 11.0 1.1 0.6 1.50 87.4 10.6 1.4 0.6 1.44 8.8 91.2 Residual Phase SARA Analyses Saturates & Aromatics, wt% Resins, wt% Asphaltenes, wt% Asphaltenes, wt% (Expected) Residual Phase Elemental Analysis3) C, wt% H, wt% N, wt% S, wt% H/C Atomic Ratio a) 87.4 11.0 1.1 0.5 1.50 Normalized to 100% 87.4 10.5 1.4 0.7 1.44 279 Table 4.11 Summary of Extraction Yields and Residual Fractions Analyses for the Sunnyside Bitumen Pressure (MPa) Reduced Pressure Temperature (K) 380 Reduced Temperature 0.92 Solvent Density, g/cc 10.4 2.3 10.4 2.3 10.4 2.3 17.3 4.1 339 1.03 380 1.03 422 1.03 380 1.14 5.6 1.2 0.533 0.566 0.553 0.545 0.56! Product Yields (%) Extract Phase Residual Phase 22.4 77.6 14.8 85.2 11.2 88.8 23.7 76.3 13.3 7.3 51.5 37.9 3.3 1.2 37.1 58.4 5.4 3.5 35.9 55.2 16.7 6.8 40.2 36.3 1.3 1.5 37.0 60.3 26.8 30.4 27.7 26.6 30.9 87.20 10.14 1.42 1.23 1.39 87.31 10.28 1.17 1.24 1.41 87.32 10.60 1.02 1.05 1.46 86.72 9.86 1.41 1.23 1.36 12.0 88.0 Residual Phase SARA Analyses Saturates, wt% Aromatics, wt% Resins, wt% Asphaltene, wt% Asphaltene, wt% (Expected) «) Residual Phase Elemental Analvsis* C, wt% H, wt% N, wt% S, wt% H/C Atomic Ratio a) 87.41 10.40 1.08 1.11 1.43 Normalized to 100% 280 Figure 4.36 Effect of Solvent Density on the Extraction of Saturates and Aromatics from the Whiterocks Bitumen 281 Saturates and Aromatics Content, wt% 50 40 30 20 10 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 282 Figure 4.37 Effect of Solvent Density on the Extraction of Saturates and Aromatics from the PR Spring Bitumen 283 50 n Saturates & Aromatics Content of the Bitumen 45 Saturates and Aromatics Content, wt% Saturates & Aromatics Content of the Residual Fractions 40 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 284 operation in SFE processes. The saturates and aromatics contents of the residual fractions for the WR and PRS bitumens were lower than the saturates and aromatics contents of the respective bitumens, and in general decreased with increase in solvent density (severity of operation). Thus, it was concluded that as the solvent density or the severity of operation or the extraction yields increased more and more of the saturates and aromatics were extracted. It has also been pointed out in the earlier sections that the cumulative extraction yields increased with increase in the solvent density meaning more of the saturates and aromatics were removed in the process. Moreover, the resin contents (Tables 4.8 through 4.11) of the residual fractions obtained during the SFE of four bitumens were higher than the resin contents of the original bitumens except for the PRS bitumen at two operating conditions where the solvent densities are higher compared to the other three conditions. This indicates that the contribution to the extract phases from resins was less than that from the saturates and aromatics. Analyses of selected extract samples indicated that no asphaltenes (pentane insolubles) were present. Thus, it was concluded that propane preferentially extracted saturates and aromatics from bitumens at the expense resins and asphaltenes. The saturates contents of the AR and SS bitumens and the respective residual fractions have been plotted and are presented in Figures 4.38 and 4.39. The trends are similar to those reported for the WR and PRS bitumens. This led 285 Figure 4.38 Effect of Solvent Density on the Extraction of Saturates from the Asphalt Ridge Bitumen 286 50 [~| Saturates Content of the Bitumen 45 Saturates Content of the Residual Fractions Saturates Content, wt% 40 35 30 25 20 15 10 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 287 Figure 4.39 Effect of Solvent Density on the Extraction of Saturates from the Sunnyside Bitumen 288 □ Saturates Content of the Bitumen Saturates Content, wt% Saturates Content of the Residual Fractions 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 289 to the conclusion that the saturates and aromatics were preferentially extracted by propane compared to the asphaltenes and resins. Asphaltene Content of Residual Fractions The residual fractions of the four bitumens after SFE were mixed with excessive amounts of n-pentane to yield pentane insolubles, i.e., asphaltenes. The asphaltene content of the residual fractions was examined to determine the effect of process variables on the removal or rejection of the asphaltenes and also to investigate the nature of the unextracted residual fraction. The measured asphaltene content of the residual fractions is compared with the expected asphaltene content of the residual fractions and the asphaltene content of the respective bitumens in Figures 4.40 through 4.43. The expected asphaltene contents of the residual fractions were calculated on a prorated basis relative to the amount of bitumen charged to the extractor and assuming that all the asphaltenes remained in the residual fractions. Selected extract samples from SFE of the AR and SS bitumens at 10.4 MPa (Pr=2.3) and 339 K (Tr=1.03) were analyzed and there was no evidence that asphaltenes were transferred to the extract phases. The asphaltene contents of the residual fractions for the four bitumens were much higher than those of the bitumens and the expected asphaltene contents which had been calculated on a prorated basis. An explanation for this phenomenon can be generated by revisiting the literature. Nellensteyn[203] proposed initially a colloidal model for the heavy asphaltic material present in 290 Figure 4.40 Effect of Solvent Density on the Extraction of Asphaltenes from the Whiterocks Bitumen 291 B Bitumen Q Residual Fractions (Expected) ~ ■ Residual Fractions (Measured) Asphaltene Content, wt% 25 0 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 292 Figure 4.41 Effect of Solvent Density on the Extraction of Asphaltenes from the Asphalt Ridge Bitumen 293 S\ 50 FU Bitumen □ Residual Fractions (Expected) Asphaltene Content, wt% - - Residual Fractions (Measured) 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 294 Figure 4.42 Effect of Solvent Density on the Extraction of Asphaltenes from the PR Spring Bitumen Asphaltene Content, wt% 295 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 296 Figure 4.43 Effect of Solvent Density on the Extraction of Asphaltenes from the Sunnyside Bitumen 297 Bitumen Residual Fractions (Expected) Asphaltene Content, wt% Residual Fractions (Measured) 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 298 naturally occurring crude oils. According to him, the asphaltic materials were made up of micelles covered or shielded from the bulk of the oil by adsorbed resins (cosolubilizing agents) and other hydrocarbon materials. The micelles were assumed to be dispersed in the hydrocarbon medium. He also suggested that the precipitating properties of asphaltic materials in different solvents are related to the surface tension. Swanson[204] and Witherspoon and Munir[205] suggested that the resins help asphaltenes stay dissolved in the distillate portion of the naturally occurring crude oil, heavy oil, bitumen, etc. Dickie and Yen[206] proposed that the petroleum resins act as an interface between the polar asphaltenes and relatively nonpolar oil fractions in petroleum. Koots and Speight[207] proposed that the resin fractions play major role in maintaining the asphaltenes in a colloidal state in crude oils. Leontaritis and Mansoori[207] proposed a thermodynamic model of the colloidal state to predict the onset of asphaltene flocculation. Mitchell and Speight[209) indicated that asphaltene precipitation increased exponentially with a decrease in carbon number in the paraffinic solvent used for precipitation. Thus, the composition of the asphaltene precipitated using different solvents was different. The asphaltene molecular weight increased and the H/C atomic ratio decreased with increase in the carbon number of the paraffinic solvent used for precipitation. There are other process variables such as the ratio of solvent-to-feedstock, the precipitation temperature and the contact time that influence the amount of asphaltene precipitated. Thus, the interaction between resins and asphaltenes is the key to the colloidal suspension of the asphaltenes in oil. This interaction is disturbed by the 299 addition of a nonpolar solvent, the equilibrium that existed between the resin (soluble in the nonpolar solvent) and asphaltenes (insoluble in the nonpolar solvent) was disturbed, and permitted the aggregation of asphaltene molecules and their precipitation from solution. The analysis of the residual fractions produced in the SFE process indicated the presence of saturates, aromatics, resins and pentane insolubles when propane was used as the solvent. At the same time, the propane-free extract phases consisted of saturates, aromatics and resins and no asphaltenes or pentane insolubles. During the SFE process, the resins or the cosolubilizing agents that kept the asphaltenes in solution in the original bitumen were transferred to the extract phase. Without these resin molecules, some of the asphaltene that did not precipitate from the original bitumen (while performing pentane insolubles) was precipitated after the bitumen sample was subjected to SFE. The observation made here was also observed by Koots and Speight[207] while studying the relationship between resin and asphaltenes in bitumen or crude oils. While using pentane and heptane as solvent, the Athabasca bitumen yielded 17 and 11 wt% of asphaltenes, respectively. Thus, 35 % of the pentane insolubles remained in solution when heptane was used as the solvent. An attempt was made to dissolve the pentane insolubles in heptane; however, only 10 % of the pentane insoluble dissolved. When resins were added, such that the asphaltene-resin ratio was same as that of the original bitumen, 33 % of the pentane insoluble asphaltenes were soluble in pentane. The authors also observed similar effects using other hydrocarbon solvents. Thus, the removal of resins (cosolubilizing agents)resulted in an 300 increase in the pentane insolubles in the residual fractions obtained from SFE of bitumens using propane. The measured asphaltene content of the residual fractions increased with increase in the solvent density for all the four bitumens (Figure 4.40 through 4.43). This correlated well with the increased extraction yields with increase in solvent density for the four bitumens and also with the decrease in the saturate and aromatic contents of the residual fractions. Thus, it was concluded that the resins (cosolubilizing agents) that kept the asphaltenes suspended in the original bitumen have been removed during SFE. Elemental Analyses The elemental (C, H, N and S) analyses of the residual fractions for the bitumens are reported in Tables 4.8 through 4.11 for the WR, AR, PRS and SS bitumens, respectively. The nitrogen and sulfur compounds are primarily present in the resin and asphaltenes with the exception of sulfur in small quantities in the oils (saturates plus aromatics). The nitrogen and sulfur compounds have a tendency to concentrate in the asphaltenes relative to resins[210]. However, an exception to this trend was observed in the case of sulfur[210]. It has been observed from the solubility fractions of the SFE residual fractions that saturates and aromatics were extracted preferentially compared to resins and asphaltenes. The analyses of the extract samples indicated the absence of asphaltenes. Thus the pentane insolubles which contain the increased amounts of asphaltenes and consequently nitrogen and sulfur were 301 left behind in the residual fractions. The resins were transferred to the extract phase; however, the relative amount of extraction was less compared to the extraction of saturates and aromatics (saturates and aromatics content decreased in the residual fractions compared to the original bitumens). This was also indicated by the increase in the resin contents of the residual fractions compared to the original resin content of the bitumens. Thus, the heteroatoms such as sulfur and nitrogen that are predominantly contained in the asphaltenes, and to a lesser extent in the resins, were concentrated in the residual fractions. The nitrogen and sulfur concentrations of the residual fractions were consistently higher than the nitrogen and sulfur contents of the bitumen confirming that the heteroatoms are concentrated in the unextracted residual fractions. This is an important observation relevant to the quality of the upgraded hydrocarbon liquids produced in the SFE process. Elemental analyses of the extract samples need to be performed to quantify the amount of heteroatoms present in the extract samples, however, the quantities of the extract samples produced in this study were not sufficient to permit elemental analyses. The H/C atomic ratios of the residual fractions are compared with the H/C ratios of the bitumens in Figures 4.44 through 4.47 for the WR, AR, PRS and SS bitumens; respectively. The H/C ratios of the residual fractions were lower than the H/C ratios of the bitumens. Furthermore, as the solvent density increased and concomitantly the extraction yields, the H/C atomic ratio of the residual fractions decreased for the four bitumens implying that the saturates and 302 Figure 4.44 Relationship Between Pure Solvent Density and the Residual Fraction H/C Atomic Ratio for the Whiterocks Bitumen 303 1.65 1.6 H/C Atomic Ratio 1.55 1.5 1.45 1.4 1.35 1.3 1.25 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 304 Figure 4.45 Relationship Between Pure Solvent Density and the Residual Fraction H/C Atomic Ratio for the Asphalt Ridge Bitumen 305 1.7 H/C Atomic Ratio 1.65 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 306 Figure 4.46 Relationship Between Pure Solvent Density and the Residual Fraction H/C Atomic Ratio for the PR Spring Bitumen 307 1.65 1.6 H/C Atomic Ratio 1.55 1.5 1.45 1.4 1.35 1.3 1.25 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 308 Figure 4.47 Relationship Between Pure Solvent Density and the Residual Fraction H/C Atomic Ratio for the Sunnyside Bitumen H/C Atomic Ratio 309 0.533 0.545 0.553 0.566 Solvent Density, g/cc 0.569 310 aromatic compounds were preferentially extracted compared to the polar resins and asphaltenes. This was confirmed by an observed reduction in the saturates and aromatic contents of the residual fractions compared to the original bitumen. Modeling SFE Using Continuous Thermodynamics Principle The principles of continuous thermodynamics and its evolution towards process industry applications were discussed in Chapter 2. An attempt has been made to model the supercritical fluid extraction of oil sands bitumen with propane as solvent using the principle of continuous thermodynamics. As discussed earlier in Chapter 2, the pseudocomponent lumps used to represent the bitumen in modeling were arbitrarily specified whereas continuous thermodynamic principles uses mathematically selected quadrature points to represent the complex hydrocarbon mixtures. The critical properties were estimated at these quadrature points and flash calculations were conducted using the Peng-Robinson[41] equation of state to simulate the supercritical extraction process and understand the effect of bitumen composition on the SFE yields. The choice of the proper continuous distribution function and the number of quadrature points required to represent ultra heavy oils such as bitumen was very critical for the success of the modeling process. Choice of Continuous Distribution Function The characterization of bitumens using simulated distillation techniques[45] to obtain boiling point and/or molecular distributions is limited to 70-90 wt% of the bitumen. The molecular weight or boiling point distributions of 311 the bitumen formed the basis in the attempt to model the SFE process using continuous thermodynamics. The gamma distribution (open ended distribution) is the most suitable function (as explained in Chapter 2) to represent a continuous hydrocarbon mixture like bitumen, which can not be completely characterized. An attempt has been made in this study to fit the alkane based AR bitumen molecular weight distribution obtained from the simulated distillation data using a gamma distribution. The simulated distillation technique assumes that the normal alkane elutes last for a particular carbon number homologous series. Thus, the carbon number or boiling point or the molecular weight distributions derived from simulated distillation is alkane based (as discussed in Chapter 3). A proper fit could not be obtained using a gamma distribution because of the two predominant bio-markers present in the differential molecular weight distribution (Figure 4.48). Hence, an attempt was made using the cumulative molecular weight distribution (instead of the differential distribution) using a higher order polynomial distribution function of the AR bitumen with molecular weight and cumulative weight fraction as dependent and independent variables; respectively. Higher order polynomials can be used to fit a variety of distributions by varying the order of the polynomial. The boiling point distribution of the AR bitumen was determined up to 973 K which constitutes only 89 wt% of the total material. The molecular weight distribution was extended beyond 973 K (C™) by extrapolating the shape of the distribution to meet the y-axis at 2650 g/gmol. An 312 Figure 4.48 Continuous Molecular Weight Distribution Function for the Asphalt Ridge Bitumen Using a Gamma Distribution Molecular Weight, g/gmol 313 acceptable fit was obtained using an eighth order polynomial. The fit obtained for the AR bitumen is presented in Figure 4.49. Mani et al.[159] used true boiling point data and four quadrature points to represent a naphtha/kerosene blendstock (Edmister and Pollock[211]) which contained no distillable residue to the predict bubble point pressure of the mixture. The boiling points for these four quadrature points were obtained by graphical interpolation of the true boiling point-volume distilled data. measured and predicted bubble point pressures agreed very well. The A similar approach was used in this study except a higher order polynomial was used to fit the molecular weight-cumulative weight fraction data. Once the relationship between the dependent and independent variables was obtained, optimization calculations were performed to identify the number of quadrature points required to adequately represent the bitumen. Calculation Procedure Using the Quadrature Method Initially the number of quadrature points required to represent the bitumen was assumed to be from 4 to 10 in increments of two. The weight fractions corresponding to the chosen sets of quadrature points were calculated using equation 4.4: Fwi = 2( q i +1) [4.4] where the qi are the zeros of the Legendre polynomial corresponding to order n. The values of qi and wi for various orders n are tabulated[212]. Selected values 315 Figure 4.49 Continuous Molecular Weight Distribution Function for the Asphalt Ridge Bitumen Using a Higher Order Polynomial Function Molecular Weight, g/gmol oo o o o o o Cumulative Weight Fraction o ro o 4* o b> o 00 316 317 used for calculations in this study are listed in Table 4.12. The larger the value of n, the more accurate the results, hence optimization calculations were performed using 4, 6, 8 and 10 quadrature points to represent the AR bitumen. Using the Fw as calculated from Equation 4.4, the molecular weights at the quadrature points were specified using the higher order polynomial equation obtained from the AR bitumen molecular weight distribution (Figure 4.49). The boiling point or molecular weight distribution of the bitumen and bitumen products was obtained using an extended simulated distillation technique[45] assuming that in a particular carbon number series, normal alkanes are the last to elute in the chromatographic analysis. Thus, the relationship obtained between the retention times and the elution of alkanes was used to obtain the alkane based boiling point and/or molecular weight distributions. The relationship was then used to obtain the boiling points at these quadrature points. A typical relationship between boiling point and molecular weight[199] of n-alkanes ranging from carbon number 10 to 90 is presented in Figure 4.50. The specific gravities at these quadrature points were evaluated by a trial and error method using the calculated weighting faction, Fwj, to represent the overall measured bitumen specific gravity. The critical properties at these quadrature points were evaluated using the Lee-Kesler correlations[43] using the boiling points, molecular weights and specific gravities. The mole fraction of each component at these quadrature point segments was estimated using Equation 4.5[159]: 318 Table 4.12 Quadrature Points and Weight Factor for Gauss-Legendre lntegration[212] Wi ±qia) n=2 1.000000 0.5773503 n=3 0.555555 0.888889 0.7745967 0.0000000 n=4 0.3478548 0.6521452 0.8611363 0.3399810 n=6 0.1713245 0.3607616 0.4679139 0.9324695 0.6612094 0.2386192 n=8 0.1012285 0.2223810 0.3137066 0.3626838 0.9602899 0.7966665 0.5255324 0.1834346 n=10 0.9739065 0.8650634 0.6794096 0.4333954 0.1488743 0.0666713 0.1494513 0.2190864 0.2692602 0.2955242 To be used as positive and negative values for computation 319 Figure 4.50 Relationship Between Molecular Weight and Boiling Point (Alkane Based) Molecular Weight, g/gmol 320 Boiling Point, K 321 2 SGM , where: Xj is the mole fraction; SGj is the specific gravity of the quadrature segment I; M is the average or measured molecular weight of the bitumen; SG is the measured bitumen specific gravity, at 288 K/288 K; and; Mi is the molecular weight of the quadrature segment, i. Flash calculations were performed using the Peng-Robinson equation of state[41] to optimize the number of quadrature points used to represent the AR bitumen. The flash calculations were performed at 10.4 MPa (Pr=2.3) and 380 K (Tr=1.03) using a propane-bitumen mixture. The numbers of quadrature points considered for the optimization were 4, 6, 8 and 10. The corresponding molecular weights, boiling points and mole fractions were calculated using the quadrature method and are presented in Table 4.13. The mole % of the AR bitumen in the vapor phase is plotted against the number of quadrature points in Figure 4.51. As the number of quadrature points used to represent the AR bitumen increased from 4 to 10, in increments of 2, the mole % of bitumen in the vapor phase increased initially, however, it remained nearly constant at quadrature points 8 and 10. The changes were attributed to the number of components used to represent complex mixtures. Hence, it was concluded that 8 or 10 quadrature points were adequate to represent complex hydrocarbon 322 Table 4.13 Quadrature Components Properties for Optimization of Number of Components for Asphalt Ridge Bitumen qi Wi Fwi MWi BP, K SGi Xi 4 Quadrature Points 0.348 -0.861 -0.340 0.652 0.652 0.340 0.348 0.861 0.069 0.330 0.670 0.931 236.837 578.007 434.523 724.719 812.788 8876.108 1660.061 1048.751 0.850 0.950 0.980 1.200 0.369 0.377 0.202 0.053 6 Quadrature Points -0.932 0.171 -0.661 0.361 -0.239 0.468 0.239 0.468 0.661 0.361 0.932 0.171 0.034 0.169 0.381 0.619 0.831 0.966 195.260 531.339 311.492 739.207 461.360 739.207 726.410 848.946 1103.456 950.019 2092.838 1104.757 0.850 0.950 0.950 1.000 1.100 1.150 0.213 0.246 0.246 0.156 0.079 0.020 8 Quadrature Points -0.960 0.101 0.222 -0.797 -0.526 0.314 -0.183 0.363 0.183 0.363 0.526 0.314 0.797 0.222 0.960 0.101 0.020 0.102 0.237 0.408 0.592 0.763 0.898 0.980 168.408 495.574 260.585 601.107 370.068 685.903 476.654 747.091 680.209 833.059 962.902 917.080 1399.700 1007.509 2313.967 1129.038 0.850 0.880 0.920 0.950 0.980 1.050 1.120 1.200 0.151 0.214 0.213 0.191 0.138 0.082 0.040 0.011 10 Quad Points -0.974 -0.865 -0.679 -0.433 -0.433 0.149 0.433 0.679 0.865 0.974 0.13 0.067 0.160 0.283 0.426 0.574 0.717 0.840 0.933 0.987 151.757 470.406 235.159 576.287 303.936 638.310 405.496 708.005 487.365 752.463 652.467 822.993 889.175 897.823 1130.190 955.806 1679.504 1051.566 2434.576 1141.322 0.850 0.880 0.900 0.930 0.960 0.980 1.030 1.090 1.130 1.180 0.110 0.160 0.181 0.167 0.152 0.114 0.076 0.049 0.022 0.00 0.067 0.149 0.219 0.269 0.296 0.296 0.269 0.219 0.149 0.067 323 Figure 4.51 Optimization Results to Choose the Number of Quadrature Points Mole % of Bitumen in Vapor Phase k to W -U o 0) 1 ro c*> I Tv ? Oi O) 324 325 mixtures like the AR bitumen for phase equilibrium calculations. Eight quadrature points were used for all further calculations. Modeling Supercritical Extraction Process The supercritical fluid extraction of oil sands bitumens from the Uinta basin of Utah has been investigated using propane as the solvent. Mining and surface recovery methods have been commercialized and synthetic crude oil is being produced from oil sands using the hot water recovery process[17]. The separated bitumen is upgraded to a refinery acceptable synthetic crude oil using various coking processes[35,36]. The application of supercritical fluid extraction is proposed for downstream upgrading of bitumen and refinery vacuum resids to produce upgraded hydrocarbon liquids. This part of the study concentrates on the development of a process to upgrade heavy oil such as bitumen and identify the process conditions (e.g., solvent-to-feed ratio, temperature and pressure) for various stages of the process. A process flow diagram was developed for upgrading the bitumen-derived heavy oil by supercritical fluid extraction using propane as the solvent. The process flow diagram is presented in Figure 4.52. The bitumen is mixed with a known amount of propane and passed through preheaters, where it is brought to the desired supercritical operating conditions and passed through the primary extractor. The primary extractor is provided with internals to thoroughly mix the solvent with the bitumen. The feed and solvent mixture are separated into vapor and liquid streams. The liquid stream, which consists primarily of asphaltenes 326 Figure 4.52 Process Flow Diagram for a Supercritical Fluid Extraction and Separation Process for Upgrading Bitumens Compressor Solvent Recycle 328 and some resins, is transferred to separator IV which operates at 0.1 MPa pressure and optimum temperature (as calculated in the study). The solvent is flashed and recycled and the asphaltenes are separated The vapor phase from the primary extractor flows through a heat exchanger where it is heated and then flashed in separator I. The light ends from the primary extractor are separated into saturates and aromatics in the overhead stream and resins in the bottomstream. The overheads and bottoms from separator I are flashed at ambient pressure subsequently in separators II and III; respectively, to recover the solvent which is recycled. The bottoms from separators II and III are collected as saturates and aromatics and resins; respectively. The solvent separated at separators II, III and IV is compressed and recycled back to the primary extractor. The commercial ROSE[33] process has been tested on various heavy oil feedstocks as discussed in Chapter 2. The published process data have been obtained at the pilot plant level with the primary extractors operated above the critical pressure and below the critical temperature of the solvent in a countercurrent mode. The advantages of supercritical fluids are utilized in the separators located downstream of the primary extractors to recover solvents from the extracted products. The process suggested in this study uses one time solvent-bitumen contact at supercritical conditions for the solvent. The extracted products are separated into solubility class compounds at elevated temperatures. The purpose of this study is to identify suitable process conditions for the supercritical extraction and separation processes along with optimum 329 solvent-to-feed ratios. The knowledge acquired in modeling the supercritical fluid extraction process using the Peng-Robinson equation of state[41] and continuous thermodynamics principle will be used to optimize the process variables. Primary Extractor The bitumens from the AR and Sunnyside oil sands deposits were used to optimize the process conditions required for the primary extractor. The main function of the extractor is to extract lighter hydrocarbons from the feed bitumen using propane as the solvent. Continuous thermodynamics was used to generate the distribution function for the bitumen distribution and the PengRobinson equation of state[41] was used in the flash calculation routines to obtain the composition of the equilibrated phases. The optimization was performed using five solvent-to-feed mole ratios (2.33, 3.0, 4.0, 5.67 and 9.0); pressures varying from 4.14 MPa to 17.3 MPa and temperatures ranging from (339 to 422 K). As indicated, eight quadrature points were used to represent the AR and SS bitumens in the phase equilibrium calculations. The molecular weights, boiling points and specific gravities of the bitumen feedstocks were specified using the procedure outlined previously. The various properties computed at quadrature points for the AR and SS bitumens are presented in Table 4.14 and 4.15; respectively. The specific gravities of the quadrature segments were Table 4.14 Properties of the Asphalt Ridge Bitumen at the Quadrature Points Properties Quad8 Specific Gravity, (288K/288K) Quadl Quad2 Quad3 Quad4 0.85 0.88 0.92 0.95 Quad5 0.98 Quad6 Quad7 1.05 1.12 1.20 Boiling Point, K 495.7 601.2 686.0 747.2 833.2 917.2 1007.6 1129.2 Molecular Weight, (g/gmol) 167.8 260.6 370.1 476.7 680.2 962.9 1399.7 2314.0 Mole fraction 0.15 0.21 0.21 0.19 0.13 0.082 0,04 0.01 Critical Pressure, MPa 2.38 1.61 1.27 1.08 0.83 0.77 0.70 0.58 Critical temperature, K 689.0 Acentric Factor 0.51 780.2 0.77 855.6 0.96 909.6 1.09 979.7 1.27 1065.7 1.36 1157.0 1.45 1274.6 1.56 330 Table 4.15 Properties of the Sunnyside Bitumen at the Quadrature Points Properties Specific Gravity, (288K/288K) Quadl 0.85 Quad2 Quad3 Quad4 Quad5 0.90 0.93 0.98 1.02 Quad6 Quad7 Quad8 1.08 1.14 1.20 Boiling Point, K 509.9 642.9 713.5 802.6 906.4 1007.5 1099.2 1158.3 Molecular Weight, (g/gmol) 178.6 309.9 414.6 599.3 921.0 1398.9 2044.5 2610.5 Mole fraction 0.17 0.21 0.22 0.18 0.11 0.07 0.03 0.01 Critical Pressure, MPa 2.23 1.42 1.16 0.95 0.71 0.59 0.52 0.52 Critical temperature, K 700.5 Acentric Factor 0.24 817.6 0.22 878.5 0.20 959.3 0.19 1045.2 0.17 1138.2 0.16 1225.8 0.15 1293.4 0.15 CO CO 332 normalized so that the computed specific gravities of the mixtures used to represent the bitumens were the same as the measured specific gravities. The critical properties at these quadrature points were estimated using the LeeKesler correlations[43], Flash Calculations were performed using CMGPROP[213], a property prediction program from Computer Maintenance Group, Calgary, Canada. The flash calculations were performed for the AR and SS bitumens at pressures varying from 4.14 MPa to 17.3 MPa and for temperatures of 339, 380 and 422 K. It was observed from the results that a twophase flash was not stable for the AR bitumen at a solvent-to-feed ratio of 2.33 and nor for the SS bitumen at solvent-to-feed ratios of 2.33 and 3.0. Hence, these two solvent-to-feed ratios were not considered for further interpretation. The vapor phase compositions obtained at various combinations of solvent-tobitumen ratios, pressures and temperatures are presented in Figures 4.53 through 4.58. It is observed from the simulation results that as the system pressure increased at constant solvent-to-bitumen ratio and temperature, the mole % of bitumen in the vapor phase for both the AR and SS bitumens increased. Conversely, as the temperature increased at constant solvent-to-bitumen ratio and system pressure, the mole % of bitumen in the vapor phase decreased. This trend is consistent with the experimental data. It is noted from the simulation results that at constant system temperature and pressure, as the solvent-tobitumen ratio increased, the mole % of bitumen transferred into the vapor phase decreased. 333 Figure 4.53 Simulated Extraction Yield Results for the Asphalt Ridge Bitumen at 339 K (Tr=0.92) Mole % of Bitumen Extracted 334 Pressure, MPa 335 Figure 4.54 Simulated Extraction Yield Results for the Asphalt Ridge Bitumen at 380 K (Tr=1.03) Mole % of Bitumen Extracted 336 Pressure, MPa 337 Figure 4.55 Simulated Extraction Yield Results for the Asphalt Ridge Bitumen at 422 K (Tr=1.14) Mole % of Bitumen Extracted 338 Pressure, MPa 339 Figure 4.56 Simulated Extraction Yield Results for the Sunnyside Bitumen at 339 K (Tr=0.92) Mole % of Bitumen Extracted 340 Pressure, MPa 341 Figure 4.57 Simulated Extraction Yield Results for the Sunnyside Bitumen at 380 K (Tr=1.03) Mole % of Bitumen Extracted 342 Pressure, MPa 343 Figure 4.58 Simulated Extraction Yield Results for the Sunnyside Bitumen at 422 K (Tr—1.14) Mole % of Bitumen Extracted 344 Pressure, MPa It is also noted from the modeling plots that the amount of bitumen transferred into the vapor phase was higher at 339 K (Tr=0.92) compared to 390 K (Tr=1.03) and 422 K (Tr=1.14) at all solvent-to-bitumen ratios for both the AR and SS bitumens. The maximum amount of the AR bitumen extracted was 10.7 mole % at a solvent-to-bitumen ratio of 3.0, a temperature of 339 K (Tr=0.92) and a pressure of 11.3 MPa (Pr=2.66). The maximum amount of the SS bitumen extracted was 10.2 mole % at a solvent-to-bitumen ratio of 4.0, a temperature of 339 K (Tr=0.92) and a pressure of 17.3 MPa (Pr=4.1) compared to the 8.7 mole % extracted at a solvent-to-bitumen ratio of 4.0,temperature 339 K (Tr=0.92) and pressure 11.3 MPa (Pr=2.66). The difference in the extraction yield at these conditions is small compared to the capital cost involved in raising the pressure from 11.3 MPa to 17.3 MPa. The optimum conditions for extraction of bitumen using the supercritical fluid extractor (one pass solvent and feed) were found to Asphalt Ridge Sunnyside Solvent/Feed Mole Ratio 3.0 4.0 Temperature K 339 339 Pressure MPa 11.3 17.3 Extract Yield, mol % 10.7 10.2 Separator I The vapor phase from the primary extractor flows to separator I through a heat exchanger where the stream is heated prior to separation into a saturates and aromatics stream and a resins stream. The saturates and aromatics solubility class compounds contain relatively few heteroatomic species and 346 could be integrated into a refinery process scheme without elaborate treatment. Heteroatoms that are detrimental in the catalytic conversion of petroleum are concentrated primarily in the asphaltene and resins solubility fractions. Thus, by separating the feedstream containing saturates, aromatics and resins into two streams that contain predominantly saturates and aromatics and resins, respectively, hydrotreating to remove heteroatoms could be restricted to the resin. The feed stream to the separator was assumed to consist of saturates, aromatics and resins of composition 0.6, 0.3 and 0.1 mole fractions; respectively, for both the AR and SS bitumens. The feedstream to separator I consists predominantly of propane (90 mole %), and the three solubility class compounds constitute only 10 mole % of the mixture. These parameters were obtained while optimizing the parameters for the primary extractor. The saturates, aromatics and resin ensembles for the AR and SS bitumens were fitted by higher order polynomial distribution functions as shown in Figure 4.59 and 4.60; respectively. Eight quadrature points were assumed for each of the saturates, aromatics and resin ensembles. The alkane based molecular weight distribution for the saturates, aromatics and resins ensembles for both of the bitumens were obtained from simulated distillation data. The saturates for both the bitumens were completely characterized using simulated distillation. The aromatics and resins were characterized only up to 973 K which accounts for approximately 40 to 55 wt% of the total material. The molecular weight distribution was extrapolated for the unknown portions of the aromatics and 347 Figure 4.59 Continuous Distributions for the Saturates, Aromatics and Resin Ensembles for the Asphalt Ridge Bitumen Using a Higher Order Polynomial Function Cumulative Weight Fraction 348 Molecular Weight, g/gmol 349 Figure 4.60 Continuous Distributions for the Saturates, Aromatics and Resin Ensembles for the Sunnyside Bitumen Using a Higher Order Polynomial Function Cumulative Weight Fraction 350 M olecular W eight, g/gm ol 351 resins so as to be consistent with the feed bitumen composition. The molecular weights and boiling points at these quadrature points were calculated using the procedure outlined in an earlier section. The specific gravities of the saturates, aromatics and resins fractions were calculated using boiling points and the Watson characteristic factors[214] as: SGj = 1.2164404Tj1/3/Kw where: J, [4.6] is the boiling point of fraction i, K; and; Kw is the Watson characteristic factor. The Watson characteristic factor varies from 12.5 to 12.8 for paraffins and is 9.8 for benzene[214]. However, for this study, the Watson characteristic factor for the saturates fractions that consist of paraffins and naphthenes was assumed to be 11.0 and was taken to be 9.8 for the aromatics and resins fractions. Critical properties at the quadrature points for the saturates, aromatics and resins ensembles for the two bitumens were estimated using the Lee-Kesler correlations[43]. The properties calculated at the quadrature points for the three solubility class ensembles for the AR bitumen are presented in Tables 4.16 through 4.18. Flash calculations were performed for the solubility fractions of both bitumens using the Peng-Robinson equation of state[41] and CMGPROP[213], The optimization calculations were performed at system pressures varying from 8.9 to 17.1 MPa and at temperatures varying from 311 K to 477 K. The mole fractions of the saturates, aromatics and resins in both the vapor and liquid Table 4.16 Properties of the Saturates Fraction at the Quadrature Points for the Asphalt Ridge Bitumen Sat2 Sat3 Sat4 Sat5 Sat6 Sat7 Sat8 Properties Sat1 Specific Gravity, (288 K/288 K) 0.88 0.92 0.94 0.97 0.98 1.00 1.02 1.05 Boiling Point, K 531.54 620.59 659.93 722.87 737.48 780.51 828.26 929.44 Molecular Weight, (g/gmol) 195.3 282.3 330.8 430.9 457.9 547.0 666.4 1012.8 Mole Fraction 0.07 0.10 0.12 0.11 0.10 0.07 0.04 0.01 Critical Pressure, MPa 2.21 1.69 1.53 1.29 1.26 1.13 0.99 0.73 Critical temperature, K 727.24 808.71 843.95 914.96 952.73 0.77 0.85 1.01 1.10 Acentric Factor Sat 0.57 900.8 0.99 993.5 1.19 1073.7 1.39 Quadrature Points for Saturates Ensemble 352 Table 4.17 Properties of the Aromatics Fraction at the Quadrature Points for the Asphalt Ridge Bitumen Properties AR08 AR01 AR02 Specific Gravity, (288 K/288 K) 0.97 Boiling Point, K Molecular Weight, (g/gmol) AR05 AR04 1.07 1.14 1.18 1.21 1.23 1.25 1.26 507.82 682.54 819.54 917.15 991.76 1052.37 1098.87 1126.93 177.05 364.8 642.77 962.5 1684.0 2041.3 2292.6 1310.7 AR06 AR07 AR03 Mole Fraction 0.07 0.07 0.06 0.05 0.03 0.02 0.01 0.01 Critical Pressure, MPa 3.13 2.06 1.51 1.21 1.02 0.88 0.79 0.73 Critical temperature, K 736.27 910.22 1036.79 1188.48 1240.53 1279.72 0.71 0.95 1.24 1.33 1.39 Acentric Factor ARO 0.42 1123.5 1.12 1303.7 1.43 Quadrature Points for Aromatics Ensemble 353 Table 4.18 Properties of the Resins Fraction at the Quadrature Points for the Asphalt Ridge Bitumen Properties Res1 Res2 Specific Gravity, (288 K/288 K) 1.03 1.11 1.16 1.20 1.23 1.24 1.26 1.27 Boiling Point, K 611.48 766.37 876.87 977.76 1037.71 1079.98 1128.54 1155.98 Molecular Weight, (g/gmol) 271.9 515.9 814.8 1584.8 1887.9 2307.8 2585.3 Res3 Res4 1236.9 Res5 Res6 Res7 Res8 Mole Fraction 0.02 0.023 0.021 0.016 0.012 0.009 0.005 0.002 Critical Pressure, MPa 2.42 1.70 2.32 1.04 0.92 0.81 0.73 0.68 Critical temperature, 841.33 K 988.04 1087.27 1174.37 1228.81 1261.82 1303.78 1326.84 1.30 1.38 1.44 1.48 Acentric Factor Res 0.058 0.86 1.06 1.22 Quadrature Points for Resin Ensemble 354 phases were obtained for both bitumens. The term selectivity is used to study the extraction of a single compound or of a group of compounds relative to other compounds or groups of compounds present in a product mixture. The selectivity in extraction of compound a over b in phase i is defined as the ratio of the mole % of a in phase i to the mole percent of b in Phase i. If the mole % of a and b are not equal in the feed, then the selectivity must be normalized. Thus, the selectivity for the distribution of these solubility class compounds in different phases, was defined as: Selectivity of Mole % of a in Phase 1 Mole % o f a in Feed to Separator a over b =------------------------------------------ --------------------- - ---------- [4.7 Mole % of b in Phase 1 / 1 /M o le % of b in Feed to Separator I J The selectivities were calculated for both the AR and SS bitumens for the following options to evaluate the optimum temperature and pressure conditions required to obtain maximum separation of the solubility class compounds: • selectivity of saturates over resins in the vapor phase; • selectivity of aromatics over resins in the vapor phase; • selectivity of saturates and aromatics over resins in the vapor phase; and; • selectivity of resins over saturates and aromatics in the liquid phase. The selectivities for the above mentioned options were calculated for both the bitumens from the modeling results and are presented in Figures 4.61 to 4.68. It is evident from the plots that the selectivities for the following options for 356 Figure 4.61 Predicted Selectivity of Saturates and Aromatics over Resins in the Vapor Phase for the AR Bitumen Selectivity of Saturates and Aromatics Over Resins in Vapor Phase 300 320 340 360 380 400 420 440 Temperature, K 460 480 500 358 Figure 4.62 Predicted Selectivity of Saturates over Resins in the Vapor Phase for the AR Bitumen Selectivity of Saturates Over Resins in Vapor Phase 300 (I) 1a u Ul *>l (O w Ol 320 340 360 380 400 420 440 Temperature, K 460 480 359 500 360 Figure 4.63 Predicted Selectivity of Aromatics over Resins in the Vapor Phase for the AR Bitumen Selectivity of Aromatics Over Resins in Vapor Phase 300 320 340 360 380 400 420 440 Temperature, K 460 480 500 CO cn 362 Figure 4.64 Predicted Selectivity of Resins over Saturates and Aromatics in the Liquid Phase for the AR Bitumen Selectivity of Resins over Saturates and Aromatics Liquid Phase 300 320 340 360 380 400 420 440 Temperature, K 460 480 363 500 364 Figure 4.65 Predicted Selectivity of Saturates and Aromatics over Resins in the Vapor Phase for the Sunnyside Bitumen Selectivity of Saturates and Aromatics Over Resins in Vapor Phase Temperature, K 366 Figure 4.66 Predicted Selectivity of Saturates over Resins in the Vapor Phase for the Sunnyside Bitumen Selectivity of Saturates Over Resins in Vapor Phase Temperature, K o 368 Figure 4,67 Predicted Selectivity of Aromatics over Resins in the Vapor Phase for the Sunnyside Bitumen Selectivity of Aromatics Over Resins in Vapor Phase Temperature, K 369 370 Figure 4.68 Predicted Selectivity of Resins over Saturates and Aromatics in the Liquid Phase for the Sunnyside Bitumen Selectivity of Resins over Saturates and Aromatics Liquid Phase Temperature, K 372 the AR bitumen increased at constant pressure with an increase in temperature up to 470 K and then decreased slightly thereafter: • saturates over resins in vapor phase; • aromatics over resins in vapor phase; • saturates plus aromatics over resins in vapor phase. The decreases in the selectivities with an increase in pressure at constant temperature were more pronounced at higher temperature. The selectivities calculated from the modeling results for the AR resins over saturates and aromatics in the liquid phase decreased with an increase in temperature below 450 K and remained constant at constant pressure. The selectivities of resins over saturates and aromatics in the liquid phase increased with increase, in pressure at constant temperature. The increase in selectivity is more pronounced at higher temperature (~ 460 K). The overall change in the selectivities for the resins over saturates and aromatics is one tenth less than the increase observed for the selectivity of saturates and aromatics over resins in the vapor phase. Similar observations have been made for the individual and combined selectivities calculated from the modeling results for saturates and aromatics individually and combined over resins in the vapor phase for the SS bitumen. However, the selectivities of the saturates and aromatics individually and combined over resins in the vapor phase is 20 % higher at the lowest temperature for the SS bitumen than the corresponding selectivities of the AR bitumen solubility class ensembles. This is due to the difference in the composition of these solubility class compounds of the two bitumens. The 373 predicted selectivities of the resins over saturates and aromatics in the liquid phase for the SS bitumen were different from the AR resin selectivities. The predicted selectivities of the resins in the liquid phase for the SS bitumen decreased with increase in temperature at constant pressure. However, the predicted selectivity of resin in the liquid phase increased with increase in pressure from 10.4 MPa to 12.0 MPa and then decreased at 13.7 MPa and remained constant up to 15.4 MPa and started increasing again. This observation is in contrast to the observation made for the corresponding resin selectivities for the AR bitumen. The predicted selectivities are composition sensitive and the lesser the distance (Figure 4.69) or overlap (in case of differential distribution) between the different compound class ensembles the higher the selectivities and consequently the separation achieved. Flash calculations were conducted using the saturates, aromatics and resins ensembles of the SS bitumen to illustrate this point. A comparison of the molecular weight distributions of the saturates, aromatics and resins for the AR and SS bitumens is presented in Figure 4.69. The saturates ensembles of both the AR and SS bitumens are similar except that the SS saturates are slightly heavier than the AR saturates. The aromatics ensembles for the bitumens are quite different. The AR bitumen aromatics are heavier than the SS bitumen aromatics ensemble. The AR bitumen aromatics ensemble is closer to its resin ensemble than are the corresponding ensembles for SS bitumen. The resin ensemble of the SS bitumen was slightly heavier than the resins ensemble of the AR bitumen. The final molecular weight for the SS 374 Figure 4.69 Comparison of the Molecular Weight Distribution of the Saturates, Aromatics and Resins for the Asphalt Ridge and Sunnyside Bitumens -r ~ fl D . ■ a X ♦ ♦ ♦ t t o t □ ♦ □ t o ARB Saturates □ ARB Aromatics + ARB Resins • SSB Saturates ■ SSB Aromatics ♦ SSB Resins 1 — 500 1500 2000 Molecular Weight, g/gmol 2500 3000 375 1000 “ ““ i -----------------1----------------- r - resins was 100 counts higher than that of the AR resins. When flash calculations were performed using these solubility class compound ensembles for the AR and SS bitumens, the selectivities reflected the difference in the composition as observed above. The predicted individual and combined selectivities of the saturates and aromatics over resins in the vapor phase for the SS bitumen were higher than the selectivities of the corresponding ensembles for the AR bitumen due to the difference in the aromatics and resins distributions. .Consequently, the difference is more pronounced in the resin selectivities in the liquid phase for SS and AR bitumens. It is observed from the predicted selectivity plots that an increase in the pressure does not facilitate the separation of the solubility class compounds. However, an increase in the temperature has a significant influence on the separation of the solubility class ensembles. The commercial ROSE process was designed to operate at higher temperature than the extractor for better separation. The optimum pressure and temperature for separation of the different bitumen solubility class compounds are listed below: Pressure, MPa Temperature, K Asphalt Ridge Bitumen Solubility Class Compounds 10.4 470 Sunnyside Bitumen Solubility Class Compounds 12.0 480 It was concluded from this study that supercritical fluid extraction and separation for representative Uinta Basin bitumens are composition dependent. The greater the difference between solubility class compounds in a cumulative 377 boiling point distribution, the higher the selectivities and separation. The higher the extraction pressure, the higher are the extraction yields and the heavier the compounds extracted. The higher the temperature, the higher the selectivity for the separation of solubility class compounds. The individual and combined saturates and aromatics were preferentially extracted and separated over asphaltenes and resins. experimental data. This agrees with the trends observed from the CHAPTER 5 CONCLUSIONS The supercritical fluid extraction (SFE) apparatus was successfully used to conduct studies with two bitumens from the Asphalt Ridge (AR) and Sunnyside (SS) oil sands deposits of Utah. The existing system[39,40] was modified by switching the backpressure valve located upstream of the extractor to downstream of the extractor between the densitometer and the low pressure separator to achieve better pressure control of the system. A data acquisition system was installed to monitor the flowrate of the solvent gas flowing out of the system and also to measure the density of the extract phase on a continuous basis. A high temperature simulated distillation technique was developed to extend the ASTM D2887 and D5307 methods from 811 K to 973 K for totally elutable and noneluting samples during the chromatographic run. This technique uses a capillary column instead of the packed column suggested by the ASTM methods. Suitable oven, injector and detector programs were developed. The new technique permits the characterization of an additional 20 to 35 wt% of bitumens and bitumen-derived products relative to the conventional ASTM techniques. User friendly software was developed using Visual Basic for reading 379 the sliced and calibration data from the gas chromatograph integrator and converting the data into boiling point and carbon number distributions. The program has good charting capabilities. The setup program facilitates installation of this software on an IBM-PC or equivalent computers under the Microsoft Windows™ environment. The SFE experiments were carried out at five different sets of operating conditions using the Asphalt Ridge and Sunnyside bitumens and the following conclusions were drawn: a) The cumulative extraction yields for both the Asphalt Ridge and Sunnyside bitumens increased with increase in pressure at constant temperature; b) The cumulative extraction yields of the two bitumens decreased with increase in temperature at constant pressure. c) The extraction yields increased with increase in propane solvent density. d) The liquid products obtained from SFE of both the bitumens were upgraded liquids which were approximately 80 wt% volatile. e) Higher molecular weight extract phases were obtained by increasing the system pressure at constant temperature. Lighter and upgraded liquids were obtained by increasing the temperature at constant pressure. The bitumens from four major Utah deposits, Whiterocks, Asphalt Ridge, PR Spring and Sunnyside were subjected to SFE using propane as the solvent. The effect of pressure, temperature, solvent density, and feed compositions on extraction yields and residual fraction characteristics has been investigated. The 380 cumulative extraction yields for the four bitumen increased with increase in pressure at constant temperature and decreased with increase in temperature at constant pressure. In general, higher extraction yields were obtained at higher solvent density for all four bitumens. The cumulative extraction yields decreased with increase in the asphaltene content and were directly proportional to the feed resin content of the feedstock at all five operating conditions. Except for the Asphalt Ridge bitumen, the extraction yields increased with an increase in feed volatility and saturates content of the bitumens. Saturates and aromatics were preferentially extracted compared to asphaltene and resins. This was confirmed by the reduction in the H/C ratio of the residual fraction. The higher the solvent density, the greater the extent of removal of saturated and aromatic compounds during SFE of all four bitumens. Modeling of the supercritical extraction of oil sands bitumen was attempted using continuous thermodynamics along with the Peng-Robinson equation of state. A process flow diagram was developed for upgrading bitumen recovered by the surface mining and aqueous floatation recovery technique. Optimization has been attempted using the modeling procedure to obtain operating conditions such as solvent-to-bitumen ratio, pressure and temperature for supercritical extraction and separation using propane as solvent and the Asphalt Ridge and Sunnyside bitumens as feedstocks. The modeling results predicted preferential extraction of saturates, and aromatics relative to resins were consistent with the experimental observation. APPENDIX A SUPERCRITICAL FLUID EXTRACTION DATA 382 Table A.1 Extraction Yields for Supercritical Fluid Extraction of Uinta Basin Bitumens at 5.6 MPa (Pf=1.2) and 380 K (Tr=1.03) Cumulative Volume of Propane Vented (I, @STP) 0 25 50 75 100 125 150 Whiterocks Bitumen[39] 0.00 3.80 8.30 11.80 14.40 16.70 18.60 Asphalt Ridge Bitumen 0.00 2.30 5.31 7.42 9.57 11.62 13.43 PR Spring Bitumen[40] 0.00 0.20 2.40 5.50 6.50 7.80 8.80 50 g of bitumen sample were initially charged to the extractor Courtesy: SFE data for Whiterocks bitumen by Hwang[39] Sunnyside Bitumen 0.00 0.60 3.36 6.28 8.74 10.72 11.96 383 Table A.2 Extraction Yields for Supercritical Fluid Extraction of Uinta Basin Bitumens at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92) Cumulative Volume of Propane Vented (I, @STP) 0 25 50 75 100 125 150 Whiterocks Bitumen[39] Asphalt Ridge Bitumen PR Spring Bitumen[40] 0.00 2.60 8.70 19.60 27.90 33.30 37.40 0.00 3.71 10.85 17.06 22.45 27.25 31.31 0.00 0.43 6.43 12.85 17.87 20.15 23.04 50 g of bitumen sample were initially charged to the extractor Courtesy: SFE data for Whiterocks bitumen by Hwang[39] Sunnyside Bitumen 0.00 1.44 8.49 13.82 17.46 20.13 22.41 384 Table A.3 Extraction Yields for Supercritical Fluid Extraction of Uinta Basin Bitumens at 10.4 MPa (Pr=2.3) and 380 K (Tr=1.03) Cumulative Volume of Propane Vented (I, @STP) 0 25 50 75 100 125 150 Whiterocks Bitumen[39] 0.00 7.10 15.7 23.10 29.00 33.60 36.80 Asphalt Ridge Bitumen PR Spring Bitumen[40] 0.00 2.24 7.99 13.79 18.13 21.81 24.53 0.00 2.90 7.00 12.20 15.50 19.00 20.80 50 g of bitumen sample were initially charged to the extractor Courtesy: SFE data for Whiterocks bitumen by Hwang[39] Sunnyside Bitumen 0.00 1.81 6.35 9.58 11.83 13.60 14.85 385 Table A.4 Extraction Yields for Supercritical Fluid Extraction of Uinta Basin Bitumens at 10.4 MPa (Pr=2.3) and 422 K (T r=1.14) Cumulative Volume of Propane Vented (I, @STP) 0 25 50 75 100 125 150 Whiterocks Bitumen[39] Asphalt Ridge Bitumen PR Spring Bitumen[40] 0.00 5.80 10.40 14.00 17.00 19.50 22.10 0.00 3.26 7.65 11.31 14.21 16.37 18.18 0.00 1.80 5.10 8.60 11.30 13.60 15.70 50 g of bitumen sample were initially charged to the extractor Courtesy: SFE data for Whiterocks bitumen by Hwang[39] Sunnyside Bitumen 0.00 1.74 5.18 7.52 9.05 10.29 11.21 386 Table A.5 Extraction Yields for Supercritical Fluid Extraction of Uinta Basin Bitumens at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03) Cumulative Volume of Propane Vented (I, @STP) 0 25 50 75 100 125 150 Whiterocks Bitumen[39] 0.00 4.10 14.40 25.30 35.10 40.70 45.00 Asphalt Ridge Bitumen PR Spring Bitumen[40] 0.00 6.77 14.14 20.73 25.55 28.81 31.41 0.00 5.46 14.08 21.74 25.75 29.09 31.70 50 g of bitumen sample were initially charged to the extractor Courtesy: SFE data for Whiterocks bitumen by Hwang[39] Sunnyside Bitumen 0.00 3.73 10.81 16.01 19.48 21.93 23.68 387 Table A.6 Experimental Reproducibility Results for Supercritical Fluid Extraction of Asphalt Ridge and Sunnyside Bitumens at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92) Cumulative Volume of Propane Vented (I, @STP) 0 25 50 75 100 125 150 Asphalt Ridge Bitumen Run #2 0.00 3.71 10.85 17.06 22.45 27.25 31.31 Asphalt Ridge Bitumen Run #11 Sunnyside Bitumen Run #7 0.00 4.59 12.82 20.57 25.46 28.62 31.26 0.00 1.44 8.49 13.82 17.46 20.13 22.41 50 g of bitumen sample were initially charged to the extractor Sunnyside Bitumen Run #12 0.00 1.29 9.10 14.80 18.85 21.84 24.24 APPENDIX B SIMULATED DISTILLATION DATA 389 Table B.1 Cumulative Carbon Number and Boiling Point Distribution for Uinta Basin Bitumens (in Cumulative Weight Fractions) Carbon Number 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Whiterocks Bitumen 0.002 0.003 0.005 0.016 0.032 0.053 0.079 0.107 0.135 0.163 0.208 0.260 0.296 0.311 0.368 0.392 0.428 0.458 0.481 0.502 0.522 0.541 0.556 0.574 0.592 0.607 0.621 0.637 0.650 0.662 0.679 0.688 0.699 0.711 0.720 0.731 0.741 0.749 0.758 0.769 0.779 0.787 Asphalt Ridge Bitumen 0.003 0.017 0.037 0.065 0.098 0.131 0.162 0.194 0.226 0.274 0.331 0.367 0.401 0.448 0.481 0.506 0.528 0.547 0.568 0.588 0.607 0.624 0.641 0.659 0.675 0.691 0.708 0.724 0.741 0.756 0.771 0.787 0.804 0.812 0.821 0.833 0.842 0.851 0.865 0.875 0.891 0.901 PR Spring Bitumen 0.001 0.002 0.005 0.018 0.039 0.061 0.086 0.111 0.138 0.169 0.215 0.266 0.304 0.320 0.368 0.387 0.418 0.446 0.469 0.49 0.509 0.528 0.543 0.56 0.578 0.593 0.607 0.622 0.636 0.648 0.664 0.673 0.685 0.696 0.706 0.718 0.729 0.739 0.750 0.763 0.776 0.781 Sunnyside Bitumen 0.002 0.009 0.019 0.036 0.059 0.084 0.109 0.136 0.167 0.210 0.250 0.280 0.306 0.336 0.360 0.383 0.403 0.421 0.439 0.457 0.473 0.489 0.505 0.521 0.536 0.551 0.566 0.580 0.595 0.607 0.621 0.635 0.650 0.658 0.665 0.676 0.684 0.692 0.705 0.714 0.728 0.734 390 Table B.2 Cumulative Carbon Number and Boiling Point Distribution of Solubility Fractions (Cumulative Weight Fractions) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Whiterocks Bitumen Saturates Aromatics Resins 0.000 0.000 0.000 0.001 0.000 0.001 0.001 0.001 0.000 0.002 0.002 0.000 0.005 0.004 0.004 0.006 0.018 0.009 0.022 0.010 0.068 0.032 0.017 0.104 0.041 0.027 0.142 0.056 0.041 0.182 0.073 0.223 0.057 0.074 0.280 0.093 0.115 0.092 0.367 0.137 0.458 0.109 0.159 0.126 0.532 0.616 0.185 0.144 0.693 0.211 0.162 0.236 0.753 0.180 0.199 0.784 0.262 0.285 0.812 0.216 0.835 0.310 0.234 0.853 0.333 0.250 0.355 0.871 0.266 0.886 0.377 0.282 0.898 0.398 0.297 0.422 0.909 0.315 0.441 0.920 0.329 0.463 0.929 0.344 0.481 0.937 0.358 0.943 0.498 0.370 0.950 0.517 0.385 0.957 0.396 0.532 0.963 0.552 0.411 0.968 0.569 0.425 0.973 0.585 0.438 0.601 0.453 0.978 0.982 0.616 0.466 0.986 0.633 0.481 0.989 0.648 0.494 0.992 0.664 0.509 0.994 0.680 0.525 0.697 0.996 0.544 0.714 0.998 0.563 Asphalt Ridge Bitumen Saturates Aromatics Resins 0.000 0.000 0.000 0.000 0.002 0.001 0.003 0.001 0.000 0.002 0.000 0.007 0.003 0.005 0.020 0.033 0.004 0.019 0.050 0.087 0.008 0.133 0.062 0.013 0.179 0.073 0.019 0.229 0.086 0.030 0.279 0.098 0.047 0.344 0.111 0.062 0.126 0.431 0.078 0.140 0.528 0.095 0.155 0.113 0.609 0.688 0.171 0.132 0.760 0.186 0.150 0.813 0.201 0.168 0.217 0.839 0.187 0.862 0.232 0.204 0.881 0.248 0.221 0.896 0.263 0.236 0.910 0.277 0.251 0.922 0.291 0.266 0.280 0.931 0.305 0.321 0.939 0.297 0.334 0.948 0.310 0.954 0.349 0.325 0.960 0.362 0.339 0.964 0.374 0.351 0.366 0.970 0.388 0.975 0.399 0.377 0.978 0.414 0.392 0.982 0.427 0.406 0.440 0.985 0.419 0.454 0.989 0.434 0.467 0.991 0.447 0.482 0.993 0.462 0.995 0.497 0.475 0.512 0.997 0.491 0.528 0.508 0.998 0.546 1.000 0.528 0.565 0.547 1.001 391 Table B.3 Cumulative Carbon Number and Boiling Point Distribution of Solubility Fractions (Cumulative Weight Fraction) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 PR Spring Bitumen Saturates Aromatics 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.000 0.001 0.002 0.003 0.008 0.026 0.056 0.044 0.093 0.134 0.057 0.179 0.075 0.226 0.103 0.293 0.139 0.398 0.176 0.497 0.214 0.560 0.250 0.286 0.628 0.689 0.319 0.733 0.351 0.758 0.382 0.409 0.785 0.807 0.437 0.826 0.461 0.846 0.486 0.510 0.862 0.876 0.532 0.887 0.558 0.900 0.578 0.910 0.599 0.920 0.618 0.926 0.633 0.934 0.652 0.664 0.942 0.948 0.681 0.955 0.695 0.961 0.708 0.967 0.722 0.972 0.734 0.976 0.747 0.980 0.757 0.984 0.768 0.988 0.779 0.991 0.791 0.803 0.994 Resins 0.000 0.001 0.001 0.002 0.003 0.023 0.029 0.040 0.056 0.074 0.099 0.121 0.144 0.167 0.190 0.213 0.235 0.257 0.279 0.298 0.318 0.335 0.352 0.368 0.383 0.401 0.415 0.430 0.444 0.455 0.469 0.479 0.493 0.505 0.517 0.530 0.541 0.554 0.565 0.578 0.592 0.608 0.624 Sunnyside Bitumen Saturates Aromatics Resins 0.000 0.000 0.000 0.000 0.000 0.001 0.000 0.000 0.001 0.000 0.001 0.002 0.002 0.005 0.003 0.009 0.016 0.004 0.090 0.028 0.007 0.138 0.049 0.011 0.191 0.063 0.018 0.087 0.030 0.245 0.109 0.298 0.040 0.368 0.133 0.051 0.466 0.152 0.067 0.561 0.169 0.084 0.626 0.189 0.101 0.692 0.214 0.119 0.236 0.751 0.136 0.793 0.257 0.154 0.280 0.173 0.817 0.841 0.302 0.190 0.861 0.326 0.208 0.877 0.347 0.224 0.893 0.368 0.240 0.906 0.389 0.256 0.917 0.410 0.271 0.925 0.433 0.289 0.935 0.452 0.303 0.943 0.474 0.318 0.949 0.492 0.333 0.954 0.509 0.345 0.959 0.528 0.361 0.543 0.965 0.372 0.563 0.969 0.388 0.973 0.580 0.403 0.977 0.596 0.417 0.613 0.981 0.433 0.628 0.984 0.447 0.645 0.986 0.463 0.661 0.989 0.477 0.991 0.677 0.493 0.694 0.993 0.510 0.530 0.995 0.711 0.729 0.550 0.996 392 Table B.4 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Sunnyside Bitumen at 5.6 MPa (Pr=1.2) and 380 K (Tr=1.03) (Cumulative Weight Fraction) Carbon Number 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Window #1 0.001 0.001 0.006 0.037 0.089 0.147 0.221 0.300 0.378 0.450 0.526 0.622 0.699 0.746 0.791 0.832 0.859 0.880 0.895 0.908 0.918 0.928 0.935 0.942 0.948 0.953 0.958 0.961 0.964 0.968 0.970 0.973 0.975 0.977 0.979 0.981 0.982 0.983 0.984 0.985 0.986 0.987 Window #2 0.001 0.002 0.007 0.040 0.093 0,161 0.246 0.332 0.411 0.486 0.576 0.668 0.722 0.763 0.808 0.840 0.863 0.882 0.896 0.909 0.918 0.927 0.934 0.940 0.946 0.951 0.955 0.959 0.961 0.965 0.967 0.969 0.972 0.973 0.975 0.977 0.978 0.979 0.980 0.981 0.982 0.983 Window #3 0.001 0.001 0.005 0.031 0.074 0.130 0.203 0.279 0.353 0.425 0.514 0.610 0.670 0.715 0.766 0.804 0.833 0.857 0.875 0.892 0.904 0.916 0.925 0.933 0.941 0.947 0.952 0.957 0.961 0.965 0.967 0.970 0.973 0.975 0.977 0.979 0.980 0.981 0.983 0.984 0.985 0.986 Window #4 0.000 0.001 0.004 0.023 0.057 0.100 0.157 0.220 0.284 0.345 0.413 0.503 0.586 0.641 0.692 0.743 0.782 0.815 0.839 0.861 0.878 0.893 0.906 0.917 0.928 0.936 0.943 0.949 0.954 0.959 0.962 0.966 0.970 0.972 0.975 0.977 0.979 0.981 0.982 0.984 0.985 0.987 Window #5 0.000 0.001 0.003 0.021 0.053 0.095 0.151 0.216 0.284 0.353 0.439 0.546 0.624 0.676 0.732 0.780 0.815 0.843 0.864 0.883 0.897 0.910 0.921 0.929 0.939 0.945 0.950 0.955 0.959 0.963 0.966 0.969 0.971 0.974 0.976 0.977 0.979 0.980 0.982 0.983 0.984 0.985 Window #6 0.000 0.001 0.002 0.016 0.044 0.069 0.100 0.153 0.213 0.278 0.354 0.452 0.555 0.641 0.712 0.776 0.825 0.860 0.885 0.905 0.920 0.932 0.942 0.950 0.958 0.963 0.968 0.972 0.975 0.978 0.980 0.982 0.984 0.985 0.987 0.988 0.989 0.990 0.991 0.991 0.992 0.993 393 Table B.5 Carbon Number and Boiling Point Distribution of Extracts from SFE of Sunnyside Bitumen at 10.4 MPa (Pr=2.3) and 339 K (Tr=0.92) (Cumulative Weight Fraction) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Window #1 0.000 0.001 0.001 0.004 0.018 0.043 0.077 0.126 0.180 0.236 0.289 0.347 0.425 0.498 0.548 0.594 0.642 0.681 0.715 0.741 0.766 0.786 0.804 0.821 0.835 0.852 0.863 0.874 0.884 0.892 0.901 0.907 0.915 0.922 0.927 0.933 0.937 0.942 0.946 0.950 0.953 0.957 0.960 Window #2 0.000 0.001 0.001 0.003 0.015 0.034 0.060 0.096 0.136 0.178 0.222 0.283 0.356 0.405 0.444 0.492 0.533 0.567 0.598 0.624 0.649 0.670 0.691 0.710 0.727 0.746 0.760 0.775 0.788 0.798 0.811 0.819 0.831 0.840 0.848 0.857 0.863 0.871 0.877 0.883 0.889 0.895 0.900 Window #3 0.000 0.000 0.001 0.003 0.013 0.030 0.052 0.083 0.119 0.157 0.200 0.261 0.334 0.384 0.425 0.475 0.518 0.555 0.588 0.616 0.643 0.666 0.688 0.709 0.727 0.747 0.762 0.777 0.790 0.801 0.814 0.823 0.834 0.843 0.852 0.860 0.867 0.874 0.880 0.886 0.892 0.898 0.903 Window #4 0.000 0.000 0.001 0.002 0.010 0.024 0.042 0.068 0.098 0.134 0.177 0.238 0.310 0.359 0.400 0.450 0.492 0.528 0.562 0.589 0.617 0.641 0.665 0.686 0.706 0.728 0.744 0.761 0.776 0.788 0.802 0.812 0.824 0.835 0.844 0.853 0.861 0.869 0.875 0.882 0.888 0.895 0.900 Window #5 0.000 0.000 0.000 0.002 0.012 0.030 0.052 0.082 0.118 0.160 0.205 0.265 0.345 0.407 0.453 0.504 0.552 0.591 0.627 0.656 0.684 0.707 0.729 0.749 0.767 0.787 0.801 0.816 0.829 0.839 0.851 0.859 0.870 0.879 0.887 0.895 0.901 0.908 0.914 0.920 0.925 0.931 0.936 Window #6 0.000 0.000 0.000 0.001 0.008 0.019 0.034 0.054 0.079 0.109 0.148 0.209 0.275 0.319 0.359 0.410 0.451 0.485 0.519 0.546 0.574 0.598 0.621 0.641 0.660 0.680 0.696 0.713 0.727 0.739 0.752 0.762 0.774 0.785 0.793 0.802 0.809 0.818 0.825 0.832 0.838 0.845 0.851 Table B.6 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Sunnyside Bitumen at 10.4 MPa (Pr=2.3) and 380 K (Tr=1.03) (Cumulative Weight Fraction) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Window #1 0.000 0.000 0.001 0.002 0.014 0.038 0.070 0.115 0.166 0.221 0.275 0.338 0.424 0.499 0.550 0.600 0.650 0.690 0.725 0.752 0.777 0.797 0.816 0.833 0.847 0.863 0.874 0.885 0.895 0.902 . 0.911 0.917 0.924 0.930 0.936 0.941 0.945 0.950 0.953 0.957 0.960 0.964 0.967 Window #2 0.000 0.001 0.001 0.003 0.009 0.019 0.030 0.047 0.068 0.095 0.133 0.195 0.265 0.312 0.356 0.413 0.459 0.497 0.533 0.563 0.593 0.619 0.644 0.666 0.686 0.707 0.723 0.741 0.755 0.768 0.782 0.792 0.804 0.814 0.822 0.830 0.837 0.844 0.850 0.856 0.862 0.867 0.873 Window #3 0.000 0.001 0.001 0.002 0.013 0.034 0.060 0.095 0.132 0.174 0.223 0.292 0.376 0.432 0.479 0.536 0.584 0.622 0.657 0.686 0.713 0.737 0.758 0.777 0.794 0.812 0.825 0.839 0.850 0.860 0.871 0.879 0.888 0.896 0.903 0.909 0.915 0.920 0.926 0.930 0.935 0.939 0.943 Window #4 0.000 0.000 0.000 0.001 0.010 0.030 0.056 0.089 0.127 0.169 0.215 0.279 0.364 0.431 0.480 0.535 0.586 0.628 0.665 0.696 0.724 0.747 0.769 0.789 0.806 0.825 0.838 0.852 0.864 0.873 0.883 0.890 0.900 0.907 0.914 0.921 0.926 0.932 0.936 0.941 0.945 0.949 0.953 Window #5 0.000 0.000 0.000 0.001 0.009 0.027 0.052 0.083 0.119 0.159 0.204 0.265 0.348 0.416 0.467 0.524 0.577 0.619 0.659 0.690 0.720 0.746 0.769 0.789 0.807 0.826 0.839 0.854 0.866 0.876 0.886 0.894 0.903 0.911 0.918 0.924 0.929 0.935 0.940 0.944 0.948 0.952 0.956 Window #6 0.000 0.000 0.000 0.001 0.007 0.021 0.040 0.064 0.091 0.124 0.167 0.234 0.311 0.363 0.410 0.467 0.517 0.555 0.593 0.623 0.653 0.677 0.702 0.723 0.741 0.761 0.775 0.790 0.804 0.814 0.826 0.834 0.845 0.853 0.861 0.868 0.874 0.880 0.886 0.891 0.896 0.900 0.905 395 Table B.7 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Sunnyside Bitumen at 10.4 MPa (Pr=2.3) and 422 K (Tr=1.14) (Cumulative Weight Fraction) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Window #1 0.000 0.001 0.001 0.004 0.022 0.052 0.091 0.148 0.213 0.282 0.346 0.414 0.499 0.587 0.650 0.700 0.750 0.792 0.825 0.848 0.869 0.885 0.899 0.911 0.920 0.930 0.937 0.943 0.949 0.953 0.957 0.960 0.964 0.967 0.969 0.972 0.974 0.976 0.978 0.979 0.981 0.982 0.984 Window #2 0.000 0.000 0.001 0.004 0.024 0.056 0.096 0,150 0.209 0.272 0.334 0.404 0.497 0.579 0.634 0.686 0,738 0.777 0.810 0.834 0.856 0.872 0.887 0.900 0.910 0.921 0.928 0.935 0.941 0.945 0.950 0.953 0.957 0.961 0.963 0.966 0.968 0.971 0.972 0.974 0.976 0.978 0.979 Window #3 0.000 0.000 0.001 0.003 0.020 0.051 0.089 0.137 0.189 0.245 0.304 0.371 0.461 0.553 0.618 0.675 0.732 0.779 0.818 0.846 0.872 0.891 0.908 0.921 0.933 0.944 0.951 0.957 0.962 0.966 0.969 0.972 0.975 0.977 0.978 0.980 0.981 0.983 0.984 0.985 0.986 0.987 0.988 Window #4 0.000 0.000 0.001 0.002 0.015 0.042 0.077 0.122 0.172 0.226 0.283 0.352 0.450 0.540 0.603 0.662 0.722 0.768 0.807 0.836 0.862 0.882 0.899 0.914 0.926 0.939 0.946 0.954 0.959 0.963 0.968 0.970 0.974 0.976 0.978 0.980 0.981 0.982 0983 0.984 0.985 0.986 0.987 Window #5 0.000 0.000 0.001 0.002 0.011 0.034 0.066 0.108 0.154 0.205 0.261 0.337 0.438 0.514 0.571 0.635 0.692 0.733 0.771 0.800 0.826 0.847 0.866 0.881 0.895 0.908 0.917 0.926 0.933 0.939 0.945 0.949 0.953 0.957 0.960 0.963 0.965 0.967 0.969 0.971 0.973 0.974 0.975 Window #6 0.000 0.001 0.001 0.002 0.009 0.028 0.057 0.096 0.139 0.187 0.242 0.319 0.420 0.495 0.552 0.616 0.673 0.717 0.756 0.785 0.812 0.834 0.854 0.870 0.884 0.899 0.909 0.918 0.926 0.932 0.938 0.942 0.947 0.951 0.954 0.957 0.960 0.962 0.964 0.966 0.968 0.969 0.971 396 Table B.8 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Sunnyside Bitumen at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03) (Cumulative Weight Fraction) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Window #1 0.000 0.001 0.001 0.005 0.021 0.047 0.076 0.114 0.155 0.199 0.244 0.298 0.374 0.445 0.496 0.545 0.596 0.639 0.677 0.706 0.735 0.757 0.779 0.798 0.815 0.834 0.847 0.861 0.873 0.882 0.892 0.899 0.909 0.916 0.923 0.930 0.935 0.941 0.945 0.950 0.954 0.958 0.962 Window #2 0.000 0.000 0.001 0.002 0.013 0.031 0.054 0.086 0.123 0.164 0.210 0.273 0.348 0.398 0.439 0.491 0.535 0.570 0.604 0.631 0.658 0.682 0.704 0.724 0.742 0.761 0.776 0.791 0.804 0.815 0.827 0.836 0.847 0.857 0.865 0.872 0.879 0.887 0.893 0.899 0.905 0.911 0.916 Window #3 0.000 0.000 0.000 0.001 0.009 0.026 0.048 0.080 0.118 0.159 0.202 0.258 0.333 0.390 0.433 0.482 0.528 0.564 0.599 0.628 0.655 0.680 0.702 0.723 0.741 0.761 0.776 0.792 0.805 0.817 0.830 0.839 0.851 0.861 0.870 0.878 0.885 0.893 0.901 0.907 0.913 0.920 0.925 Window #4 0.000 0.000 0.000 0.001 0.004 0.013 0.027 0.045 0.068 0.095 0.128 0.180 0.237 0.277 0.313 0.358 0.397 0.431 0.464 0.492 0.520 0.544 0.569 0.591 0.612 0.636 0.653 0.671 0.688 0.701 0.717 0.728 0.742 0.753 0.764 0.775 0.784 0.793 0.801 0.809 0.817 0.825 0.833 Window #5 0.000 0.000 0.000 0.000 0.003 0.009 0.019 0.033 0.051 0.073 0.103 0.151 0.206 0.247 0.283 0.330 0.370 0.407 0.442 0.472 0.503 0.529 0.556 0.581 0.604 0.631 0.650 0.671 0.689 0.704 0.722 0.734 0.750 0.763 0.775 0.786 0.796 0.806 0.814 0.822 0.830 0.838 0.845 Window #6 0.000 0.000 0.000 0.000 0.002 0.005 0.012 0.022 0.034 0.052 0.077 0.119 0.170 0.206 0.241 0.285 0.323 0.359 0.393 0.422 0.453 0.480 0.507 0.533 0.557 0.584 0.605 0.626 0.647 0.663 0.683 0.696 0.714 0.729 0.742 0.755 0.766 0.777 0.786 0.795 0.803 0.812 0.820 397 Table B.9 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Asphalt Ridge Bitumen at 5.6 MPa (Pr=1.2) and 380 K (Tr=1.03) (Cumulative Weight Fraction) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Window #1 0.001 0.003 0.005 0.013 0.043 0.078 0.126 0.186 0.248 0.309 0.365 0.420 0.492 0.564 0.618 0.671 0.726 0.767 0.800 0.822 0.842 0.858 0.873 0.886 0.897 0.908 0.916 0.924 0.931 0.936 0.943 0.947 0.953 0.957 0.961 0.964 0.967 0.970 0.973 0.975 0.978 0.980 0.981 Window #2 0.001 0.004 0.005 0.015 0.047 0.082 0.129 0.186 0.252 0.316 0.374 0.431 0.505 0.582 0.642 0.697 0.753 0.796 0.831 0.852 0.871 0.886 0.899 0.911 0.920 0.930 0.936 0.943 0.948 0.953 0.958 0.961 0.966 0.969 0.972 0.974 0.976 0.978 0.980 0.982 0.983 0.985 0.986 Window #3 0.000 0.001 0.001 0.002 0.009 0.016 0.025 0.037 0.051 0.068 0.088 0.123 0.175 0.208 0.240 0.293 0.337 0.373 0.405 0.432 0.461 0.487 0.514 0.539 0.561 0.588 0.608 0.630 0.651 0.668 0.690 0.705 0.724 0.740 0.754 0.768 0.780 0.792 0.802 0.812 0.822 0.831 0.840 Window #4 0.000 0.002 0.003 0.009 0.037 0.065 0.106 0.157 0.218 0.282 0.344 0.408 0.488 0.576 0.646 0.708 0.769 0.817 0.854 0.876 0.895 0.910 0.923 0.934 0.942 0.950 0.956 0.962 0.966 0.969 0.973 0.976 0.979 0.981 0.983 0.984 0.986 0.987 0.988 0.989 0.990 0.991 0.992 Window #5 0.000 0.001 0.002 0.007 0.032 0.066 0.109 0.165 0.227 0.290 0.349 0.419 0.524 0.607 0.654 0.720 0.784 0.823 0.850 0.869 0.888 0.904 0.918 0.929 0.938 0.947 0.953 0.959 0.964 0.967 0.972 0.975 0.978 0.980 0.982 0.984 0.985 0.987 0.988 0.989 0.990 0.991 0.991 Window #6 0.000 0.001 0.001 0.005 0.025 0.055 0.091 0.138 0.191 0.249 0.304 0.363 0.443 0.542 0.607 0.661 0.729 0.788 0.826 0.850 0.871 0.889 0.906 0.920 0.931 0.942 0.949 0.956 0.962 0.966 0.971 0.975 0.979 0.982 0.984 0.986 0.987 0.989 0.990 0.991 0.992 0.993 0.994 398 Table B.10 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Asphalt Ridge Bitumen at 10.4 MPa (Pr=2.3) and 339 K (Tr=0,92) (Cumulative Weight Fraction) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Window #1 0.000 0.003 0.004 0.011 0.041 0.079 0.122 0.170 0.216 0.261 0.306 0.351 0.415 0.484 0.527 0.572 0.628 0.671 0.704 0.727 0.749 0.768 0.787 0.804 0.819 0.835 0.846 0.859 0.869 0.878 0.888 0.896 0.906 0.914 0.921 0.927 0.933 0.939 0.944 0.949 0.953 0.957 0.961 Window #2 0.001 0.004 0.006 0.014 0.045 0.083 0.127 0.175 0.221 0.266 0.311 0.357 0.425 0.492 0.532 0.580 0.637 0.677 0.709 0.731 0.753 0.773 0.791 0.808 0.823 0.838 0.849 0.862 0.872 0.881 0.891 0.899 0.909 0.916 0.923 0.929 0.934 0.940 0.945 0.950 0.954 0.958 0.962 Window #3 0.000 0.002 0.004 0.010 0.038 0.074 0.115 0.160 0.204 0.247 0.290 0.334 0.396 0.464 0.507 0.551 0.606 0.649 0.682 0.706 0.728 0.748 0.767 0.785 0.800 0.816 0.828 0.842 0.852 0.862 0.872 0.881 0.892 0.900 0.908 0.915 0.921 0.927 0.933 0.939 0.944 0.949 0.953 Window #4 0.000 0.003 0.004 0.010 0.035 0.069 0.107 0.151 0.193 0.235 0.277 0.320 0.381 0.450 0.493 0.535 0.589 0.634 0.667 0.691 0.714 0.733 0.753 0.771 0.787 0.805 0.817 0.830 0.842 0.851 0.863 0.871 0.883 0.892 0.900 0.908 0.914 0.921 0.926 0.932 0.938 0.944 0.949 Window #5 0.000 0.002 0.004 0.009 0.033 0.065 0.101 0.143 0.184 0.224 0.265 0.307 0.367 0.435 0.478 0.520 0.573 0.618 0.652 0.676 0.699 0.719 0.739 0.758 0.774 0.792 0.805 0.819 0.832 0.841 0.853 0.862 0.874 0.884 0.892 0.901 0.907 0.915 0.920 0.927 0.933 0.939 0.944 Window #6 0.000 0.000 0.001 0.006 0.028 0.057 0.091 0.130 0.169 0.208 0.250 0.291 0.352 0.417 0.458 0.503 0.557 0.599 0.633 0.657 0.681 0.703 0.723 0.743 0.760 0.778 0.792 0.807 0.819 0.830 0.842 0.852 0.865 0.875 0.884 0.892 0.899 0.907 0.914 0.921 0.927 0.933 0.939 399 Table B.11 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Asphalt Ridge Bitumen at 10.4 MPa (Pr=2.3) and 380 K (Tr=1.03) (Cumulative. Weight Fraction) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Window #1 0.001 0.004 0.007 0.017 0.057 0.104 0.155 0.207 0.256 0.300 0.357 0.405 0.472 0.529 0.563 0.608 0.656 0.689 0.714 0.732 0.750 0.766 0.781 0.795 0.807 0.820 0.830 0.840 0.849 0.856 0.865 0.872 0.880 0.887 0.893 0.899 0.904 0.910 0.914 0.919 0.924 0.929 0.933 Window #2 0.000 0.004 0.006 0.017 0.059 0.113 0.171 0.233 0.291 0.344 0.393 0.444 0.516 0.585 0.623 0.669 0.722 0.757 0.783 0.800 0.817 0.832 0.846 0.858 0.868 0.879 0.886 0.894 0.901 0.906 0.913 0.917 0.923 0.928 0.932 0.936 0.939 0.943 0.945 0.948 0.951 0.954 0.957 Window #3 0.000 0.003 0.005 0.013 0.048 0.086 0.136 0.193 0.254 0.310 0.362 0.413 0.480 0.550 0.603 0.655 0.709 0.752 0.786 0.809 0.829 0.846 0.861 0.874 0.885 0.896 0.904 0.913 0.919 0.925 0.931 0.936 0.941 0.946 0.949 0.953 0.956 0.959 0.961 0.964 0.966 0.969 0.971 Window #4 0.000 0.002 0.004 0.011 0.044 0.073 0.114 0.170 0.232 0.290 0.343 0.397 0.467 0.538 0.591 0.646 0.701 0.744 0.779 0.802 0.823 0.840 0.856 0.870 0.881 0.893 0.902 0.910 0.917 0.923 0.930 0.935 0.941 0.945 0.949 0.952 0.955 0.958 0.961 0.964 0.966 0.968 0.970 Window #5 0.000 0.002 0.003 0.010 0.044 0.089 0.142 0.201 0.258 0.311 0.362 0.421 0.510 0.573 0.613 0.671 0.726 0.760 0.785 0.804 0.823 0.839 0.854 0.867 0.877 0.888 0.896 0.904 0.911 0.916 0.923 0.927 0.932 0.936 0.940 0.943 0.946 0.949 0.951 0.953 0.956 0.958 0.960 Window #6 0.000 0.001 0.002 0.008 0.033 0.062 0.104 0.153 0.208 0.264 0.315 0.368 0.437 0.516 0.574 0.627 0.687 0.738 0.776 0.802 0.825 0.843 0.861 0.876 0.889 0.902 0.911 0.920 0.928 0.934 0.941 0.946 0.952 0.956 0.960 0.964 0.967 0.969 0.972 0.974 0.976 0.978 0.980 400 Table B.12 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Asphalt Ridge Bitumen at 10.4 MPa (Pr=2.3) and 422 K (Tr=1.14) (Cumulative Weight Fraction) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Window #1 0.000 0.003 0.005 0.014 0.047 0.081 0.131 0.191 0.259 0.325 0.384 0.442 0.515 0.594 0.652 0.704 0.757 0.800 0.831 0.850 0.867 0.880 0.892 0.902 0.911 0.919 0.925 0.931 0.935 0.939 0.944 0.947 0.950 0.954 0.956 0.959 0.961 0.963 0.965 0.967 0.969 0.971 0.973 Window #2 0.000 0.002 0.004 0.010 0.038 0.069 0.112 0.166 0.231 0.295 0.353 0.410 0.484 0.564 0.624 0.679 0.737 0.782 0.817 0.839 0.859 0.875 0.889 0.901 0.912 0.921 0.928 0.936 0.941 0.946 0.951 0.954 0.959 0.962 0.965 0.967 0.969 0.972 0.974 0.975 0.977 0.978 0.980 Window #3 0.000 0.002 0.003 0.009 0.040 0.078 0.126 0.186 0.248 0.309 0.365 0.434 0.535 0.599 0.642 0.708 0.762 0.794 0.819 0.836 0.854 0.869 0.882 0.892 0.901 0.909 0.915 0.922 0.926 0.931 0.935 0.938 0.942 0.945 0.947 0.949 0.951 0.953 0.955 0.957 0.958 0.960 0.962 Window #4 0.000 0.001 0.002 0.006 0.031 0.067 0.109 0.162 0.219 0.278 0.333 0.392 0.476 0.570 0.624 0.678 0.745 0.794 0.828 0.849 0.869 0.886 0.901 0.914 0.924 0.934 0.941 0.948 0.953 0.957 0.963 0.966 0.970 0.973 0.975 0.977 0.979 0.981 0.982 0.984 0.985 0.986 0.987 Window #5 0.000 0.001 0.001 0.005 0.024 0.043 0.070 0.107 0.154 0.205 0.258 0.314 0.388 0.472 0.545 0.612 0.677 0.735 0.783 0.815 0.842 0.862 0.880 0.896 0.909 0.922 0.931 0.939 0.946 0.951 0.957 0.961 0.965 0.969 0.972 0.975 0.977 0.980 0.981 0.983 0.984 0.986 0.987 Window #6 0.000 0.000 0.001 0.003 0.019 0.046 0.080 0.126 0.177 0.231 0.285 0.343 0.425 0.523 0.586 0.641 0.712 0.771 0.810 0.836 0.858 0.877 0.895 0.910 0.922 0.933 0.941 0.948 0.954 0.959 0.964 0.968 0.972 0.975 0.977 0.979 0.981 0.983 0.984 0.986 0.987 0.988 0.989 401 Table B.13 Cumulative Carbon Number and Boiling Point Distribution of Extracts from SFE of Asphalt Ridge Bitumen at 17.3 MPa (Pr=4.1) and 380 K (Tr=1.03) (Cumulative Weight Fraction) Carbon Number 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 Boiling Point, K 342 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 973 Window #1 0.000 0.003 0.005 0.013 0.045 0.084 0.128 0.175 0.22 0.262 0.302 0.352 0.428 0.479 0.514 0.567 0.616 0.649 0.675 0.695 0.716 0.734 0.752 0.768 0.782 0.797 0.808 0.82 0.831 0.839 0.85 0.857 0.867 0.875 0.882 0.889 0.895 0.901 0.906 0.912 0.917 0.923 0.928 Window #2 0.000 0.002 0.003 0.008 0.031 0.062 0.098 0.139 0.179 0.219 0.256 0.297 0.358 0.422 0.459 0.502 0.555 0.596 0.628 0.651 0.674 0.694 0.714 0.732 0.748 0.766 0.779 0.793 0.806 0.816 0.829 0.838 0.85 0.86 0.869 0.878 0.886 0.894 0.901 0.908 0.915 0.922 0.928 Window #3 0.000 0.001 0.002 0.006 0.027 0.055 0.089 0.128 0.167 0.205 0.242 0.289 0.362 0.409 0.443 0.497 0.546 0.578 0.605 0.627 0.649 0.67 0.69 0.708 0.724 0.741 0.754 0.769 0.781 0.792 0.805 0.815 0.827 0.837 0.846 0.854 0.862 0.87 0.878 0.885 0.892 0.899 0.906 Window #4 0.000 0.001 0.002 0.006 0.024 0.049 0.078 0.111 0.144 0.176 0.208 0.255 0.322 0.362 0.396 0.45 0.492 0.523 0.549 0.571 0.594 0.615 0.635 0.653 0.669 0.688 0.702 0.717 0.731 0.742 0.757 0.767 0.78 0.79 0.799 0.809 0.817 0.826 0.834 0.842 0.85 0.859 0.867 Window #5 0.000 0.001 0.001 0.004 0.019 0.041 0.069 0.101 0.135 0.168 0.202 0.253 0.325 0.367 0.403 0.461 0.506 0.539 0.567 0.59 0.614 0.636 0.658 0.677 0.694 0.714 0.729 0.745 0.76 0.772 0.787 0.797 0.81 0.821 0.831 0.841 0.849 0.858 0.865 0.873 0.88 0.888 0.895 Window #6 0.000 0.000 0.000 0.002 0.013 0.03 0.051 0.076 0.103 0.13 0.159 0.203 0.267 0.305 0.339 0.393 0.436 0.47 0.499 0.523 0.549 0.573 0.597 0.619 0.639 0.663 0.681 0.7 0.718 0.733 0.752 0.764 0.781 0.794 0.806 0.818 0.828 0.839 0.848 0.857 0.865 0.874 0.882 402 Table B.14 Cumulative Carbon Number and Boiling Point Distribution of Residual Fractions Obtained from SFE of Asphalt Ridge Bitumen (Cumulative Weight Fraction) Carbon Number 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 Boiling Point, K 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 5.6 MPa 380 K 0.000 0.000 0.000 0.000 0.002 0.010 0.017 0.027 0.039 0.054 0.080 0.112 0.133 0.158 0.187 0.21 0.232 0.247 0.266 0.284 0.3 0.317 0.332 0.345 0.357 0.372 0.383 0.393 0.401 0.411 0.42 0.428 0.437 0.445 0.453 0.46 0.466 0.472 0.478 0.484 0.488 10.4 MPa 339 K 0.000 0.000 0.001 0.004 0.009 0.017 0.020 0.024 0.028 0.033 0.040 0.049 0.056 0.064 0.073 0.081 0.09 0.096 0.104 0.112 0.119 0.127 0.135 0.142 0.149 0.158 0.165 0.172 0.177 0.185 0.194 0.201 0.21 0.219 0.23 0.241 0.252 0.262 0.275 0.29 0.304 10.4 MPa 380 K 0.000 0.000 0.000 0.000 0.003 0.014 0.021 0.030 0.039 0.051 0.070 0.095 0.112 0.131 0.155 0.174 0.191 0.204 0.22 0.235 0.248 0.263 0.277 0.289 0.301 0.315 0.327 0.338 0.346 0.358 0.37 0.381 0.393 0.405 0.419 0.432 0.445 0.457 0.473 0.49 0.507 10.4 MPa 422 K 0.000 0.000 0.000 0.000 0.000 0.004 0.007 0.012 0.017 0.024 0.037 0.054 0.067 0.080 0.099 0.113 0.126 0.136 0.149 0.162 0.173 0.186 0.198 0.209 0.220 0.232 0.243 0.254 0.263 0.275 . 0.287 0.299 0.312 0.327 0.344 0.36 0.376 0.393 0.412 0.435 0.46 17.3 MPa 380 K 0.000 0.001 0.001 0.006 0.010 0.014 0.020 0.025 0.031 0.038 0.049 0.060 0.068 0.073 0.089 0.095 0.108 0.121 0.134 0.146 0.157 0.168 0.178 0.189 0.202 0.212 0.223 0.234 0.245 0.255 0.269 0.277 0.288 0.299 0.309 0.321 0.333 0.345 0.356 0.372 0.389 403 Table B.15 Cumulative Carbon Number and Boiling Point Distribution of Residual Fractions Obtained from SFE of Sunnyside Bitumen (Cumulative Weight Fraction) Carton Number 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 Boiling Point, K 399 447 489 527 560 589 617 641 664 669 704 722 739 754 769 782 795 807 818 829 839 848 857 865 873 881 888 895 902 908 914 920 926 931 937 943 948 954 959 964 968 5.6 MPa 380 K 0.000 0.000 0.000 0.001 0.002 0.005 0.008 0.012 0.018 0.026 0.040 0.056 0.069 0.084 0.099 0.115 0.131 0.143 0.158 0.171 0.183 0.197 0.209 0.221 0.231 0.244 0.255 0.266 0.274 0.285 0.297 0.308 0.321 0.334 0.35 0.365 0.381 0.396 0.416 0.438 0.461 10.4 MPa 339 K 0.000 0.000 0.000 0.001 0.002 0.005 0.008 0.011 0.015 0.020 0.027 0.036 0.045 0.054 0.063 0.073 0.084 0.093 0.103 0.113 0.122 0.132 0.142 0.15 0.159 0.169 0.178 0.187 0.194 0.203 0.214 0.224 0.235 0.247 0.261 0.276 0.29 0.304 0.323 0.343 0.364 10.4 MPa 380 K 0.000 0.000 0.000 0.001 0.002 0.005 0.008 0.011 0.015 0.02 0.027 0.036 0.045 0.054 0.063 0.073 0.084 0.093 0.103 0.113 0.122 0.132 0.142 0.15 0.159 0.169 0.178 0.187 0.194 0.203 0.214 0.224 0.235 0.247 0.261 0.276 0.29 0.304 0.323 0.343 0.364 10.4 MPa 422 K 0.000 0.000 0.000 0.000 0.001 0.005 0.009 0.014 0.021 0.031 0.047 0.065 0.081 0.097 0.116 0.133 0.151 0.165 0.181 0.196 0.209 0.224 0.237 0.249 0.26 0.274 0.285 0.295 0.303 0.313 0.325 0.335 0.347 0.358 0.372 0.385 0.398 0.411 0.427 0.445 0.463 17.3 MPa 380 K 0.001 0.001 0.002 0.007 0.011 0.015 0.018 0.023 0.029 0.037 0.048 0.058 0.068 0.074 0.091 0.098 0.112 0.127 0.141 0.155 0.168 0.181 0.191 0.203 0.216 0.227 0.238 0.25 0.26 0.27 0.284 0.292 0.302 0.313 0.323 0.335 0.346 0.357 0.368 0.383 0.399 APPENDIX C ANALYTICAL TEST PROCEDURE FOR SARA ANALYSIS 405 Procedure for Asphaltenes Test Chemicals Required Pentane and toluene of spectrograde. Procedure Heat the bitumen in a water bath until softened. 1. Approximately 3 to 4 g of the bitumen or hydrocarbon sample was placed in a 250 cm3 Erlenmeyer flask. 2. Toluene was added in the ratio of 1 cm3 per 1 g of bitumen. 3. The Erlenmeyer flask was placed in a beaker containing hot water and was shaken occasionally so that the bitumen in the flask was completely mixed with the toluene. 4. The flask was cooled under tap water and 40 cm3 of pentane was added for every 1 cm3 of toluene added and the contents of the flask were thoroughly mixed. 5. The flask was covered with parafilm paper and placed in the dark overnight until complete precipitation of the asphaltenes took place. 6. The fritted glass funnel was weighed. The vacuum filtration unit was set up with a 1 L Buchner flask. 7. The contents of the Erlenmeyer flask were thoroughly mixed. The contents of the flask was transferred to the fritted funnel, the flask was washed several times with pentane until the solution decanted from the flask was colorless. 406 8. The precipitate on the funnel was washed several times with pentane until the wash liquid was colorless. 9. The funnel was completely drained. The funnel and the Erlenmeyer flask were placed in an oven maintained at 383 K and dried it for 1-2 hrs. 10. The funnel was cooled to ambient temperature and weighed The asphaltenes were computed by difference. The calculation was corrected for the asphaltenes that adhered to the walls of the Erlenmeyer flask. 11. The pentane was evaporated from the pentane-maltene solution in a rotoevaporator with the water bath temperature maintained at 323 K and the matlenes thus obtained were used for further fractionation by adsorption chromatography. Fractionation of Maltenes Chemicals Required 1. Pentane, tetrahydrofuran and methanol (all of spectrograde). 2. Fuller’s Earth(FE) of 30 - 60 mesh size and neutral alumina(NA). Procedure 1. The FE was placed in a glass thimble and saturated with pentane. 2. The ratio of FE to maltene solution was 10 to 1. While transferring the maltenes to the FE, care was taken to ensure that all the maltenes were adsorbed on the FE with the result that a clear colorless solution was obtained as a filtrate. 3. The maltenes-FE mixture in the thimble was capped with glass wool, and the thimble was transferred to the Soxhlet extraction unit. 4. The extraction of the maltenes fraction was carried out with a solvent sequence of increasing polarity: pentane, tetrahydrofuran and methanol. Each extraction was carried out for 24 hrs or until the extract from the glass thimble was colorless. 5. The extracts were labeled as saturates and aromatics(pentane soluble), resin-1 (tetrahydrofuran soluble) and resin-2(methanol soluble). In each case the solvent was removed by rotoevaporation. After rotoevoporation, each of the fractions was dried and weighed. 6. The NA was placed in a glass thimble and dried in a oven at 393 K for 24 hours until all the residual water was removed from the neutral alumina. 7. The ratio of NA to saturates and aromatics was 10 to 1. While transferring the saturates and aromatics to the NA, care was taken that all the maltenes were adsorbed on the NA with the result that a clear colorless solution was obtained as filtrate. 8. The saturates and aromatics - NA mixture in the thimble was capped with glass wool, and the thimble was transferred to the Soxhlet extraction unit. 9. The extraction of the saturates and aromatics fraction was carried out with a solvent sequence of increasing polarity: pentane, tetrahydrofuron and methanol. Each extraction was carried out for 24 hrs or until the extract from the glass thimble was colorless. 408 10. The extracts were labeled as saturates(pentane soluble), aromatics1(tetrahydrofuran) and aromatics-2(methanol soluble). In each case the solvent was removed by rotoevaporation. After rotoevoporation, each of the fractions were weighed. The fractionation scheme was proposed by Bukka et al.[44,192]. A schematic of the adsorption chromatography fractionation scheme is presented in Figure C.1. An overall material balance of 96 to 102 wt% was achieved and the results are reported on a normalized basis. 409 Figure C.1 Schematic of the Adsorption Chromatography Technique Bitumen 40 times Pentane Solubles Insolubles Maltenes Asphaltenes Adsorbed on Fullers Earth Pentane Tetrahydrofuran Saturates & Aromatics Resin I Adsorbed on Neutral Alumina Pentane Saturates Tetrahydrofuran Aromatics I Methanol Aromatcs II Methanol Resin II APPENDIX D FIGURES PERTAINING TO MODELING 412 The thermodynamic modeling was performed using Asphalt Ridge and Sunnyside bitumens with their boiling point distributions as a single ensemble. The modeling results and the observations are presented in Chapter 4. As explained in the preceding chapters the bitumens used in this study were fractionated into solubility fractions such as saturates, aromatics, resins and asphaltenes. The boiling point distributions obtained from these solubility fractions was used as separate ensembles in the thermodynamic modeling to study the effect of pressure, temperature and solvent-to-feed ratio on the extraction yields and also on the selectivity of saturates plus aromatics plus resins over asphaltenes in the vapor phase in the supercritical extractor system proposed in line with the procedure outlined in Chapter 4. The purpose of this alternate approach was to use the boiling point distributions of the solubility fractions of bitumen as the starting point for modeling instead of a single bitumen ensemble. This will describe the bitumen system better than the single boiling point ensemble. The results obtained from modeling will help us to interpret the selective extraction of saturates plus aromatics plus resins over asphaltenes in the vapor phase of the extractor. The boiling point distributions obtained from simulated distillation of saturates, aromatics and resins were used along with an assumed boiling point profile for asphaltenes. The asphaltenes are solid at room temperature and it is not possible to subject them to simulated distillation analyses due to the high concentration of nonvolatile components. The four solubility fraction ensembles for the Asphalt Ridge and Sunnyside bitumens were fitted to higher order 413 polynomials. The distribution equations obtained were used in modeling the extractor of the supercritical fluid extraction system as suggested in Chapter 4. The results obtained from modeling of the extractor are presented in this Appendix. 414 Pressure, MPa Figure D. 1 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Asphalt Ridge Bitumen at 339 K (Tr=0.92) Using Separate Solubility Ensembles ! 415 Pressure, MPa Figure D.2 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Asphalt Ridge Bitumen at 380 K (Tr=1.03) Using Separate Solubility Ensembles 416 Pressure, MPa Figure D.3 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Asphalt Ridge Bitumen at 422 K (Tr=1.14) Using Separate Solubility Ensembles 417 Pressure, MPa Figure D.4 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Sunnyside Bitumen at 339 K (Tr=0.92) Using Separate Solubility Ensembles Mole % of Bitumen in Vapor Phase 418 Pressure, MPa Figure D.5 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Sunnyside Bitumen at 380 K (Tr=1.03) Using Separate Solubility Ensembles Mole % of Bitumen in Vapor Phase 419 Pressure, MPa Figure D.6 Effect of Pressure and Solvent-to-Feed Ratio on the Extraction Yields for Sunnyside Bitumen at 422 K (Tr=1.14) Using Separate Solubility Ensembles Selectivity of Saturates, Aromatics and Resins over Asphaltenes 420 Pressure, MPa Figure D.7 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor Phase for Asphalt Ridge Bitumen at 339 K (Tr=0.92) Using Separate Solubility Ensembles Selectivity of Saturates, Aromatics and Resins over Asphaltenes 421 Figure D.8 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor Phase for Asphalt Ridge Bitumen at 380 K (Tr=1.03) Using Separate Solubility Ensembles 203 183 163 143 123 103 83 63 43 23 3 1 Pressure, MPa 19 E Tect of Pressure and Solvent-to-Feed Ratio on the Selectiv ;s ar i Aromatics and Resins over Asphaltenes in Vapor Phas Bitumen at 422 K (Tr=1.14) Using Separate Solubility Ensem 423 Pressure, MPa Figure D.10 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor Phase for Sunnyside Bitumen at 339 K (Tr=0.92) Using Separate Solubility Ensembles 424 Pressure, MPa Figure D.11 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor Phase for Sunnyside Bitumen at 380 K (Tr=1.03) Using Separate Solubility Ensembles 425 Pressure, MPa Figure D.12 Effect of Pressure and Solvent-to-Feed Ratio on the Selectivity of Saturates and Aromatics and Resins over Asphaltenes in Vapor Phase for Sunnyside Bitumen at 422 K (Tr=1.14) Using Separate Solubility Ensembles APPENDIX E SIMULATED DISTILLATION SOFTWARE CODE 427 This program was designed and compiled in Visual Basic professional edition, version 3.0 supplied by Microsoft Corporation, One Microsoft Way, Richmond, Washington. This subroutine closes the start screen and loads the opening screen for estimation of boiling point and carbon number distributions by ASTM D2887 and the extended method as discussed in Chapter 4. Sub d2887_Click () Load d2887open d2887open.Show End Sub This subroutine closes the start screen and loads the opening screen for estimation of boiling point and carbon number distribution by ASTM D5307 and extended method explained in Chapter 4 Sub d5307_Click () Unload start Load d5307open d5307open.Show End Sub This subroutine allows the user to exit from the software Sub ex_Click () End End Sub This subroutine closes the start screen and loads the opening screen for converting three or more column format data files obtained from the HP 3396 series II integrator to one column format suitable for this software. Sub one_Click () Unload start Load convert convert. Show End Sub This subroutine closes the start screen and loads the notepad software which could be used to edit the calibration and output files generated by this software. Sub text_Click () Unload Start Load Frmmdi Frmmdi.Show End Sub 428 Declaring the necessary constants and the Windows API routines for floating pop up menu used in the opening screen, define A-Z Declare Function GetMenu% Lib "User" (ByVal hWnd%) Declare Function GetSubMenu% Lib "User" (ByVal hMenu%, ByVal nPos%) Declare Function TrackPopupMenu% Lib "User" (ByVal hMenu%, ByVal wFlags%, ByVal X%, ByVal Y%, ByVal nReserved%, ByVal hWnd%, ByVal lpReserved&) This subroutine is activated when the left or the right mouse button is down and the pop up menu displayed on the screen to choose the above mentioned five options. Sub Form_MouseDown (Button As Integer, Shift As Integer, X As Single, Y As Single) PopupX = (X + Left) / screen.TwipsPerPixelX PopupY = (Y + Top) I screen.TwipsPerPixelY hMenu = GetMenu(hWnd) hSubMenu = GetSubMenu(hMenu, 0) I = TrackPopupMenu(hSubMenu, 0, PopupX, PopupY, 0, hWnd, 0) End Sub Subroutine to open a text file using the notepad. Sub FOpenProc () Dim RetVal On Error Resume Next Dim OpenFileName As String frmMDI.CMDialogl.Filename = "" frmMDI.CMDialogl.Action = 1 If Err <> 32755 Then 'user pressed cancel OpenFileName = frmMDI.CMDialogl Filename OpenFile (OpenFileName) UpdateFileMenu (OpenFileName) End If End Sub Subroutine to get the name of the text file. Function GetFileName () 'Displays a Save As dialog and returns a file name 'or an empty string if the user cancels On Error Resume Next frmMDI.CMDialogl.Filename = "" frmMDI.CMDialogl.Action = 2 If Err <> 32755 Then 'User cancelled dialog GetFileName = frmMDI.CMDialogl Filename Else 429 GetFileName = "" End If End Function Subroutine to display name and path of the recently used four files Function OnRecentFilesList (Filename) As Integer Dim i For i = 1 To 4 If frmMDI.mnuRecentFile(i).Caption = Filename Then OnRecentFilesList = True Exit Function End If Next i OnRecentFilesList = False End Function Subroutine to open a file in the notepad Sub OpenFile (Filename) Dim NL, Textln, GetLine Dim flndex As Integer NL = Chr$(13) + Chr$(10) On Error Resume Next ' open the selected file Open Filename For Input As #1 If Err Then MsgBox "Can't open file :" + Filename Exit Sub End If ' change mousepointer to an hourglass screen.MousePointer = 11 ' change form's caption and display new text flndex = FindFreelndex() document(flndex).Tag = flndex document(flndex). Caption = UCase$(Filename) document(flndex).Text1.Text = lnput$(LOF(1), 1) FState(flndex).Dirty = False document(flndex).Show Close #1 ' reset mouse pointer screen.MousePointer = 0 End Sub Subroutine for saving a file in the notepad Sub SaveFileAs (Filename) On Error Resume Next 430 Dim Contents As String ' open the file Open Filename For Output As #1 ' put contents of the notepad into a variable Contents = frmMDI.ActiveForm.Text1.Text ' display hourglass screen. MousePointer = 11 ' write variable contents to saved file Print #1, Contents Close #1 ' reset the mousepointer screen. MousePointer = 0 ' set the Notepad's caption If Err Then MsgBox Error, 48, App.Title Else frmMDI.ActiveForm.Caption = UCase$(Filename) ' reset the dirty flag FState(frmMDI.ActiveForm.Tag).Dirty = False End If End Sub Subroutine to update the FileMenu in the notepad Sub UpdateFileMenu (Filename) Dim RetVal ' Check if OpenFileName is already on MRU list. RetVal = OnRecentFilesList(Filename) If Not RetVal Then ' Write OpenFileName to MDINOTEPAD.INI WriteRecentFiles (Filename) End If ' Update menus for most recent file list. GetRecentFiles End Sub Subroutine to load the opening screen for the notepad. Sub Form_Load () Dim I As Integer mnuFontName(O).Caption = screen. Fonts(O) For I = 1 To screen.FontCount -1 Load mnuFontName(l) mnuFontName(O).Caption = screen. Fonts( I) Next End Sub 431 Subroutine to unload the opening screen of the notepad. Sub Form_QueryUnload (Cancel As Integer, UnloadMode As Integer) Dim Msg, Filename, NL Dim Response As Integer If FState(Me.Tag).Dirty Then Filename = Me.Caption NL = Chr$(10) & Chr$(13) Msg = "The text in [" & Filename & "] has changed." Msg = Msg & NL Msg = Msg & "Do you want to save the changes?" Response = MsgBox(Msg, 51, frmMDI.Caption) Select Case Response ' User selects Yes Case 6 'Get the filename to save the file Filename = GetFileName() 'If the user did notspecify a file name, 'cancel the unload; otherwise, save it. If Filename = "" Then Cancel = True Else SaveFileAs (Filename) End If ' User selects No ' Ok to unload Case 7 Cancel = False ' User selects Cancel ' Cancel the unload Case 2 Cancel = True End Select End If End Sub Subroutine to resize the opening screen of the notepad. Sub Form_Resize () If windowstate <> 1 And ScaleHeight <> 0 Then TextlVisible = False T e x tl Height = ScaleHeight Textl Width = ScaleWidth TextlVisible = True End If End Sub 432 Subrouine for unloading the opening screen of the notepad Sub Form_Unload (Cancel As Integer) FState(Me.Tag). Deleted = True 'Hide toolbar edit buttons if no notepad windows If Not AnyPadsLeftQ Then frmMDIIimgCutButton.Visible = False frmMDIIimgCopyButton. Visible = False frmMDIIimgPasteButton. Visible = False End If End Sub Subroutine for copying text in the notepad Sub mnuECopy_Click () EditCopyProc End Sub Subroutine for cutting text in the notepad Sub mnuECut_Click () EditCutProc End Sub Subroutine to delete a block of text in the notepad Sub mnuEDelete_Click () ' If cursor is not at the end of the notepad. If screen.ActiveControl.SelStart <> Len(screen.ActiveControl.Text) Then ' If nothing is selected, extend selection by one. If screen.ActiveControl.SelLength = 0 Then screen. ActiveControl.SelLength = 1 ' If cursor is on a blank line, extend selection by two. If Asc(screen.Acti veControl.SelText) = 13 Then screen.ActiveControl.SelLength = 2 End If End If ' Delete selected text, screen. ActiveControl.SelText ="" End If End Sub Subroutine for pasting copied or cut text in the notepad Sub mnuEPaste_Click () EditPasteProc End Sub 433 Subroutine to select all the text in the text in the notepad Sub mnuESelectAII_Click () frmMDI.ActiveForm.Textl.SelStart = 0 frmMDI.ActiveForm.Textl .SelLength =Len(frmMDI.ActiveForm.Text1 .Text) End Sub Subroutine to display the time on the notepad status bar. Sub mnuETime_Click () Dim TimeStr As String, DateStr As String Textl.SelText = Now End Sub Subroutine for closing a file in the notepad Sub mnuFCIose_Click () Unload Me End Sub Subroutine for unloading the text file windows in the notepad Sub mnuFExit_Click () ' Unloading the MDI form invokes the QueryUnload event ' for each child form, then the MDI form - before unloading ' the MDI form. Setting the Cancel argument to True in any of the ' QueryUnload events aborts the unload. Unload frmMDI End Sub Subroutine for opening a text new text file window in the notepad Sub mnuFNew_Click () FileNew End Sub Subroutine for changing the text font in the notepad Sub mnuFontName_Click (index As Integer) Textl.FontName = mnuFontName(index). Caption End Sub Subroutine for saving the file with already existing filename in the notepad Sub mnuFSave_Click () Dim Filename As String If Left(Me.Caption, 8) = "Untitled" Then ' The file hasn't been saved yet, ' get the filename, then call the ' save procedure Filename = GetFileName() 434 Else ' The caption contains the name of the open file Filename = Me.Caption End If ' call the save procedure, if Filename = Empty then ' the user selected Cancel in the Save As dialog, otherwise ' save the file If Filename <> "" Then SaveFileAs Filename End If End Sub Subroutine for saving text file using user input filename in notepad Sub mnuFSaveAs_Click () Dim SaveFileName As String SaveFileName = GetFileNameQ If SaveFileName <> "" Then SaveFileAs (SaveFileName) ' Update the recent files menu UpdateFileMenu (SaveFileName) End Sub Subroutine for activating option menu “option” in notepad Sub mnuOptions_Click () mnuOToolbar.Checked = frmMDIIpicToolbar. Visible End Sub Sub mnuOToolbar_Click () OptionsToolbarProc Me End Sub Subroutineto display most recent files used in notepad Sub mnuRecentFile_Click (index As Integer) OpenFile (mnuRecentFile(index).Caption) ' Update recent files list for new notepad. GetRecentFiles End Sub Subroutine for activating find option in notepad Sub mnuSFind_Click () If Me!Text1 .SelText <> "" Then frmFind!Text1 Text = Me!Text1 SelText Else frmFind!Text1.Text = FindString End If gFirstTime = True 435 frmFind.Show End Sub Subroutine to continue search for the next user input search word in notepad Sub mnuSFindNext_Click () If Len(gFindString) > 0 Then Findlt Else mnuSFind_Click End If End Sub Subroutine to arrange mdi windows in notepad Sub mnuWArrange_Click () frmMDI.Arrange ARRANGEJCONS End Sub Subroutine to arrange mdi windows in cascade mode in notepad Sub mnuWCascade_Click () frmMDI.Arrange CASCADE End Sub Subroutine to arrange mdi windows in tile mode in notepad Sub mnuWTile_Click () frmMDI.Arrange TILE_HORIZONTAL End Sub Subroutine to receive user input for search option in notepad Sub Text1_Change () FState(Me.Tag).Dirty = True End Sub Subroutine to focus the user input text box in notepad Sub Text1_GotFocus () If frmFind.Visible Then frmFind.ZOrder 0 End If End Sub Declaration of constants for the notepad Option Explicit Global Const modal = 1 Global Const CASCADE = 0 Global Const TILE HORIZONTAL = 1 436 Global Const TILE_VERTICAL = 2 Global Const ARRANGEJCONS = 3 Type FormState Deleted As Integer Dirty As Integer Color As Long End Type Global FStateQ As FormState Global Document() As New frmNotePad Global gFindString, gFindCase As Integer, gFindDirection As Integer Global gCurPos As Integer, gFirstTime As Integer Global ArrayNum As Integer ' API functions used to read and write to MDINOTE.INI. ' Used for handling the recent files list. Declare Function GetPrivateProfileString Lib "Kernel" (ByVal IpApplicationName As String, ByVal IpKeyName As String, ByVal IpDefault As String, ByVal IpReturnedString As String, ByVal nSize As Integer, ByVal IpFileName As String) As Integer Declare Function WritePrivateProfileString Lib "Kernel" (ByVal IpApplicationName As String, ByVal IpKeyName As String, ByVal IpString As String, ByVal IplFileName As String) As Integer 'Declaration of constants strings and variables for the SIMDIS part of the ‘program Global blankr$ Global samplers Global outputrS Global calibr$ Global solpeaktime Global blankr1$ Global blankr2$ Global sampler1$ Global isr$ Global outputr1$ Global calibr1$ Global first Global last Global iswe Global sampwt Global solpeaktimel ‘Subroutine to estimate carbon number and boiling point distribution according to ‘ASTM D2887 Method Sub astmd2887_Click () 437 Static a(2500), b(2500), X(2500), t(2500), cn(2500), rt(2500), bp(2500), ctime(3), s(2500), c0(50), ca(50), co(50), cum(50), e(50) Open blankr$ For Input As #1 Open samplers For Input As #2 Open outputr$ For Append As #3 Open calibrS For Input As #4 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k=0 While Not EOF(1) k=k+1 Input #1, a(k) Wend k=0 While Not EOF(2) k=k+1 Input #2, b(k) Wend For I = 1 To k s(l) = b(l) - a(l) Next I w = 0 p = solpeaktime *30 + 1 v=k For I = p To v w = w + s(l) Next I For I = 1 To 20 ca(0) = 0 cum(0) = 0 d =0 co(0)= 0 X(l) = (rt(l) *30 + 7) For u = p To X(l) d = d + s(u) Next u ca(l) = d co(l) = ca(l) - ca(l -1 ) e(l) = co(l) / w cum(l) = cum(l -1 ) + e(l) Next I Print #3, 438 Print #3,"" Print #3 Print #3, Tab(50); Date$ Print #3, Tab(50); Time$ Print #3,"" Print #3,"" Print #3,"" Print#3, "Total Area-'; w Print #3,"" Print #3,"" Print #3,"" Print #3, "Blank Filename blankr$ Print #3, "sample Filename ", samplers Print #3, "Calibration Filename ", calibr$ Print #3, "Output Filename ", outputr$ Print #3,"" Print #3, "" Print #3, "Carbon", "Boiling", "Weight", "Cum. Weight" Print #3, "Number", "Point", "Fraction", "Fraction" Print # 3 , "Deg.F" Print #3,"" Print #3,"" For I = 1 To 20 cn1$ = Format$(cn(l), "###") bp1$ = Format$(bp(l), "####") e1$ = Format$(e(l), "##.###") cum1$ = Format$(cum(l), "##.###") Print #3, cn1$, bp1$, e1$, cum1$ Next I Close Unload d2887open Load d2887res d2887res.Show d2887res.d2887resresults.FontSize = 13 d2887res.d2887resresults.Print," ASTM D2887 Calculation Results" d2887res.d2887resresults.Print,"" d2887res.d2887resresults.FontSize = 7 d2887res.d2887resresults.Print , "Carbon", "Boiling", "Weight", "Cum. Weight" d2887res.d2887resresults.Print, "Number", "Point", "Fraction", "Fraction" d2887res.d2887resresults.Print, "Deg.F" d2887res.d2887resresults. P rint,"" For I = 1 To 20 cn2$ = Format$(cn(l), "###") bp2$ = Format$(bp(l), "####") 439 e2$ = Format$(e(l), M ##.###,,) cum2$ = Format$(cum(l), ”##.###") d2887res.d2887resresults.Print, cn2$, bp2$, e2$, cum2$ Next I d2887open.astmd2887.Enabled = -1 d2887open.astmd2887ex.Enabled = -1 d2887open.d2887cuts. Enabled = -1 d2887open.d2887plotchro. Enabled = -1 d2887open.light.Enabled = -1 d2887open.cal.Enabled = -1 d2887open.diffplotd2887. Enabled = -1 End Sub ‘Subroutine to estimate carbon number and boiling point distribution according to ‘extended method proposed in this dissertation Sub astmd2887ex_Click () Static a(2500), b(2500), X(2500), t(2500), cn(2500), rt(2500), bp(2500), ctime(3), s(2500), c0(50), ca(50), co(50), cum(50), e(50) Open blankr$ For Input As #1 Open samplers For Input As #2 Open outputr$ For Append As #3 Open calibr$ For Input As #4 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k2 = k1 k=0 While Not EOF(1) k=k+1 Input #1, a(k) Wend k=0 While Not EOF(2) k=k+1 Input #2, b(k) Wend For I = 1 To k s(l) = b(l) - a(l) Next I w=0 p = solpeaktime *30 + 1 v=k For I = p To v 440 w = w + s(l) Next I For I = 1 To k2 ca(0)= 0 cum(0) = 0 d=0 co(0) = 0 X(l) = (rt(l) *30 + 7) For u = p To X(l) d = d + s(u) Next u ca(l) = d co(l) = ca(l) - ca(l -1) e(l) = co(l) / w cum(l) = cum(l -1 ) + e(l) Next I Pr nt #3,"" Pr nt #3,"" Pr nt #3,"" Pr nt #3, Tab(50); Date$ Pr nt #3, Tab(50); Time$ Pr nt #3,"" Pr nt #3, "" Pr nt #3,"" Pr nt #3, "Total Area="; w Pr nt #3,"" Pr nt #3,"" Pr nt #3,"" Pr nt #3, "Blank Filename blankr$ Pr nt #3, "sample Filename sampler$ Pr nt#3, "Calibration Filename ", calibr$ Pr nt #3, "Output Filename ", outputr$ Pr nt #3,"" Pr nt #3,"" Pr nt #3, "Carbon", "Boiling", "Weight", "Cum. Weight Pr nt #3, "Number", "Point", "Fraction", "Fraction" Pr nt #3,"", "Deg.F" Pr nt #3,"" Pr nt #3,"" For I = 1 To k2 cn1$ = Format$(cn(l), "###") bp1$ = Format$(bp(l), "####") e1$ = Format$(e(l), "##.###") cum1$ = Format$(cum(l), "##.###") Print #3, cn1$, bp1$, e1$, cum1$ 441 Next I Close Unload d2887open Load d2887res d2887res.Show d2887res.d2887resresults.FontSize = 13 d2887res.d2887resresults.Print, " ASTM D2887 Calculation Results" d2887res.d2887resresults.Print,"" d2887res.d2887resresults.FontSize = 7 d2887res.d2887resresu!ts.Print , "Carbon", "Boiling", "Weight", "Cum. Weight" d2887res.d2887resresults.Print, "Number", "Point", "Fraction", "Fraction" d2887res.c'°S87resresults.Print, "Deg.F" d2887res __387resresults.Print, "" For I = 1 To k2 Step 2 cn2$ = Format$(cn(l), "###") bp2$ = Format$(bp(l), "####") e2$ = Format$(e(l), "##.###") cum2$ = Format$(cum(l), "##.###") d2887res.d2887resresults.Print, cn2$, bp2$, e2$, cum2$ Next I d2887open.astmd2887. Enabled = -1 d2887open.astmd2887ex. Enabled = -1 d2887open.d2887cuts.Enabled = -1 d2887open.d2887plotchro.Enabled = -1 d2887operdight.Enabled = -1 d2887open.cal. Enabled = -1 d2887open.diffplotd2887. Enabled = -1 End Sub ‘Subroutine to read the calibration data and plot a chart between carbon number ‘and retention time for the ASTM D2887 method. Sub cal_Click () Static a(2500), b(2500), X(2500), t(2500), cn(2500), rt(50), bp(50), ctime(3), s(2500), c0(50), ca(50), co(50), cum(50), e(50) Open calibr$ For Input As #4 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k2 = k1 -1 Unload d2887open Load d2887chro d2887chro.Show 442 d2887chro. Graph. FontSize = 150 d2887chro.Graph.FontStyle = 6 d2887chro. Graph. LeftTitle = "Retention Time" d2887chro.Graph.BottomTitle = "Boiling Point, Deg.F" d2887chro. Graph. GraphTitle = "Calibration" d2887chro.Graph.NumPoints = k2 d2887chro.Graph.NumSets = 1 d2887chro.Graph.Autolnc = 0 d2887chro.Graph.ThisSet = 1 For j = 1 To k2 d2887chro.Graph.ThisPoint = j d2887chro.Graph.XPosData = bp(j) Next j For j = 1 To k2 d2887chro.Graph.ThisPoint = j d2887chro.Graph.GraphData = rt(j) Next j d2887chro. Graph. YAxisStyle = 2 d2887chro.Graph.YAxisMax = 45 d2887chro.Graph.YAxisMin = 0 d2887chro.Graph.DrawMode = 2 Close d2887chro. reset. Visible = False End Sub ‘Subroutine to estimate the distillation cuts using ASTM D2887 method. Sub d2887cuts_Click () Static a(2500), b(2500), X(2500), t(2500), cn(2500), rt(2500), bp(2500), ctime(3), s(2500), c0(50), ca(50), co(50), cum(50), e(50) Open blankr$ For Input As #1 Open sampler$ For Input As #2 Open outputr$ For Append As #3 Open calibr$ For Input As #4 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k2 = k1 k=0 While Not EOF(1) k=k+1 . Input #1, a(k) Wend k=0 443 While Not E0F(2) k=k+1 Input #2, b(k) Wend For I = 1 To k s(l) = b(l) - a(l) Next I p = solpeaktime *30 + 1 v=k w =0 For I = p To v w = w + s(l) Next I ctime(1) = u rt(4) - rt(3)) / (421 - 345) *65) + rt(3) ctime(2) = rt(8) ctime(3) = ((rt(20) - rt(19)) / (1013 - 993) * 7) + rt(19) For I = 1 To 3 ca(0) = 0 cum(0) = 0 d=0 co(0) = 0 X(l) = (ctime(l) *30 + 7) For u = p To X(l) d = d + s(u) Next u ca(l) = d co(l) = ca(l) - ca(l -1) e(l) = co(l) / w cum(l) = cum(l -1) + e(l) Next I Print #3,"" Print #3,"" Print #3,"" Print #3, Tab(50); Date$ Print #3, Tab(50); Time$ Print #3,"" Print #3, "" Print #3,"" Print#3, "Total Area-'; w Print #3,"" Print #3,"" Print #3,"" Print #3, "Blank Filename blankrS Print #3, "sample Filename ”, samplers ", calibr$ Print #3, "Calibration Filename 444 Print #3, "Output Filename ", outputr$ Print #3,"" Print #3,"" Print # 3 ,"" Print #3,"" cum1$ = Format$(e(1), "#.###") cum2$ = Format$(e(2), "#.###") cum3$ = Format$(e(3), "#.###") cum4$ = Format$((1 - cum(3)), "#.###") Print #3, "Gasoline =", Tab(35); cum1$ Print #3, "Middle Distillates =", Tab(35); cum2$ Print #3, "Gas Oil =", Tab(35); cum3$ Print #3, "Residue =", Tab(35); cum4$ Close Unload d2887open Load d2887res d2887res.Show d2887res.d2887resresults.FontSize = 13 d2887res.d2887resresults.Print," Distillation Cuts (by ASTM D2887)" d2887res.d2887resresults. Print,"" d2887res.d2887resresults.FontSize = 8.25 d2887res.d2887resresults.Print,"" d2887res.d2887resresults.Print, "Gasoline =", cum1$ d2887res.d2887resresults. P rint, "" d2887res.d2887resresults.Print, "Middle Distillates =", cum2$ d2887res.d2887resresults. Print, "" d2887res.d2887resresults.Print, "Gas Oil =", cum3$ d2887res.d2887resresults.Print, "" d2887res.d2887resresults.Print, "Residue =", cum4$ d2887open.astmd2887. Enabled = -1 d2887open.astmd2887ex. Enabled = -1 d2887open.d2887cuts. Enabled = -1 d2887open.d2887displot.Enabled = -1 d2887open.d2887plotchro. Enabled = -1 d2887open.light.Enabled = -1 d2887open.cal. Enabled = -1 d2887open.diffplotd2887. Enabled = -1 End Sub ‘Subroutine to read the chromatography data and plot a chart between boiling point distribution and cumulative weight percent for the ASTM D2887. Sub d2887displot_Click () Static a(2500), b(2500), X(2500), t(2500), cn(2500), rt(2500), bp(2500), ctime(3), s(2500), c0(50), ca(50), co(50), cum(50), e(50) Open blankrS For Input As #1 445 Open samplers For Input As #2 Open outputr$ For Append As #3 Open calibr$ For Input As #4 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k2 = k1 -1 k=0 While Not EOF(1) k=k+1 Input #1, afk) Wend k=0 While Not EOF(2) k=k+1 Input #2, b(k) Wend For I = 1 To k s(l) = b(l) - a(l) Next I w=0 p = solpeaktime *30 + 1 v = rt(k2) *30 + 7 For I = p To v w = w + s(l) Next I For I = 1 To k2 ca(0) = 0 cum(0) = 0 d=0 co(0) = 0 X(l) = (rt(l) *30 + 7) For u = p To X(!) d = d + s(u) Next u ca(l) = d co(l) = ca(l) - ca(l -1) e(l) = co(l) / w cum(l) = cum(l -1 ) + e(l) Next I Unload d2887open Load d2887chro d2887chro.Show 446 d2887chro.Graph.FontSize = 150 d2887chro. Graph. FontStyle = 6 d2887chro.Graph.BottomTitle = "Boiling Point, Deg.F" d2887chro.Graph.GraphTitte = "Boiling Point Distribution" d2887chro.Graph. LeftTitle = "Cumulative Weight Fraction" d2887chro. Graph.TickEvery = 200 d2887chro. Graph. YAxisMax = 1 d2887chro.Graph.YAxisMin = 0 d2887chro.Graph.NumPoints = k2 d2887chro.Graph.NumSets = 1 d2887chro.Graph.YAxisTicks = 10 d2887chro.Graph.Autolnc = 0 d2887chro Graph.ThisSet = 1 For I = 1 1 w k2 d2887chro. Graph.ThisPoint = I d2887chro.Graph.XPosData = bp(l) Next I For I = 1 To k2 d2887chro.Graph.ThisPoint = I d2887chro. Graph. GraphData = cum(l) Next I d2887chro. Graph. DrawMode = 2 Close d2887chro.reset.Visible = False End Sub Subroutine to unload the ASTM D2887 opening screen and load the input data ‘screen for the ASTM D2887 method Sub d2887inp_Click () Unload d2887open Load d2887inpu d2887inpu.Show End Sub ‘Subroutine to read the chromatography data and plot a chart between time and ‘signal intensity for the sample and balnk runs for the ASTM D2887. Sub d2887plotchro_Click () screen. MousePointer = 11 Static a(2, 2500), s(2500) Open blankr$ For Input As #1 Open samplers For Input As #2 Open outputr$ For Append As #3 Open calibr$ For Input As #4 k=0 While Not EOF(1) 447 k=k+1 Input #1, a(1, k) Wend k=0 While Not E0F(2) k=k+1 Input #2, a(2, k) Wend Unload d2887open Load d2887chro d2887chro.Show d2887chro.Graph.FontSize = 150 d2887chro.Graph.FontStyle = 6 d2887chro. Graph. LeftTitle = "Signal Intensity" d2887chro. Graph. BottomTitle = "Time, min" d2887chro. Graph. GraphTitle = "Chromatograms" d2887chro.Graph.YMax = 150000 d2887chro.Graph.YMin = 0 d2887chro.Graph.NumPoints = k d2887chro. Graph. NumSets = 2 d2887chro.Graph.Autolnc = 0 For I = 1 To 2 d2887chro.Graph.ThisSet = I For j = 1 To k d2887chro.Graph.ThisPoint = j d2887chro.Graph.GraphData = a(l, j) Next j s(1) = 1 / 30 For j = 1 To k d2887chro.Graph.ThisPoint = j d2887chro.Graph.XPosData = s(j) s(j + 1) = sG) + 1 / 30 Next] Next I d2887chro.Graph.YAxisStyle = 2 d2887chro. Graph. YTicks = 6 d2887chro.Graph.DrawMode = 2 d2887chro.Text1 .Visible = False d2887chro.Text2.Visible = False d2887chro.Text5.Visible = False d2887chro.Text6.Visible = False screen. MousePointer = 0 Close End Sub 448 ‘Subroutine to read the charomatographic data and plot a chart between boiling ‘point and differential weight fraction for the ASTM D2887. Sub diffp!otd2887_C Iick () Static a(2500), b(2500), X(2500), t(2500), cn(2500), rt(2500), bp(2500), ctime(3), s(2500), c0(50), ca(50), co(50), cum(50), e(50) Open blankrS For Input As #1 Open samplers For Input As #2 Open outputr$ For Append As #3 Open calibrS For Input As #4 k1 =0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k2 = k1 -1 k=0 While Not EOF(1) k=k+1 Input #1, a(k) Wend k=0 While Not EOF(2) k=k+1 Input #2, b(k) Wend For I = 1 To k s(D = b(l) - a(l) Next I w=0 p = solpeaktime *30 + 1 v = rt(k2) *30 + 7 For I = p To v w = w + s(l) Next I For I = 1 To k2 ca(0) = 0 cum(0) = 0 d=0 co(0) = 0 X(l) = (rt(l) *30 + 7) For u = p To X(l) d = d + s(u) Next u ca(l) = d co(l) = ca(l) - ca(l -1) 449 e(l) = co(l) / w cum(l) = cum(l -1 ) + e(!) Next I Unload d2887open Load d2887chro d2887chro.Show d2887chro.Graph.FontSize = 150 d2887chro. Graph. FontStyle = 6 d2887chro.Graph.BottomTitle = "Boiling Point, Deg.F" d2887chro.Graph.GraphTitle = "Differential Boiling Point Distribution" d2887chro.Graph.LeftTitle = "Weight Fraction" d2887chro. Graph.TickEvery = 200 d2887chro. Graph. YAxisMax = 1 d2887chro. Graph. YAxisMin = 0 d2887chro.Graph.NumPoints = k2 d2887chro.Graph.NumSets = 1 d2887chro. Graph.Ticks = 1 d2887chro. Graph. YAxisTicks = 10 d2887chro. Graph. IndexStyle = 0 d2887chro. Graph. Autolnc = 0 d2887chro.Graph.ThisSet = 1 For I = 1 To k2 d2887chro.Graph.ThisPoint = I d2887chro.Graph.XPosData = bp(l) Next I For I = 1 To k2 d2887chro.Graph.ThisPoint - I d2887chro.Graph.GraphData = e(l) Next I d2887chro. Graph. DrawMode = 2 Close d2887chro. reset. Visible = False End Sub Subroutine to exit from the ASTM D2887 opening screen to the program startup ‘screen Sub exi_Click () Unload d2887open Load start start. Show End Sub Subroutine to exit from the program Sub exit_Click () End 450 End Sub ‘Subroutine to swich from ASTM D2887 method to D5307 method. Sub switchtoastmd5307_Click () Unload d2887open Load d5307open d5307open.Show End Sub Subroutine for loading the user input screen for the ASTM D2887 Sub Form_Load () On Error GoTo errhandler cmdialogl .DialogTitle = "Choose a Blank Data Filename" cmdialogl.Filter = "RPT Files (*.rpt)|Blank5*.rpt|" cmdialogl.Action = 1 d2887blank.Text = cmdialogl.Filename cmdialog2.DialogTitle = "Choose a Sample Data Filename" cmdialog2.Filter = "RPT Files (* rpt)|a*rpt|" cmdialog2.Action = 1 d2887sample.Text = cmdialog2. Filename cmdialog3.DialogTitle = "Choose a Calibration Data Filename" cmdialog3.Filter = "CAI Files (*.cal)|d*.cal|" cmdialog3.Action = 1 d2887calib.Text = cmdialog3.Filename errhandler: Exit Sub End Sub Subroutine for unloading the user input screen and back to the ASTM D2887 'opening screen. Sub d2887inpuok_Click () solpeaktime = Val(solpeak.Text) blankrS = d2887inpu.d2887blank.Text samplers = d2887inpu.d2887sample.Text calibrS = d2887inpu.d2887calib.Text outputrS = d2887inpu.d2887output.Text Unload d2887inpu Load d2887open d2887open.Show d2887open.astmd2887. Enabled = -1 d2887open.astmd2887ex.Enabled = -1 d2887open.d2887cuts. Enabled = -1 d2887open.d2887displot.Enabled = -1 d2887open.d2887plotchro.Enabled = -1 d2887open.diffplotd2887. Enabled = -1 451 d2887open.light.Enabled = -1 d2887open.cal.Enabled = -1 End Sub Subroutine to close the d2887results screen and open d2887open screen Subroutine Sub d2887resok_Click () Unload d2887res Load d2887open d2887open.Show d2887open.astmd2887. Enabled = -1 d2887open.astmd2887ex. Enabled = -1 d2887open.d2887cuts.Enabled = -1 d2887open.d2887displot.Enabled = -1 d2887open.d2887plotchro. Enabled = -1 d2887open.diffplotd2887. Enabled = -1 d2887open. light. Enabled = -1 d2887open.cal. Enabled = -1 End Sub Subroutine to close the d2887chromatogram screen and open d2887open screen Sub Done_Click () Unload d2887chro Load d2887open d2887open.Show d2887open.astmd2887. Enabled = -1 d2887open.astmd2887ex.Enabled = -1 d2887open.d2887cuts.Enabled = -1 d2887open.d2887displot.Enabled = -1 d2887open.d2887plotchro. Enabled = -1 d2887open.diffplotd2887. Enabled = -1 d2887open.light.Enabled = -1 d2887open.cal. Enabled = -1 End Sub Subroutine to open d2887 chromatogram screen Sub Form_Load () Graph. SDKMouse = 0 End Sub Subroutine to activate the Hotgraphing for the d2887 chromatogram screen Sub Graph_HotHit (HITSET As Integer, hitpoint As Integer) Graph.ThisSet = HITSET Graph.ThisPoint = hitpoint text5.Text = Val(Graph.XPosData) 452 text6.Text = Val(Graph.GraphData) End Sub Subroutine to redraw the chromatograms in the d2887 chromatogram screen Sub reset_Click () Graph.XAxisStyle = 0 Graph.YAxisStyle = 0 Graph. DrawMode = 2 End Sub Subroutine to read the data from the textbox in the d2887 chromatogram screen Sub Text5_KeyDown (keycode As Integer, Shift As Integer) If keycode = 13 Then Graph.XPosData = Val(text5.Text) End If End Sub Subroutine to read the data from the textbox in the d2887 chromatogram screen Sub Text6_KeyDown (keycode As Integer, Shift As Integer) If keycode = 13 Then Graph.GraphData = Val(text6.Text) Graph. DrawMode = 2 End If End Sub Subroutine to read the chromatogram data from file and estimates the boiling point distribution according to ASTM D5307 method. Sub astmd5307_Click () Static A(2500), b(2500), c(2500), S(2500), si(2500), d(2500), X(2500), t(2500), cn(2500), rt(2500), bp(2500), ctime(3), c0(50), ca(50), co(50), cum(50), e(50) Unload d5307open Load d5307res d5307res. Show Open blankrl $ For Input As #1 Open samplerl $ For input As #2 Open outputr1$ For Append As #3 Open calibrl $ For Input As #4 Open blankr2$ For Input As #5 Open isr$ For Input As #6 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend 453 k2 = k1 -1 k=0 While Not E0F(1) k=k+1 Input #1, A(k) Wend k=0 While Not E0F(2) k=k+1 Input #2, b(k) Wend k=0 While Not E0F(5) k=k+1 Input #5, c(k) Wend k=0 While Not E0F(6) k=k+ 1 Input #6, d(k) Wend For I = 1 To k S(l) = b(l) - A(l) si(l) = d (l)-c (l) Next I isratio = iswe / (iswe + sampwt) firstt = first * 30 * .95 lastt = last * 30 * 1.05 p = solpeaktimel *30 + 1 h=0 I= 0 v = (((rt(20) - rt(19)) / ((1013 - 993) * 7)) + rt(19)) *30 + 7 For I = p To v h = h + si(l) I = I + S(l) Next I ais = 0 bis = 0 For I = firstt To lastt ais = ais + si(l) bis = bis + S(l) Next I r = (I - bis) / (h - ais) w = ((ais * r) - bis) * ((1 - isratio) / isratio) For I = 1 To 20 454 ca(0) = 0 cum(O) = 0 o =0 co(0)= 0 X(l) = (rt(l) *30 + 7) For u = p To X(l) o = o + S(u) Next u ca(I) = o co(l) = ca(l) - ca(l -1) e(l) = co(l) / w cum(l) = cum(l -1 ) + e(l) Next I Pr nt #3, "ASTM D5307 Calculation" Pr nt #3, "" Pr nt #3, "" Pr nt #3, Tab(50); Date$ Pr nt #3, Tab(50); Time$ Pr nt #3,"" Pr nt #3, "" Pr nt #3, "" Pr nt #3, "Total Area-'; w Pr nt #3,"" Pr nt #3, "" Pr nt #3,"" Pr nt #3, "Blank Filename blankr1$ Pr nt #3, "sample Filename sampler1$ Pr nt#3, "Calibration Filename ", calibr1$ Pr nt #3, "Output Filename outputr1$ Pr nt #3, "" Pr nt #3, "" Pr nt #3, "Carbon", "Boiling", "Weight", "Cum. Weight Pr nt #3, "Number", "Point",. "Fraction", "Fraction" Pr nt #3, "", "Deg.F" Pr nt #3, "" Pr nt #3, "" For I = 1 To 20 cn1$ = Format$(cn(l), "###") bp1$ = Format$(bp(l), "####") e1$ = Format$(e(l), "##.###") cum1$ = Format$(cum(l), "##.###") Print #3, cn1$, bp1$, e1$, cum1$ Next I Close d5307res.d5307resresults.FontSize = 13 455 d5307res.d5307resresults. P rint," ASTM D5307 Calculation Results" d5307res.d5307resresults. Print,"" d5307res.d5307resresults.FontSize = 7 d5307res.d5307resresults.Print , "Carbon", "Boiling", "Weight", "Cum. Weight" d5307res.d5307resresults.Print, "Number", "Point", "Fraction", "Fraction" d5307res.d5307resresu!ts.Print, "Deg.F" d5307res.d5307resresults.Print, "" For I = 1 To 20 cn2$ = Format$(cn(l), "###") bp2$ = Format$(bp(l), "####") e2$ = Format$(e(l), "##.###") cum2$ = Format$(cum(l), "##.###") d5307res.d5307resresults.Print, cn2$, bp2$, e2$, cum2$ Next I d5307open.astmd5307. Enabled = -1 d5307open.astmd5307ex.Enabled = -1 d5307open.d5307cuts. Enabled = -1 d5307open.sampisblank.Enabled = -1 d5307open.cal.Enabled = -1 d5307open.d5307diff. Enabled = -1 End Sub Subroutine to read the chromatogram data from file and estimates the boiling point distribution according to ASTM D5307 extended method explained in the dissertation. Sub astmd5307ex_Click () Static A(2500), b(2500), c(2500), S(2500), si(2500), d(2500), X(2500), t(2500), cn(2500), rt(2500), bp(2500), ctime(3), c0(50), ca(50), co(50), cum(50), e(50) Open blankr1$ For Input As #1 Open sampler1$ For Input As #2 Open outputr1$ For Append As #3 Open calibrl $ For Input As #4 Open blankr2$ For Input As #5 Open isr$ For Input As #6 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k2 = k1 -1 k=0 While Not EOF(1) 456 k=k+ 1 Input #1, A(k) Wend k=0 While Not E0F(2) k=k+ 1 Input #2, b(k) Wend k=0 While Not E0F(5) k=k+1 Input #5, c(k) Wend k=0 While Not E0F(6) k=k+1 Input #6, d(k) Wend For I = 1 To k S(l) = b(I) - A(I) si(l) = d(l) - c(l) Next I isratio = iswe / (iswe + sampwt) firstt = first * 30 * .95 lastt = last * 30 * 1.05 p = solpeaktimel *30 + 1 h=0 I= 0 v = rt(k2) *30 + 7 For I = p To v h = h + si(l) I = I + S(l) Next I ais = 0 bis = 0 For I = firstt To lastt ais = ais + si(l) bis = bis + S(l) Next I r = (I - bis) / (h - ais) w = ((ais * r) - bis) * ((1 - isratio) / isratio) For I = 1 To k2 ca(0)= 0 cum(0) = 0 o=0 457 co(0) = 0 X(l) = (rt(l) *30 + 7) For u = p To X(l) o = o + S(u) Next u ca(l) = o co(l) = ca(l) - ca(l -1 ) e(l) = co(l) / w cum(l) = cum(l -1 ) + e(l) Next I Print #3, "ASTM D5307 Extended Method Calculation" Print #3,"" Print #3,"" Print #3, Tab(50); Date$ Print #3, Tab(50); Time$ Print #3,"" Print #3,"" Print #3,"" Print #3, "Total Area="; w Print # 3 ,"” Print #3,"" Print #3,"" Print #3, "Blank Filename ", blankr1$ Print #3, "sample Filename sampler1$ Print #3, "Calibration Filename calibr1$ Print #3, "Output Filename ", outputr1$ Print #3,"" Print #3,"" Print #3, "Carbon", "Boiling", "Weight", "Cum. Weight" Print #3, "Number”, "Point", "Fraction", "Fraction" Print # 3 , "Deg.F" Print #3,"" Print #3,"" For I = 1 To k2 cn1$ = Format$(cn(l), "###”) bp1$ = Format$(bp(l), "####") e1$ = Format$(e(l), "##.###”) cum1$ = Format$(cum(l), "##.###") Print #3, cn1$, bp1$, e1$, cum1$ Next I Close Unload d5307open Load d5307res d5307res. Show d5307res.d5307resresults.FontSize = 13 458 d5307res.d5307resresu!ts.Print , " ASTM D5307 Extended Method Results" d5307res.d5307resresults.Print,"" d5307res.d5307resresults.FontSize = 7 d5307res.d5307resresults.Print , "Carbon", "Boiling", "Weight", "Cum. Weight" d5307res.d5307resresults.Print, "Number", "Point", "Fraction", "Fraction" d5307res.d5307resresults.Print, "Deg.F" d5307res.d5307resresults. Print,"" For I = 1 To k2 Step 2 cn2$ = Format$(cn(l), "###") bp2$ = Format$(bp(l), ”####") e2$ = Format$(e(l), "##.###") cum2$ = Format$(cum(l), "##.###") d5307res.d5307resresults.Print, cn2$, bp2$, e2$, cum2$ Next I d5307open.astmd5307.Enabled = -1 d5307open.astmd5307ex. Enabled = -1 d5307open.d5307cuts. Enabled = -1 d5307open.sampisblank. Enabled = -1 d5307open.cal. Enabled = -1 d5307open.d5307diff. Enabled = -1 End Sub Subroutine to read the calibration data from the file and draws a graph between boilint point and retention time. Sub cal_Click () Static cn(2500), rt(50), bp(50) Open calibr1$ For Input As #4 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k2 = k1 -1 Unload d5307open Load d5307chro d5307chro.Show d5307chro.reset.Visible = False d5307chro.Graph NumPoints = k2 d5307chro.Graph.NumSets = 1 d5307chro. Graph. Autolnc = 0 d5307chro.Graph.ThisSet = 1 For j = 1 To k2 d5307chro.Graph.ThisPoint = j 459 d5307chro.Graph.XPosData = bp(j) d5307chro.Graph.GraphData = rt(j) Next j d5307chro.Graph.YAxisStyle = 2 d5307chro.Graph.YAxisMax = 45 d5307chro. Graph. YAxisMin = 0 d5307chro.Graph.YAxisTicks = 9 d5307chro. Graph. DrawMode = 2 Close End Sub Subroutine to read the chromatogram data from the file and estimates the distillation cuts according to the ASTM D5307 method. Sub d5307cuts_Click () Static A(2500), b(2500), c(2500), S(2500), si(2500), d(2500), X(2500), t(2500), cn(50), rt(50), bp(2500), ctime(3), c0(50), ca(50), co(50), cum(50), e(50) Open blankrl $ For Input As #1 Open samplerl $ For Input As #2 Open outputr1$ For Append As #3 Open calibr1$ For Input As #4 Open blankr2$ For Input As #5 Open isr$ For Input As #6 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k2 = k1 -1 k=0 While Not EOF{1) k=k+ 1 Input #1, A(k) Wend k=0 While Not EOF(2) k=k+ 1 Input #2, b(k) Wend k=0 While Not EOF(5) k=k+1 Input #5, c(k) Wend k=0 460 While Not E0F(6) k=k+1 Input #6, d(k) Wend For I = 1 To k S(l) = b(l) - A(l) si(l) = d(l) -c(l) Next I isratio = iswe / (iswe + sampwt) firstt = first * 30 * .95 lastt = last *30 *1.05 p = solpeaktimel *30 + 1 h=0 I=0 v = (((rt(20) - rt(19)) / (1013 - 993) * 7) + rt(19)) *30 + 7 For I = p To v h = h + si(l) I = I + S(l) Next I ais = 0 bis = 0 For I = firstt To lastt ais = ais + si(l) bis = bis + S(l) Next I r = (I - bis) / (h - ais) w = ((ais * r) - bis) * ((1 - isratio) / isratio) ctime(1) = ((rt(4) - rt(3)) / (421 - 345) * 65) + rt(3) ctime(2) = rt(8) ctime(3) = ((rt(20) - rt(19)) / (1013 - 993) * 7) + rt(19) For I = 1 To 3 ca(0) = 0 cum(0) = 0 f=0 co(0) = 0 X(l) = (ctime(l) *30 + 7) For u = p To X(l) f = f + S(u) Next u ca(l) = f co(l) = ca(l) - ca(l -1) e(l) = co(l) / w cum(l) = cum(l -1) + e(l) Next I Print # 3 ,"" 461 Pr nt #3, Pr nt #3, Pr nt #3, Pr nt #3, Pr nt #3, "Distillation Cuts Calculation Using ASTM D5307 Method" Pr nt #3, Pr nt #3, Pr nt #3, "Total Area="; w Pr nt #3, Pr nt #3, Pr nt #3, blankr1$ Pr nt #3, ", sampler1$ Pr nt #3, calibrl $ Pr nt #3, outputr1$ Pr nt #3, Pr nt #3, Pr nt #3, Pr nt #3, Pr nt #3, cum1$ = Format$(e(1), '■#.###") cum2$ = Format$(e(2), "#.###") cum3$ = Format$(e(3), "#.###") cum4$ = Format$((1 - cum(3)), ’"#.###") Print #3, "Gasoline =", Tab(35); cum1$ Print #3, "Middle Distillates =", Tab(35); cum2$ Print #3, "Gas Oil Tab(35); cum3$ Print #3, "Residue =", Tab(35); cum4$ Close Unload d5307open Load d5307res d5307res.Show d5307res.d5307resresults.FontSize = 13 d5307res.d5307resresults.Print," Distillation Cuts Calculation" d5307res.d5307resresults.Print, "" d5307res. d5307resresults. FontSize = 8.25 d5307res. d5307resresults. Pr nt d5307res. d5307resresults. Pr nt d5307res.d5307resresults.Pr nt "Gasoline =", cum1$ d5307res.d5307resresults.Pr nt d5307res.d5307resresults.Pr nt "Middle Distillates =", cum2$ d5307res.d5307resresults.Pr nt d5307res.d5307resresults. Pr nt "Gas Oil =", cum3$ mi d5307res.d5307resresults. Pr nt d5307res. d5307resresults. Pr nt "Residue cum4$ d5307open.astmd5307. Enabled = -1 I HI «««« 462 d5307open.astmd5307ex.Enabled = -1 d5307open.d5307cuts. Enabled = -1 d5307open.sampisblank.Enabled = -1 d5307open cal. Enabled = -1 d5307open.d5307diff.Enabled = -1 End Sub Subroutine to read the chromatogram data from file and plots the chromatograms. Sub d5307diff_Click () Static A(2500), b(2500), c(2500), S(2500), si(2500), d(2500), X(2500), t(2500), cn(2500), rt(2500), bp(2500), ctime(3), c0(50), ca(50), co(50), cum(50), e(50) Open blankr1$ For Input As #1 Open sampler1$ For Input As #2 Open outputrl $ For Append As #3 Open calibr1$ For Input As #4 Open blankr2$ For Input As #5 Open isr$ For Input As #6 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k2 = k1 -1 k=0 While Not EOF(1) k=k+1 Input #1, A(k) Wend k=0 While Not EOF(2) k=k+1 Input #2, b(k) Wend k=0 While Not EOF(5) k=k+ 1 Input #5, c(k) Wend k=0 While Not EOF(6) k=k+1 Input #6, d(k) Wend 463 For I = 1 To k S(l) = b(l) - A(l) si(l) = d(l) - c(l) Next I isratio = iswe / (iswe + sampwt) firstt = first * 30 * .95 lastt = last * 30 * 1.05 p = solpeaktime 1 *30 + 1 h=0 I=0 v = rt(k2) *30 + 7 For I = p To v h = h + si(l) I = I + S(l) Next I ais = 0 bis = 0 For I = firstt To lastt ais = ais + si(l) bis = bis + S(l) Next I r = (I - bis) / (h - ais) w = ((ais * r) - bis) * ((1 - isratio) / isratio) For I = 1 To k2 ca(0)= 0 cum(0) = 0 o=0 co(0) = 0 X(l) = (rt(l) *30 + 7) For u = p To X(l) o = o + S(u) Next u ca(l) = o co(l) = ca(l) - ca(l -1 ) e(l) = co(l) / w cum(l) = cum(l -1 ) + e(l) Next I Unload d5307open Load d5307chro d5307chro.Show d5307chro. Graph. FontStyle = 6 d5307chro.Graph.FontSize = 90 d5307chro.Graph.LeftTitle = "Weight Fraction" d5307chro.Graph.BottomTitle = "Boiling Point, Deg.F" d5307chro. Graph. FontStyle = 6 464 d5307chro.Graph.FontSize = 150 d5307chro. Graph.GraphTitle = "Diff. Boiling Point Distribution" d5307chro. Graph, NumPoints = k2 d5307chro. Graph. NumSets = 1 d5307chro. Graph.Ticks = .01 d5307chro.Graph.YAxisTicks = 10 d5307chro.Graph.lndexStyle = 0 d5307chro.Graph.AutoInc = 0 d5307chro.Graph.ThisSet = 1 For I = 1 To k2 d5307chro.Graph.ThisPoint = I d5307chro.Graph.XPosData = bp(l) d5307chro.Graph.GraphData = e(I) Next I d5307chro. Graph. YAxisStyle = 2 d5307chro.Graph.YAxisMax = .1 d5307chro.Graph.YAxisMin = 0 d5307chro.Graph.TickEvery = 200 d5307chro.Graph.LabelEvery = 200 d5307chro.Graph.TickEvery = 200 d5307chro,Graph.YAxisTicks = 5 d5307chro. Graph. DrawMode = 2 Close d5307chro.reset.Visible = False End Sub Subroutine to read the chromatogram data from file and estimates the boiling point distribution and draws a graph between boiling point distribution and cumulative weight fraction according to ASTM D5307 method. Sub d5307displ_Click () Static A(2500), b(2500), c(2500), S(2500), si(2500), d(2500), X(2500), t(2500), cn(2500), rt(2500), bp(2500), ctime(3), c0(50), ca(50), co(50), cum(50), e(50) Open blankr1$ For Input As #1 Open sampler1$ For Input As #2 Open outputrl $ For Append As #3 Open calibr1$ For Input As #4 Open blankr2$ For Input As #5 Open isr$ For Input As #6 k1 = 0 While Not EOF(4) k1 = k1 + 1 Input #4, cn(k1), rt(k1), bp(k1) Wend k2 = k1 -1 465 k=0 While Not E0F(1) k=k+1 Input #1, A(k) Wend k=0 While Not E0F(2) k=k+ 1 Input #2, b(k) Wend k=0 While Not E0F(5) k=k+1 Input #5, c(k) Wend k=0 While Not E0F(6) k=k+1 Input #6, d(k) Wend For I = 1 To k S(l) = b(l) - A(l) si(l) = d(l)-c(l) Next I isratio = iswe / (iswe + sampwt) firstt = first * 30 * .95 lastt = last * 30 * 1.05 p = solpeaktimel *30 + 1 h=0 I= 0 v = rt(k2) *30 + 7 For I = p To v h = h + si(l) I = I + S(l) Next I ais = 0 bis = 0 For I = firstt To lastt ais = ais + si(l) bis = bis + S(l) Next I r = (I - bis) / (h - ais) w = ((ais * r) - bis) * ((1 - isratio) I isratio)For I = 1 To k2 ca(0) = 0 cum(0) = 0 466 o=0 co(0) = 0 X(l) = (rt(l) * 30 + 7) For u = p To X(l) o = o + S(u) Next u ca(l) = o co(l) = ca(l) - ca(l -1) e(l) = co(l) / w cum(l) = cum(l -1) + e(l) Next I Unload d5307open Load d5307chro d5307chro.Show d5307chro. Graph. FontStyle = 6 d5307chro. Graph. FontSize = 90 d5307chro.Graph.LeftTitle = "Cumulative Weight Fraction" d5307chro.Graph.BottomTitle = "Boiling Point, Deg.F" d5307chro. Graph. FontStyle = 6 d5307chro. Graph. FontSize = 150 d5307chro.Graph.GraphTitle = "Boiling Point Distribution" d5307chro.Graph.NumPoints = k2 d5307chro.Graph.NumSets = 1 d5307chro.Graph.Ticks = 1 d5307chro.Graph.YAxisTicks = 10 d5307chro. Graph. IndexStyle = 0 d5307chro. Graph. Autolnc = 0 d5307chro.Graph.ThisSet = 1 For I = 1 To k2 d5307chro.Graph.ThisPoint = I d5307chro. Graph. XPosData = bp(l) d5307chro.Graph.GraphData = cum(l) Next I d5307chro. Graph. YAxisStyle = 2 d5307chro. Graph. YAxisMax = 1 d5307chro. Graph. YAxisMin = 0 d5307chro.Graph.TickEvery = 200 d5307chro, Graph. LabelEvery = 200 d5307chro. Graph.TickEvery = 200 d5307chro. Graph. YAxisTicks = 5 d5307chro. Graph. DrawMode = 2 Close d5307open.astmd5307. Enabled = -1 d5307open.astmd5307ex.Enabled = -1 d5307open.d5307cuts.Enabled = -1 467 d5307open.sampisblank.Enabled = -1 d5307chro.reset.Visible = False End Sub Subroutine to end the program Sub d5307ex_Click () End End Sub Subroutine to load the d5307 input screen for input Sub d5307in_Click () Load d5307inpu d5307inpu.Show End Sub Subroutine to unload the d5307 opening screen and loads the software starting screen. Sub d5307retun_Click () Unload d5307open Load start start. Show End Sub Subroutine to unload the d5307 opening screen and loads the d2887 opening screen. Sub switchtod2887_CIick () Unload d5307open Load d2887open d2887open.Show End Sub Subroutine to read the input from d5307 input screen and returns to d5307 opening screen. Sub d5307inpuok_Click () blankr1$ = d5307blank1.Text sampler1$ = d5307sample.Text blankr2$ = d5307blank2.Text isr$ = d5307is.Text calibr1$ = d5307calib.Text outputr1$ = d5307output.Text Unload d5307inpu Load d5307para d5307para.Show End Sub 468 Subroutine to load the dialog box for obtaining paths for the blank, sample and sample plus internal standard chromatogram data. Sub Form_Load () On Error GoTo errhandler cmdialogl.DialogTitle = "Choose a Blank Data Filename for Sample" cmdialogl.Filter = "RPT Files (*.rpt)|BLANKd28.rpt|" cmdialogl.Action = 1 d5307blank1.Text = cmdialogl .Filename cmdialog2.DialogTitle = "Choose a Sample Data Filename" cmdialog2.Filter = "RPT Files ( * rpt)|a*rpt|" cmdialog2.Action = 1 d5307sample.Text = cmdialog2.Filename cmdialog3.DialogTitle = "Choose a Blank Data Filename for IS" cmdialog3.Filter = "RPT Files (* rpt)|BLANKd28* rpt|" cmdialog3.Action = 1 d5307blank2.Text = cmdialog3.Filename cmdialog4.DialogTitle = "Choose a IS Data Filename" cmdialog4.Filter = "RPT Files (* rpt)|a* rpt|" cmdialog4.Action = 1 d5307is.Text = cmdialog4.Filename cmdialog5.DialogTitle = "Choose a Calibration Data Filename" cmdialog5.Filter = "CAI Files (* cal)|d2887d28.cal|" cmdialog5.Action = 1 d5307calib.Text = cmdialog5.Filename errhandler: Exit Sub End Sub Subroutine to close the d5307 parameter input screen and switch to the d5307 opening screen. Sub d5307paraok_Click () first = Val(d5307para.firstime.Text) last = Val(d5307para.lastime.Text) iswe = Val(d5307para.isweight.Text) sampwt = Val(d5307para.sampleweight.Text) solpeaktimel = Val(text2.Text) Unload d5307para Load d5307open d5307open.Show d5307open.astmd5307. Enabled = -1 d5307open.astmd5307ex.Enabled = -1 d5307open.d5307cuts.Enabled = -1 d5307open.d5307displ. Enabled = -1 d5307open.sampisblank.Enabled = -1 d5307open.cal.Enabled = -1 469 d5307open.d5307diff.Enabled = -1 End Sub Subroutine to close the d5307 results display screen and open the d5307 opening screen. Sub Command 1_Click () Unload d5307res Load d5307open d5307open.Show d5307open.astmd5307.Enabled = -1 d5307open.astmd5307ex.Enabled = -1 d5307open.d5307cuts.Enabled = -1 d5307open.d5307displ.Enabled = -1 d5307open.sampisblank.Enabled = -1 End Sub Declaration of constants for the graph zooming and hot graphing options. Dim newbox Dim badpoint Dim x1 As Double, y1 As Double Dim x2 As Double, y2 As Double Dim xorigin As Double Dim yorigin As Double Dim xaxislength As Double Dim yaxislength As Double Dim xtemp As Double Dim ytemp As Double Dim xmax As Integer, ymax As Integer Dim xmin As Integer, ymin As Integer Rem GSWDEFS.TXT Rem Graphics Server Version 3 Rem Prototypes of GS ver 3 functions for Visual Basic Rem (c) Bits Per Second Ltd, Brighton, England Declare Function AG3DStyle Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nDepth%, ByVal nXGap%, ByVal nZGap%) As Integer Declare Function AGAmp Lib "GSWAG16.DLL" (ByVal nPts%, ByVal nGroup%, fAmp#) As Integer Declare Function AGAmpError Lib "GSWAG16.DLL" (ByVal nPts%, ByVal nGroup%, fAmp#) As Integer Declare Function AGAux Lib "GSWAG16.DLL" (ByVal nSize%, nAux%) As Integer Declare Function AGCageStyle Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nClrWall%, ByVal nClrSide%) As Integer Declare Function AGCIose Lib "GSWAG16.DLL" () As Integer 470 Declare Function AGCIr Lib "GSWAG16.DLL" (ByVal nGroup%, nClr%) As Integer Declare Function AGCurveStyle Lib "GSWAG16.DLL" (ByVal nType%, ByVal nOrder%, ByVal nSteps%) As Integer Declare Function AGDataZ Lib "GSWAG16.DLL" (ByVal nPts%, fDataZ#) As Integer Declare Function AGDist Lib "GSWAG16.DLL" (ByVal nPts%, nDist#) As Integer Declare Function AGDistError Lib "GSWAG16.DLL" (ByVal nPts%, nDist#) As Integer Declare Function AGErrorBar Lib "GSWAG16.DLL" (ByVal nSel%, ByVal nSymStyle%, ByVal nClr%, ByVal nErrSrc%, ByVal fValue#) As Integer Declare Function AGFGColor Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nClr%) As Integer Declare Function AGFFT Lib "GSWAG16.DLL" (ByVal nPts%, fData#, ByVal nMode%) As Integer Declare Function AGFontStyle Lib "GSWAG16.DLL" (ByVal nUse%, ByVal nFamily%, ByVal nAttribs%, ByVal nSize%) As Integer Declare Function AGGraphBG Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nClr%) As Integer Declare Function AGGridStyle Lib "GSWAG16.DLL" (ByVal nSel%, ByVal nStyleMaj%, ByVal nStyleMin%) As Integer Declare Function AGInfo Lib "GSWAG16.DLL" (ByVal nlndex%) As Double Declare Function AGLegendStyle Lib "GSWAG16.DLL" (ByVal nVert%, ByVal nHoriz%, ByVal nSize%, ByVal nClr%, ByVal nMode%) As Integer Declare Function AGOpen Lib "GSWAG16.DLL" () As Integer Declare Function AGPatt Lib "GSWAG16.DLL" (ByVal nGroup%, nPatt%) As Integer Declare Function AGRefresh3D Lib "GSWAG16.DLL" (ByVal nMode%) As Integer Declare Function AGReset Lib "GSWAG16.DLL" () As Integer Declare Function AGSetPerspective Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nRot%, ByVal nElev%, ByVal nEyePos%) As Integer Declare Function AGShow Lib "GSWAG16.DLL" (ByVal nGType%, ByVal nStyle%, ByVal nStats%) As Integer Declare Function AGSurfaceClr Lib "GSWAG16.DLL" (ByVal nClrMin%, ByVal nClrMax%, ByVal nClrSide%) As Integer Declare Function AGSym Lib "GSWAG16.DLL” (ByVal nGroup%, nSym%) As Integer Declare Function AGTimeGraph Lib "GSWAG16.DLL" (ByVal nPts%, ByVal nGroup%, ByVal fDataMax#, ByVal fDataMin#, ByVal nStyle%) As Integer Declare Function AGTimeUpdate Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nGroup%, fData#) As Integer Declare Function AGTitleBG Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nClr%) As Integer Declare Function AGTitleG Lib "GSWAG16.DLL" (ByVal szTitle$) As Integer 471 Declare Function AGTitleX Lib "GSWAG16.DLL" (ByVal szTitle$) As Integer Declare Function AGTitleY Lib "GSWAG16.DLL" (ByVal szTitle$) As Integer Declare Function AGTitleYR Lib "GSWAG16.DLL" (ByVal szTitle$) As Integer Declare Function AGXAxisStyle Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nTicks%, ByVal nLabeEvery%, ByVal fMax#, ByVal fMin#) As Integer Declare Function AGYAxisStyle Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nTicks%, ByVal nLabeEvery%, ByVal fMax#, ByVal fMin#) As Integer Declare Function AGYRAxisStyle Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nTicks%, ByVal nLabeEvery%, ByVal fMax#, ByVal fMin#) As Integer Declare Function AGZAxisStyle Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nTicks%, ByVal nLabeEvery%, ByVal fMax#, ByVal fMin#) As Integer Declare Function GSArc Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fRad#, ByVal fAngl#, ByVal fAng2#, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSArea Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fHt#, ByVal nMode%, ByVal nGroup%) As Integer Declare Function GSArea3D Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fHt#, ByVal fDepth#, ByVal fAng#, ByVal nMode%) As Integer Declare Function GSArrow Lib "GSWDLL16.DLL" (ByVal fxA#, ByVal fya#, ByVal fxB#, ByVal fyB#, ByVal fHeadLen#, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSAxis Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fLen#, ByVal fTickLen#, ByVal nMajDivs%, ByVal nMinDivs%, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSBar2D Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fSpace#, ByVal fStackHt#, ByVal nMode%, ByVal nGroup%) As Integer Declare Function GSBar3D Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fSpace#, ByVal fStackHt#, ByVal fDepth#, ByVal fAng#, ByVal nMode%, ByVal nGroup%) As Integer Declare Function GSBox2D Lib "GSWDLL16 DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fWid#, ByVal fHt#, ByVal nPatt%, ByVal nClr%) As Integer Declare Function GSBox3D Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fWid#, ByVal fHt#, ByVal fDepth#, ByVal fAng#, ByVal nPatt%, ByVal nClr1%, ByVal nClr2%) As Integer Declare Function GSBoxWhisker Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fSpace#, ByVal nMode%) As Integer Declare Function GSBubbleChart Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal nMode%) As Integer Declare Function GSCage3D Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fxLen#, ByVal fyLen#, ByVal fzLen#, ByVal fAng#, ByVal fThick#, ByVal nxGrid%, ByVal nyGrid%, ByVal nzGrid%, ByVal nMode%, ByVal nClr1%, ByVal nClr2%) As Integer 472 Declare Function GSCircle Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fRad#, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSCIearView Lib "GSWDLL16.DLL" (ByVal nMode%) As Integer Declare Function GSCIipRead Lib "GSWDLL16.DLL" (ByVal fxBL#, ByVal fyBL#, ByVal fxTR#, ByVal fyTR#, ByVal nFormat%, ByVal nMode%) As Integer Declare Function GSCIipWrite Lib "GSWDLL16.DLL" (ByVal fxBL#, ByVal fyBL#, ByVal fxTR#, ByVal fyTR#, ByVal nFormat%, ByVal nMode%) As Integer Declare Function GSCIosePrn Lib "GSWDLL16.DLL" () As Integer Declare Function GSCIoseServer Lib "GSWDLL16.DLL" () As Integer Declare Function GSCIoseView Lib "GSWDLL16.DLL" (ByVal nWin%, ByVal nView%, ByVal nMode%) As Integer Declare Function GSCIoseWin Lib "GSWDLL16.DLL" (ByVal nWin%) As Integer Declare Function GSCurveFit Lib "GSWDLL16.DLL" (ByVal nType%, ByVal nOrder%, ByVal nSteps%, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSDataAmp Lib "GSWDLL16.DLL" (ByVal nPts%, ByVal nGroup%, fAmp#) As Integer Declare Function GSDataAmpErr Lib "GSWDLL16.DLL" (ByVal nP%, ByVal nG%, fAmpErr#) As Integer Declare Function GSDataAux Lib "GSWDLL16.DLL" (ByVal nPts%, nAux%) As Integer Declare Function GSDataClr Lib "GSWDLL16.DLL" (ByVal nPts%, nClr%) As Integer Declare Function GSDataDim Lib "GSWDLL16.DLL" (ByVal nPts%, ByVal nGroup%) As Integer Declare Function GSDataDist Lib "GSWDLL16.DLL" (ByVal nPts%, fDist#) As Integer Declare Function GSDataDistErr Lib "GSWDLL16.DLL" (ByVal nP%, fDistErr#) As Integer Declare Function GSDataGetAmp Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal nGroup%) As Double Declare Function GSDataGetAmpErr Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal nGroup%, ByVal nMode%) As Double Declare Function GSDataGetAux Lib "GSWDLL16.DLL" (ByVal nPt%) As Integer Declare Function GSDataGetClr Lib "GSWDLL16.DLL" (ByVal nPt%) As Integer Declare Function GSDataGetDist Lib "GSWDLL16.DLL" (ByVal nPt%) As Double Declare Function GSDataGetDistErr Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal nMode%) As Double Declare Function GSDataGetPatt Lib "GSWDLL16.DLL" (ByVal nPt%) As Integer Declare Function GSDataGetSym Lib "GSWDLL16.DLL" (ByVal nPt%) As Integer 473 Declare Function GSDataGetZ Lib "GSWDLL16.DLL" (ByVal nPt%) As Double Declare Function GSDataPatt Lib "GSWDLL16.DLL" (ByVal nPts%, nPatt%) As Integer Declare Function GSDataRange Lib "GSWDLL16.DLL" (ByVal nFirst%, ByVal nLast%) As Integer Declare Function GSDataReset Lib "GSWDLL16.DLL" () As Integer Declare Function GSDataScale Lib "GSWDLL16.DLL" (ByVal fScale#) As Integer Declare Function GSDataStoreAmp Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal nGroup%, ByVal fAmp#) As Integer Declare Function GSDataStoreAmpErr Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal nGroup%, ByVal fErrP#, ByVal fErrM#) As Integer Declare Function GSDataStoreAux Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal nAux%) As Integer Declare Function GSDataStoreClr Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal nClr%) As Integer Declare Function GSDataStoreDist Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal fDist#) As Integer Declare Function GSDataStoreDistErr Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal fErrP#, ByVal fErrM#) As Integer Declare Function GSDataStorePatt Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal nPatt%) As Integer Declare Function GSDataStoreSym Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal nSym%) As Integer Declare Function GSDataStoreZ Lib "GSWDLL16.DLL" (ByVal nPt%, ByVal fDataZ#) As Integer Declare Function GSDataSym Lib "GSWDLL16.DLL" (ByVal nPts%, nSym%) As Integer Declare Function GSDataTrans Lib "GSWDLL16.DLL" (ByVal nPts%, ByVal nGroup%, fA#, fD#, nPatt%, nSymbol%, nAux%, nClr%) As Integer Declare Function GSDataZ Lib "GSWDLL16.DLL" (ByVal nPts%, fDataZ#) As Integer Declare Function GSDefPatt Lib "GSWDLL16.DLL" (ByVal nBitmap%, wBitmap%) As Integer Declare Function GSEIIipse Lib "GSWDLL16.DLL" (ByVal fxBL#, ByVal fyBL#, ByVal fxTR#, ByVal fyTR#. ByVal fxA#, ByVal fya#, ByVal fxB#, ByVal fyB#, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSErrorBar Lib "GSWDLL16.DLL" (ByVal nSel%, ByVal nSymStyle%, ByVal nClr%, ByVal nErrSrc%, ByVal fValue#, ByVal fOff#) As Integer Declare Function GSFixPos Lib "GSWDLL16.DLL" (ByVal fx#, ByVal fy#) As Integer Declare Function GSGantt Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal nMode%, ByVal nGroup%) As Integer Declare Function GSGetACos Lib "GSWDLL16.DLL" (ByVal fVal#) As Double 474 Declare Function GSGetALoglO Lib "GSWDLL16.DLL" (ByVal fVal#) As Double Declare Function GSGetALogE Lib "GSWDLL16.DLL" (ByVal fVal#) As Double Declare Function GSGetASin Lib "GSWDLL16.DLL" (ByVal fVal#) As Double Declare Function GSGetATan Lib "GSWDLL16.DLL" (ByVal fVal#) As Double Declare Function GSGetAXExt Lib "GSWDLL16.DLL" (ByVal hWnd%) As Double Declare Function GSGetAYExt Lib "GSWDLL16.DLL" (ByVal hWnd%) As Double Declare Function GSGetBG Lib "GSWDLL16.DLL" () As Integer Declare Function GSGetCC Lib "GSWDLL16.DLL" () As Double Declare Function GSGetCos Lib "GSWDLL16.DLL" (ByVal fVal#) As Double Declare Function GSGetCurX Lib "GSWDLL16.DLL" () As Double Declare Function GSGetCurY Lib "GSWDLL16.DLL" () As Double Declare Function GSGetE Lib "GSWDLL16.DLL" () As Double Declare Function GSGetLoglO Lib "GSWDLL16.DLL" (ByVal fVal#) As Double Declare Function GSGetLogE Lib "GSWDLL16.DLL" (ByVal fVal#) As Double Declare Function GSGetMax Lib "GSWDLL16.DLL" () As Double Declare Function GSGetMean Lib "GSWDLL16.DLL" () As Double Declare Function GSGetMF Lib "GSWDLL16.DLL" (ByVal nMode%) As Integer Declare Function GSGetMin Lib "GSWDLL16.DLL" () As Double Declare Function GSGetPi Lib "GSWDLL16.DLL" () As Double Declare Function GSGetPrnHt Lib "GSWDLL16.DLL" (ByVal nUnits%) As Double Declare Function GSGetPrnWid Lib "GSWDLL16.DLL" (ByVal nUnits%) As Double Declare Function GSGetRTextHt Lib "GSWDLL16.DLL" (ByVal nCset%, ByVal nMode%, ByVal szStringS) As Double Declare Function GSGetRTextWid Lib "GSWDLL16.DLL" (ByVal nCset%, ByVal nMode%, ByVal szStringS) As Double Declare Function GSGetSD Lib "GSWDLL16.DLL" () As Double Declare Function GSGetSFHt Lib "GSWDLL16.DLL" () As Double Declare Function GSGetSFWid Lib "GSWDLL16.DLL" () As Double Declare Function GSGetSin Lib "GSWDLL16.DLL" (ByVal fVal#) As Double Declare Function GSGetSXExt Lib "GSWDLL16.DLL" () As Double Declare Function GSGetSYExt Lib "GSWDLL16.DLL" () As Double Declare Function GSGetTan Lib "GSWDLL16.DLL" (ByVal fVal#) As Double Declare Function GSGetVer Lib "GSWDLL16.DLL" (ByVal nVer%) As Integer Declare Function GSGetVXExt Lib "GSWDLL16.DLL" () As Double Declare Function GSGetVYExt Lib "GSWDLL16.DLL" () As Double Declare Function GSGetWXExt Lib "GSWDLL16.DLL" (ByVal nMode%, ByVal nUnits%) As Double Declare Function GSGetWYExt Lib "GSWDLL16.DLL" (ByVal nMode%, ByVal nUnits%) As Double 475 Declare Function GSGrid Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fAxisLen#, ByVal fGridLen#, ByVal nDivs%, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSHLC Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal nMode%, ByVal nClr%) As Integer Declare Function GSHotGraph Lib "GSWDLL16.DLL" (ByVal nMode%) As Integer Declare Function GSLabelnPie Lib "GSWDLL16.DLL" (ByVal fxOff#, ByVal fRad#, ByVal fWid#, ByVal fHt#, ByVal nPrec%, ByVal nMode%, ByVal nCset%, ByVal nTMode%, ByVal nClr%) As Integer Declare Function GSLabelnX Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fWid#, ByVal fHt#, ByVal fBaseVal#, ByVal fStepVal#, ByVal nPrec%, ByVal nNLabs%, ByVal nCset%, ByVal nMode%, ByVal nClr%) As Integer Declare Function GSLabelnY Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fWid#, ByVal fHt#, ByVal fBaseVal#, ByVal fStepVal#, ByVal nPrec%, ByVal nNLabs%, ByVal nCset%, ByVal nMode%, ByVal nClr%) As Integer Declare Function GSLineAbs Lib "GSWDLL16.DLL" (ByVal fxA#, ByVal fya#, ByVal fxB#, ByVal fyB#, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSLineFit Lib "GSWDLL16.DLL" (ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSLineRel Lib "GSWDLL16.DLL" (ByVal fxr#, ByVal fya#, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSLinLog Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fWid#, ByVal fBase#, ByVal nMode%, ByVal nClr%) As Integer Declare Function GSLoadRFont Lib "GSWDLL16.DLL" (ByVal nFamily%, ByVal nAttrib%, ByVal nSize%, ByVal nPitch%) As Integer Declare Function GSLoadVFont Lib "GSWDLL16.DLL" (ByVal nFamily%, ByVal nAttrib%, ByVal nPitch%) As Integer Declare Function GSLogAxis Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fLen#, ByVal fTickLen#, ByVal nCycles%, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSLogGrid Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fAxisLen#, ByVal fGridLen#, ByVal nCycles%, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSLogLin Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fCycleHt#, ByVal fBaseVal#, ByVal nMode%, ByVal nClr%) As Integer Declare Function GSLogLog Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fCycleHt#, ByVal fBaseY#, ByVal fCycleWid#, ByVal fBaseX#, ByVal nMode%, ByVal nClr%) As Integer Declare Function GSMCIrRgn Lib "GSWDLL16.DLL" (ByVal nRgn%) As Integer 476 Declare Function GSMean Lib "GSWDLL16.DLL" (ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSMGetX Lib "GSWDLL16.DLL" () As Double Declare Function GSMGetY Lib "GSWDLL16.DLL" () As Double Declare Function GSMinMax Lib "GSWDLL16.DLL" (ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSMMotion Lib "GSWDLL16.DLL" () As Integer Declare Function GSMNotify Lib "GSWDLL16.DLL" (ByVal hWnd%, ByVal nMsg%, ByVal nMode%) As Integer Declare Function GSMovePos Lib "GSWDLL16.DLL" (ByVal fxr#, ByVal fya#, ByVal nMode%) As Integer Declare Function GSMPtrOff Lib "GSWDLL16.DLL" () As Integer Declare Function GSMPtrOn Lib "GSWDLL16.DLL" () As Integer Declare Function GSMPtrType Lib "GSWDLL16.DLL" (ByVal nType%) As Integer Declare Function GSMSetRgn Lib "GSWDLL16.DLL" (ByVal fxr#, ByVal fya#, ByVal nMode%, ByVal fAng#) As Integer Declare Function GSMStatus Lib "GSWDLL16.DLL" () As Integer Declare Function GSOffView Lib "GSWDLL16.DLL" (ByVal nWin%, ByVal nView%) As Integer Declare Function GSOnView Lib "GSWDLL16.DLL" (ByVal nWin%, ByVal nView%) As Integer Declare Function GSOpenChildWin Lib "GSWDLL16.DLL" (ByVal hWndParent%, ByVal fxOrg#, ByVal fyOrg#, ByVal fWid#, ByVal fHt#, ByVal fyExt#, ByVal nStyle%, ByVal nMode%, ByVal szTitle$) As Integer Declare Function GSOpenPrn Lib "GSWDLL16.DLL" (ByVal szDevice$, ByVal szFile$, ByVal nMode%) As Integer Declare Function GSOpenServer Lib "GSWDLL16.DLL" (ByVal szKey$, ByVal szHost$) As Integer Declare Function GSOpenView Lib "GSWDLL16.DLL" (ByVal nWin%, ByVal fxOrg#, ByVal fyOrg#, ByVal fWid#, ByVal fHt#, ByVal fyExt#) As Integer Declare Function GSOpenWin Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fWid#, ByVal fHt#, ByVal fyExt#, ByVal nStyle%, ByVal nMode%, ByVal szTitle$) As Integer Declare Function GSPicRead Lib "GSWDLL16.DLL" (ByVal fxBL#, ByVal fyBL#, ByVal fxTR#, ByVai fyTR#, ByVal nFormat%, ByVal nMode%, ByVal szFile$) As Integer Declare Function GSPicWrite Lib "GSWDLL16.DLL" (ByVal fxBL#, ByVal fyBL#, ByVal fxTR#, ByVal fyTR#, ByVal nFormat%, ByVal nMode%, ByVal szFile$) As Integer Declare Function GSPie2D Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fRad#, ByVal nMode%) As Integer Declare Function GSPie3D Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fRad#, ByVal fDepth#, ByVal fAng#, ByVal nMode%) As Integer 477 Declare Function GSPolar Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fAng#, ByVal nMode%, ByVal nClr%) As Integer Declare Function GSPolarAxes Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fRad#, ByVal nRadDivs%, ByVal nAngDivs%, ByVal nMode%, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSPolyFill Lib "GSWDLL16.DLL" (ByVal fxr#, ByVal fya#, ByVal nMode%, ByVal fAng#, ByVal nPatt%, ByVal nClr%) As Integer Declare Function GSPolyVec Lib "GSWDLL16.DLL" (ByVal fxr#, ByVal fya#, ByVal nMode%, ByVal fAng#, ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSPrnOut Lib "GSWDLL16.DLL" (ByVal nView%, ByVal nNCopies%, ByVal nMode%) As Integer Declare Function GSPrnSetup Lib "GSWDLL16.DLL" (ByVal fxBL#, ByVal fyBL#, ByVal fxTR#, ByVal fyTR#, ByVal nUnits%, ByVal nMode%) As Integer Declare Function GSRText Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal nCset%, ByVal nTMode%, ByVal nClr%, ByVal szStringS) As Integer Declare Function GSScatter Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal nMode%, ByVal nClr%) As Integer Declare Function GSSD Lib "GSWDLL16.DLL" (ByVal nStyle%, ByVal nClr%) As Integer Declare Function GSSelectPalette Lib "GSWDLL16.DLL" (ByVal nMode%) As Integer Declare Function GSSetBG Lib "GSWDLL16.DLL" (ByVal nClr%) As Integer Declare Function GSSetPal Lib "GSWDLL16.DLL" (ByVal nClr%, ByVal nR%, ByVal nG%, ByVal nB%) As Integer Declare Function GSSetBG Lib "GSWDLL16.DLL" (ByVal nClr%) As Integer Declare Function GSSetRFontFace Lib "GSWDLL16.DLL" (ByVal nFamily%, ByVal szFaceName$) As Integer Declare Function GSSetROP Lib "GSWDLL16.DLL" (ByVal nROP%) As Integer Declare Function GSSetVFontFace Lib "GSWDLL16.DLL" (ByVal nFamily%, ByVal szFaceNameS) As Integer Declare Function GSShade Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal nPatt%, ByVal nClr%) As Integer Declare Function GSSizeSymbol Lib "GSWDLL16.DLL" (ByVal fDiam#) As Integer Declare Function GSStatsArr Lib "GSWDLL16.DLL" (ByVal nlndex%) As Integer Declare Function GSStatsWin Lib "GSWDLL16.DLL" (ByVal fxBL#, ByVal fyBL#, ByVal fxTR#, ByVal fyTR#) As Integer Declare Function GSSymbol Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal nSymbol%, ByVal nClr%) As Integer Declare Function GSTapeGraph Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fDepth#, ByVal fAng#, ByVal nMode%, ByVal nClr%) As Integer Declare Function GSTimeGraph Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal nPts%, ByVal nGroups%, ByVal nMode%) As Integer 478 Declare Function GSTimellpdate Lib "GSWDLL16.DLL" (ByVal nMode%, ByVal nGroup%, fData#) As Integer Declare Function GSUseView Lib "GSWDLL16.DLL" (ByVal nWin%, ByVal nView%) As Integer Declare Function GSVText Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fWid#, ByVal fHt#, ByVal fAng#, ByVal nCset%, ByVal nMode%, ByVal nClr%, ByVal szStringS) As Integer Declare Function GSWinHandle Lib "GSWDLL16.DLL" (ByVal nWindow%) As Integer Declare Function GSWinNotify Lib "GSWDLL16.DLL" (ByVal hWnd%, ByVal nWM%, ByVal nEvents) As Integer Declare Function GSWinPaint Lib "GSWDLL16.DLL" (ByVal nMode%) As Integer Declare Function GSXDataScale Lib "GSWDLL16.DLL" (ByVal fScale#) As Integer Declare Function GSXYGraph Lib "GSWDLL16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal nMode%, ByVal nClr%) As Integer Declare Function VBAGDataLabels Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nLabs%, ByVal szLabs$) As Integer Declare Function VBAGLabels Lib "GSWAG16.DLL" (ByVal nNLabs%, ByVal szLabs$) As Integer Declare Function VBAGLabelY Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nLabs%, ByVal szLabs$) As Integer Declare Function VBAGLabelZ Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nLabs%, ByVal szLabs$) As Integer Declare Function VBAGLegend Lib "GSWAG16.DLL" (ByVal nNLegs%, ByVal szLegsS) As Integer Declare Function VBGSDataLabels Lib "GSWAG16.DLL" (ByVal nMode%, ByVal nPrec%, ByVal nCset%, ByVal nTMode%, ByVal nClr%, ByVal fOff#, ByVal nLabs%, ByVal szLabs$) As Integer Declare Function VBGSLabelPie Lib "GSWAG16.DLL" (ByVal fxOff#, ByVal fRad#, ByVal fWid#, ByVal fHt#, ByVal nNLabs%, ByVal nMode%, ByVal nCset%, ByVal nTMode%, ByVal nClr%, ByVal szLabs$) As Integer Declare Function VBGSLabelX Lib "GSWAG16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fWid#, ByVal fHt#, ByVal nNLabs%, ByVal nCset%, ByVal nMode%, ByVal nClr%, ByVal szLabs$) As Integer Declare Function VBGSLabelY Lib "GSWAG16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fine#, ByVal fWid#, ByVal fHt#, ByVal nNLabs%, ByVal nCset%, ByVal nMode%, ByVal nClr%, ByVal szLabs$) As Integer Declare Function VBGSLegend Lib "GSWAG16.DLL" (ByVal fxOrg#, ByVal fyOrg#, ByVal fWid#, ByVal fHt#, ByVal nNLeg%, ByVal nRows%, ByVal nMode%, ByVal nCset%, ByVal nTMode%, ByVal nTCIr%, nBCIr%, nBPatt%, ByVal szLegs$) As Integer 479 Subroutine to switch from the d5307 chromatograms screen to the d5307 opening screen. Sub Done_Click () Unload d5307chro Load d5307open d5307open.Show d5307open.astmd5307. Enabled = -1 d5307open.astmd5307ex. Enabled = -1 d5307open.d5307cuts. Enabled = -1 d5307open.d5307displ. Enabled = -1 d5307open.sampisblank.Enabled = -1 d5307open.cal. Enabled = -1 d5307open.d5307diff. Enabled = -1 End Sub Subroutine to activate the chromatogram zooming option. Sub GraphJHotHit (HITSET As Integer, hitpoint As Integer) Graph.ThisSet = HITSET Graph.ThisPoint = hitpoint text3.Text = Val(Graph.XPosData) text4.Text = Val(Graph.GraphData) End Sub Subroutine gets activated when the mouse left button is activated. This routine returns the present coordinates of the left mouse. Sub Graph_MouseDown (Button As Integer, Shift As Integer, X As Single, Y As Single) 'get and store the begining coordinates for the begining of the box If Button = 1 Then 'get the axis position coodinates xaxislength = Graph.SDKInfo(5) yaxislength = Graph. SDKInfo(6) xorigin = Graph. SDKInfo(6) yorigin = Graph. SDKInfo(8) 'transform between form coordinates and logical view units x1 = 1000 * X / Graph. Height y1 = 1000 * (Graph. Height - Y) / Graph. Height 'Draw only within coordinate axis boundaries If x1 >= xorigin And x1 <= xorigin + xaxislengh Then If y1 >= yorigin And y1 <= yorigin + yaxislength Then newbox = 1 'save the initial position xmin = (x1 - xorigin) * (Graph.SDKInfo(1) - Graph.SDKInfo(2)) / xaxislength 480 ymax = (y1 - yorigin) * (Graph.SDKInfo(3) - Graph.SDKInfo(4)) / xaxislength Else badpoint = True End If badpoint = True End If End If End Sub Subroutine gets activated when the mouse left button is moved along the graph. This routine returns the present coordinates of the mouse on the screen as it moves along. Sub Graph_MouseMove (Button As Integer, Shift As Integer, X As Single, Y As Single) If Button = 1 Then 'if left button is pressed If badpoint = False Then If newbox Then 'perform coordinate transform for second box point x2 = 1000 * X / Graph.Height y2 = 1000 * (Graph.Height) / Graph.Height ’draw newbox ’top a% = GSLineAbs(x1, yi. x2, yi. o, 2, 15) ’left a% = GSLineAbs(x1, y2, x1, yi. o, 2, 15) 'right a% = GSLineAbs(x2, y2, x1, y2, o, 2, 15) 'bottom a% = GSLineAbs(x2, y1, x2, y2, o, 2, 15) newbox = 0 Else 'get axis position coordinates xaxislength = Graph. SDKInfo(5) yaxislength = Graph. SDKInfo(6) xorigin = Graph.SDKInfo(7) yorigin = Graph. SDKInfo(8) 'transform between form and logical view units xtemp = 1000 * X / Graph.Height ytemp = 1000 * (Graph.Height - Y) / Graph.Height 'draw only within coordinate axis boundaries If xtemp >= xorigin And xtemp <= xorigin + xaxislength Then If ytemp >= yorigin And ytemp <= yorigin + yaxislength Then 'xor out the previous box 'top 481 a% = GSLineAbs(x1, y1, x2, y1, 0, 2, 15) ’left a% = GSLineAbs(x1, y2, x1, y1, 0, 2, 15) 'right a% = GSLineAbs(x2, y2, x1, y2, 0, 2, 15) 'bottom a% = GSLineAbs(x2, y1, x2, y2, 0, 2, 15) 'perform coordinate transform for second box point x2= 1000 *X/Graph.Height y2 = 1000 * (Graph.Height - Y) / Graph.Height 'set graphic server xor drawing mode r% = GSSetROP(2) 'draw new box 'top a% = GSLineAbs(x1, y1, x2, y1, 0, 2, 15) 'left a% = GSLineAbs(x1, y2, x1, y1, 0, 2, 15) 'right a% = GSLineAbs(x2, y2, x1, y2, 0, 2, 15) 'bottom a% = GSLineAbs(x2, y1, x2, y2, 0, 2, 15) End If End If End If End If End If End Sub Subroutine gets activated when the mouse left button is released. This routine returns the present coordinates of the mouse. Sub Graph_MouseUp (Button As Integer, Shift As Integer, X As Single, Y As Single) If badpoint = False Then 'draw newbox 'top a% = GSLineAbs(x1, y1, x2, y1, 0, 2, 15) 'left a% = GSLineAbs(x1, y2, x1, y1, 0, 2, 15) 'right a% = GSLineAbs(x2, y2, x1, y2, 0, 2, 15) 'bottom a% = GSLineAbs(x2, y1, x2, y2, 0, 2, 15) 'save final poition xmax = (x2 - xorigin) * (Graph.SDKInfo(1) - Graph.SDKInfo(2)) / xaxislength 482 ymin = (y2 - yorigin) * (Graph.SDKInfo(3) - Graph.SDKInfo(4)) I yaxislength If xmax < xmin Then tmp% = xmax xmax = xmin xmin = tmp% End If If ymax < ymin Then tmp% = ymax ymax = ymin ymin = tmp% End If Graph.XAxisMax = xmax Graph.XAxisMin = xmin Graph.XAxisTicks = 5 Graph.XAxisStyle = 2 Graph.YAxisMax = ymax Graph.YAxisMin = ymin Graph.YAxisTicks = 5 Graph.YAxisStyle = 2 r% = GSSetROP(O) Graph. DrawMode = 2 Else badpoint = False End If End Sub Subroutinereset the graph. Sub reset_Click () Graph.XAxisStyle = 0 Graph.YAxisStyle = 0 Graph. DrawMode = 2 End Sub REFERENCES 1. Meyer, R. F., and J. M. Duford, "Resources of Heavy Oil and Natural Bitumen Worldwide," in.(1988). Proc. Fourth Unitar/UNDP International Conference on Heave Crude and Tar Sands. Meyer, R.F., and E. J. Wiggins, eds., Paper #147 (August 7-12,1988). 2. Speight, J.G., The Chemistry and Technology of Petroleum 2 ed., Marcel Dekker, Inc., New York, New York (1991). 3. “The Production of Oil from the Intermountain West Tar Sands Deposits” NTIS (National Technical Information Service). U.S. Dept, of Commerce, PB 256516, 9, 175(1979). 4. Ball, D., L. C. Marchant, and A. Goldbert, eds., “The IOCC Monograph Series: Tar Sands,” Interstate Oil Compact Commission, U of Oklahoma Press, Oklahoma City, OK. Norman, OK, (1982). 5. “Major Tar Sands and Heavy Oil Deposits of the United States,” Interstate Oil Compact Commission, U of Oklahoma Press, Norman, Oklahoma (1984). 6. Oblad, A. G., J. W. Bunger, F. V. Hanson, J. D. Miller, H. R. Ritzma, and J. D. Seader, "Tar Sand Research and Development at the University of Utah," Ann. Rev. Energy, 12, 283 (1987). 7. Sepulveda, J. E., "Hot Water Separation of Bitumen from Utah Tar Sands," M.S. Thesis, University of Utah, Salt Lake City (1977). 8. Sepulveda, J. E., and J. D. Miller, "Extraction of Bitumen from Utah Tar Sands by a Hot Water Digestion-Floatation Technique," Mining Eng., 30, 1311 (1978). 9. Miller, J. D., and M. Misra, "Hot Water Process Development for Utah Tar Sands," Fuel Proc. Tech., 6, 27 (1982). 10. Venkatesan, V. N., "Fluid Bed Thermal Recovery of Synthetic Crude from Bituminous Sands of Utah," Ph.D. Dissertation, University of Utah, Salt Lake City (1980). 484 11. Dorius, J. C., "The Pyrolysis of Bitumen-Impregnated Sandstone from the PR Spring (Utah) Deposit in a Fluidized Bed," Ph.D. Dissertation, University of Utah, Salt Lake City (1985). 12. Cha, S., "Pyrolysis of Oil Sands from the Whiterocks Tar Sand Deposit in a Rotary Kiln," Ph.D. Dissertation, University of Utah, Salt Lake City (1991). 13. Hanson, F.V., Personnel Communication, February, 1994. 14. Penner, S.S., S. W. Benson, F. W. Camp, J. Clardy, J. Deutch, A. E. Kelley, A. E. Lewis, F. X. Mayer, A. G. Oblad, R. P. Sieg, W. C. Skinner, and D. D. Whitehurst, "Assessment of Research Needs for Oil Recovery from Heavy-oil Sources and Tar Sands," New Sources of Oil and Gas: Gases from Coal: Liquid Fuel from Coal. Shale Tar Sands, and Heavy Oil Sources. Penner, S.S and 24 other members, eds., Pergamaon Press, Oxford, U. K„ 69(1982). 15. Penner, S.S., S. W. Benson, F. W. Camp, J. Clardy, J. Deutch, A. E. Kelley, A. E. Lewis, F. X. Mayer, A. G. Oblad, R. P. Sieg, W. C. Skinner, and D. D. Whitehurst, "Research Need for Shale Oil Recovery," New Sources of Oil and Gas: Gases from Coal: Liquid Fuel from Coal. Shale Tar Sands, and Heavy Oil Sources. Penner, S.S and 24 other members, eds., Pergamaon Press, Oxford, U. K., 27 (1982). 16. Rhodes, A. K., “Brent Blend, U.K. North Sea Marker Crude, Assayed,” Oil & Gas J., 93, 63 (1995). 17. Clark, H. A., and D. S. Pasternack, "Hot Water Separation of Bitumen from Alberta Bituminous Sands," Ind. Eng. Chem., 24, 1410 (1932). 18. Schumacher, M. M., Heavy Oil and Tar Sands Recovery and Upgrading, Noyes Data Corporation, Park Ridge, New Jersey (1982). 19. “AOSTRA; A 15-year Portfolio of Achievement,” Alberta Oil Sands Technology and Research Authority (1990). 20. Shephard, D. W., "Steam Stimulation of Cold Lake Bitumen," Proc. First Unitar/UNDP International Conference on Heave Crude and Tar Sands, Paper #22 (June 4-9, 1979). 21. McRory, R. E., “Energy Heritage-Oil Sands and Heavy Oils of Alberta,” Prepared for Alberta Energy and Natural Resources, Edmonton, Alberta, Canada (1982). 485 22. Huang, E. T. S., and J. H. Tracht, “Displacement of Residual Oil by C 02,” Paper presented at the 1974 SPE Symposium on Improved Oil Recovery, Tulsa, Oklahoma, SPE Paper #4735 (April 22-24, 1974). 23. Duerkson, J.H., and A. Eloyan, “Evaluation of Solvent-Based In-situ Processes for Upgrading and Recovery of Heavy Oil and Bitumen,” Proc. Sixth Unitar/UNDP International Conference on Heave Crude and Tar Sands, Houston, Texas, Paper #122 (Feb 11-14, 1995). 24. Chu, C., "State-of-the-Art Review of Steamflood Field Projects," JPT, 37, 1887 (1985). 25. Joshi, S., M. C. Wood, L. M. Castanier, and Brigham., “Steamflooding a Waterflooded Reservoir-Performance and Evaluation,” Proc. Sixth Unitar/UNDP International Conference on Heave Crude and Tar Sands, Houston, Texas, Paper #37 (Feb 11-14, 1995). 26. Roychoudray, S., S. Sur, S. K. Sinha, N. S. Rao, A. V. Sapkal, and S. Prakash, “Experience with In-situ Combustion Pilot in Presence of Edge Water,” Proc. Sixth Unitar/UNDP International Conference on Heave Crude and Tar Sands, Houston, Texas, Paper #154 (Feb 11-14, 1995). 27. Camp, F. W., “Tar Sands,” Kirk-Othmer Encyclopedia of Chemical Technology, 3 ed., Interscience Publishers, New York, 22, 601 (1982). 28. Hupka, J., A. G. Oblad, and J. D. Miller, "Diluent Assisted Hot Water Processing of Tar Sands," AOSTRA J. Res., 3, 95 (1987). 29. Funk, E. W., "Behavior of Tar Sand Bitumen with Paraffinic Solvents and Its Application to Separations for Athabasca Tar Sands," Can. J. Chem. Eng., 57, 333 (1979). 30. Parks, B. D., and F. W. Meadus, "A Combined Solvent Extraction and Agglomeration Technique for the Recovery of Bitumen from Tar Sands," Proc. 27th Can. Chem. Eng. Conference on Oil Sands, Ottawa, Ontario, 45 (1977). 31. Sparks, B. D., and F. W. Meadus, "A Combined Solvent Extraction and Agglomeration Technique for the Recovery of Bitumen from Tar Sands," Energy Proc., Can., 72, 55 (1979). 32. Sparks, B. D., and F. W. Meadus, "A Study of Some Factors Affecting Solvent Losses in the Solvent Extraction-Spherical Agglomeration of Oil Sands," Fuel Process. Technol., 4, 251 (1981). 486 33. Lynch, K. Z., R. I. Hood, 0. G. Maraven, and L. G. Aquino, “Rose - A Flexible Process for Upgrading Heavy Crude and Atmospheric or Vacuum Residue,” Proc. Sixth Unitar/UNDP International Conference on Heave Crude and Tar Sands, Houston, Texas , Paper #27 (Feb 11-14,1995). 34. Sprague, S., K. Washimi, and M. Ishibashi, "Upgrading Cold Lake Bitumen Using the HSC-ROSE Process," Paper presented in the 1986 Tar Sands Symposium (1986). 35. Sanford, E. C., and K. H. Chung, “The Mechanism of Pitch Conversion During Coking, Hydrocracking and Catalytic Hydrocracking of Athabasca Bitumen,” AOSTRA J. Res., 7, 37 (1991). 36. Weiss, H., and Schmalfeld, “Coking of Residue Oil by LP-Process," Erdol Kohle, 42, 235 (1989). 37. Nelson, L.R., M. MacKinnon, and J. R. Gulley, “Environmental Issues on Reclamation of Oil Sand Fine Tails," Proc. Sixth Unitar/UNDP International Conference on Heave Crude and Tar Sands, 1995, Houston, Texas , Paper #163 (Feb 11 -14,1995). 38. McHugh, M. A., and V. J. Krukonis, Supercritical Fluid ExtractionPrinciples and Practice, Butterworths, Boston, (1993). 39. Hwang, J., "Supercritical Fluid Extraction of Complex Hydrocarbon Mixtures Using C02 and Propane," Ph.D. Dissertation, University of Utah, Salt Lake City (1992). 40. Subramanian, M., "Application of Supercritical Fluid Extraction to Upgrading Oil Sand Bitumens," M.S. Thesis, University of Utah, Salt Lake City (1994). 41. Peng, D. Y., and D. B. Robinson, "A New Two-Constant Equation of State," Ind. Eng. Chem. Fundam., 15, 59 (1976). 42. Whitson, C.H., "Characterizing the Hydrocarbon Plus Fractions," SPEJ, 23, 683 (1983). 43. Lee, B. I., and M. G. Kesler, "A Generalized Thermodynamic Correlation Based on Three-Parameter Corresponding States,” AICHE J., 21, 510 (1975). 487 44. Bukka, K., F. V. Hanson, J. D. Miller, and A. G. Oblad, "Fractionation and Characterization of Whiterocks Tar Sand Bitumen," Energy & Fuels, 6, 160(1992). 45. Subramanian, M., M. D. Deo, and F. V. Hanson, “Compositional Analysis of Bitumen and Bitumen-Derived Products," J. Chromo. Sci., 34, 20 (1996). 46. Kehlen, H., M. T. Ratzsch, and J. Bergmann, "Continuous Thermodynamics of Multicomponent System,” AICHE J., 31, 1136 (1985). 47. Cotterman, R. L., Phase Equilibria for Complex Fluid Mixtures at High Pressures. Development and Application of Continuous Thermodynamics," Ph.D. Dissertation, Chapter 2, 3. University of California, Berkeley (1985). 48. Cotterman, R. L., “Phase Equilibria for Complex Fluid Mixtures at High Pressures. Development and Application of Continuous Thermodynamics,” Ph.D. Dissertation, Chapter 6, 7. University of California, Berkeley (1985). 49. Paulaitis, M. E., V. J. Krukonis. R. T. Kurnik. and R. C. Reid, "Supercritical Fluid Extraction," Reviews in Chemical Engineering, 1(2), 181 (1983). 50. Levelt Sengers, J. M. H., “Critical Behavior of Fluids: Concepts and Applications,” Supercritical Fluids: Fundamentals for Application, Kiran E., and J. M. H. Levelt Sengers, eds., Kluwer Academic Publishers, Dordrecht, The Netherlands, 3 (1994). 51. Irani, C. A., and F. W. Funk, Recent Developments in Separation Science, Vol. Ill Part A, CRC Press, West Palm Beach, Florida, 171 (1977). 52. Zosel, K., “Separation with Supercritical Gases: Practical Applications,” Angew Chem. Int. Ed. Engl., 17, 702 (1978). 53. Ewald, A. H., “The Solubility of Solids in Gases: Part II,” Trans. Faraday Soc., 49, 1401 (1953). 54. Paul, P. F. M., and W. S. Wise, Principle o f Gas Extraction, Boon and Boon ltd., London, U. K. (1971). 55. Eisenbeiss, J., "A Basic Study of the Solubility of Solids in Gases at High Pressures,” Final Report Contract No. DA-18-108-AMC-244(A), Southwest Research Institute, San Antonio, Texas (1964). 488 56. Prausnitz, J. M., Molecular Thermodynamics of Fluid-Phase Equilibria, Prentice-Hall, Engelwood Cliffs, New Jersey (1986). 57. Liphard, K. G., and G. M. Schneider, “Phase Equilibrium and Critical Phenomena in Fluid Mixtures of C 02 + 2, 6, 10, 15, 19, 23 Hexamethyltetracosane up to 423 K and 100 MPa,” J. Chem. Thermo., 7, 805 (1975). 58. McKinley, C., J. Brewer, and E. S. Wang, “Solubility of Solid Oxygen in Compressed Hydrogen,” J. Advan. Cryog. Eng., 7, 114 (1961). 59. . Rowlinson, J. S., and M. J. Richardson, “The Solubility of Solids in Compressed Gases,” Advan. Chem. Phys., 2, 85 (1959). 60. Williams, D. F., "Extraction with Supercritical Gases," Chem. Eng. Sci., 56, 1769(1981). 61. Eckert, C. A., J. G. Van Alsten, and T. Stoicos, "Supercritical Fluid Processing," Environ. Sci. Technol., 20, 319(1986). 62. Hoyer, G. G., “Extractions with Supercritical Fluids: Why, How and So What,” CHEMTECH , 15, 440 (1985). 63. Brennecke, J. F., and C. A. Eckert, “Phase Equilibria for Supercritical Fluid Process Design”, AICHE J., 35, 1409 (1989). 64. Chemical Engineering at Supercritical Fluid Conditions, Paulaitis, M.E., J. M. Penninger, R. D. Gray, Jr., and P. Davidson, eds., Ann Arbor Science, Ann Arbor, Michigan (1983). 65. Roselius, W., O. Vitzthum, and P. Hubert, “Methods for Production of Coffiene-Free Coffee Extract”, U.S. Patent 3,843,824 (1974). 66. Margolis, G., and J. Chiovini, “Decaffineation Process”, U.S. Patent 4,251,559(1981). 67. Hubert, P., and O. G. Vitzthum, “Fluid Extraction of Hops, Spices and Tobacco with Supercritical Gases,” Angew. Chem. Int. Ed. Engl., 17, 710, (1978). 68. Roselius, W., and P. Hubert, “Process for Extraction of Nicotine from Tobacco,” U.S. Patent 4,153,063 (1979). 489 69. Friedrich., J. P., G. R. List, and A. J. Heakin, “Petroleum Free Extraction of Oil From Soybeans with Supercritical C 02,” Paper presented at the 72nd Ann. Am. Chem. Soc. Meeting, New Orleans (May, 1981). 70. Schultz, W. G., and J. M. Randall, “Liquid C 02 for Selective Aroma Extraction,” Food Technology., 24, 94 (1970). 71. Akgerman, A., R. L. Roop, R. K. Hess, S. D. Yeo, “Supercritical Extraction in Environmental Control,” Supercritical Fluid Technology, Penninger, J. M. L., M. Radosz, M. A. McHugh, and V. J. Krukonis, eds., 14, 479, (1993). 72. Roop, R.L., R. K. Hess, and A. Akgerman, “Supercritical Fluid Extraction of Pollutants from Water and Soil,” Supercritical Fluid Science and Technology, Johnston, K.P., and Penninger, J. M. L., eds., 29, 468 (1989). 73. Akgerman, A., and S. D. Yeo, “Supercritical Extraction of Organic Components from Aqueous Slurries," Supercritical Fluid Engineering Science, Kiran, E., and Brennecke, J. F., eds.,23, 294 (1993). 74. Madras, G., C. Thiabaud, C. Erkey, and A. Akgerman, “Modeling of Supercritical Extraction of Organics from Solid Matrices,” AICHE J., 40, 111 (1994). 75. McNally, M. E. O., “Advances in Environmental SFE,” Anal. Chemistry, 67, 308(1995). 76. Worthy, W., "Supercritical Fluids offer Improved Separations,” Chem. Eng. News, 59, 1617 (1981). 77. Hunter, E., and R. B. Richards, “Fractionation of Polymeric Ethylene”, U.S. Patent 2,457, 238 (1948). 78. Copelin, H. B., “Method for Reducing Oligomeric Cyclic Ether Content of a Polymerizate,” U.S. Patent 4,306,058 (1981). 79. de Philippi, R. P., V. J. Krukonis, R. J. Robey, and M. Modell, “Supercritical Regeneration of Activated Carbon for Adsorption of Pesticides”, EPA-600/2-80-054, March (1980). 80. Dean, J. R., Application o f Supercritical Fluids in Industrial Analysis, Blackie Academic & Professional, London, U K, (1993). 490 81. Wilson, I. D., P. Davis, and R. J. Ruane, “Supercritical Fluid Chromatography and Extraction of Pharmaceuticals,” Application of Supercritical Fluids in Industrial Analysis, Dean, J.R., ed., Blackie Academic & Professional, London, U.K74 (1993). 82. Knowles, D. E., and T. K. Hoge, “Application of Supercritical Fluids in Polymer Analysis,” Application o f Supercritical Fluids in Industrial Analysis, Dean, J.R., ed., Blackie Academic & Professional, London, U K 104 (1993). 83. Dean, J. R., “Application of Supercritical Fluids in Food Science," Application o f Supercritical Fluids in Industrial Analysis, Dean, J.R., ed., Blackie Academic & Professional, London, U K 130 (1993). 84. Janda, V., K Bartle, and A. A. Clifford, “Supercritical Fluid Extraction in Environmental Analysis,” Application o f Supercritical Fluids in Industrial Analysis, Dean, J.R., ed., Blackie Academic & Professional, London, U.K 158 (1993). 85. Fong, W. S., P. C. F. Chan, P. Pichaichanarong, W. H. CorCoran, and D. D. Lawson, “Experimental Observations on a Systematic Approach to Supercritical Extraction of Coal," Chemical Engineering at Supercritical Fluid Conditions, Paulaitis, M. E., J. M. Penninger, R. D. Gray, Jr. and P. Davidson, eds., Ann Arbor Science, Ann Arbor, Michigan, 377 (1983). 86. Bott, T. R., "Supercritical Gas Extraction," Chem. and Ind., 15, 228 (1980). 87. Shelton, J. L., and L. Yarborough, "Multiple Phase Behavior in Porous Media During C02 or Rich Gas Flooding,” JPT, 29, 1171 (1977). 88. Cheng, H., J. A. Zollweg, and W. B. Streett, “Experimental Measurement of Supercritical Fluid-Liquid Phase Equilibrium,” Supercritical Fluid Science and Technology, Johnston, K. P., and J. M. L. Penninger, eds., 86 (1989). 89. Zou, M., S. B. Lim, S. S. H. Rizvi, and J. A. Zollweg, “Vapor Liquid Equilibria of Fatty Acid Esters in Supercritical Fluids,” Supercritical Fluid Science and Technology, Johnston, K. P., and J. M. L. Penninger, eds., 98 (1989). 90. Ramirez, L. P., C. L. Galeana, and C. T. Medina, “Application of SPHCT Model to the Prediction of Phase Equilibrium in C 02-Hydrocarbon Systems,” Fluid Phase Equilibria, 70, 1 (1991). 491 91. Patten, C. L., S. H. Kisler, and K. D. Luks, “Multiphase Equilibrium Behaviour of a Mixture of Carbon Dioxide, 1-Decanol, and NTetradecane,” Supercritical Fluid Engineering Science, Kiran, E., and J. F. Brennecke, eds., 55 (1993) 92. Jennings, D.W., M. T. Gude, and A. S. Teja, “High Pressure Vapor-Liquid Equilibria in Carbon Dioxide and 1-Alkanol Mixtures,” Supercritical Fluid Engineering Science, Kiran, E., and J. F. Brennecke, eds., 10 (1993). 93. Elgin, J. C., and J. J. Weinstock, “Phase Equilibrium at Elevated Pressures in Ternary Systems of Ethylene and Water with Organic Liquids,” J. Chem. Eng. Data, 4, 3 (1959). 94. Paulaitis, M. E., M. L. Gilbert, and C. A. Nash, “Separation of EthanolWater Mixtures with Supercritical Fluids,” Paper presented at the 2nd World Congress of Chemical Engineering, Montreal, Canada (October 4­ 9, 1981). 95. Huie, N. C., K. D. Luks, and J. P. Kohn, “Phase Equilibrium Behaviour of Systems C 02-n-Eiocene and C 02-n-Decane-n-Eiocene,” J. Chem Eng. Data, 18, 311 (1973). 96. Kulkarni, A. A., K. D. Luks, and J. P. Kohn, “Phase-Equilibria Behaviour of Systems C 02-2-Methylnaththalene and C 02-n-Decane-2-MethylNaphthalene,” J. Chem. Eng. Data, 19, 349 (1974). 97. Zarah, B. Y., K. D. Luks, and J. P. Kohn, “Phase-Equilibria Behaviour of C 02 in Binary and Ternary Systems with Several Hydrocarbon Components,” AICHE Symp. Series, 140, 91 (1974). 98. Orr, F. M., Jr., and M. K. Silva, "Equilibrium Phase Compositions of C02/Crude Oil Mixtures Part 1: Measurement by a Continuous Multiple Contact Experiment,” SPEJ, 23, 272 (1983). 99. Orr, F. M., Jr., M. K. Silva, and C. Lien, "Equilibrium Phase Compositions of C02/Crude Oil Mixtures Part 2: Comparison of Continuous Multiple Contact and Slim Tube Displacement Tests,” SPEJ, 23, 281 (1983). 100. Stalkup, F. I., “Carbon Dioxide Miscible Flooding: Past, Present and Outlook for the Future,” JPT, 30(8), 1102 (1978) 101. Yellig, W. F., and R. S. Metcalf, “Determination and Prediction of C 02 Minimum Miscibility Pressure”, JPT, 32, 160 (1980). 492 102. Orr, F. M., Jr., A. D. Yu, and C. Lien, "Phase Behaviour of C 02 and Crude Oil in Low-temperature Reservoirs," SPEJ, 24, 480 (1981). 103. Radosz, M., R. L. Cotterman, and J. M. Prausnitz, "Supercritical Fluid Extraction of Continuous Oil Mixtures," Ind. Eng. Chem., 26, 731 (1987). 104. Deo, M.D., J. Hwang, and F. V. Hanson, "Supercritical Fluid Extraction of a Crude Oil, Bitumen-Derived Liquid and Bitumen by Carbon Dioxide and Propane,” Fuel, 71, 1519 (1992). 105. Martin, T.G. and D. F. Williams, "Gaseous Solvent Extraction of Oil Shales and Tar Sands," U.S. Patent No. 4,108,760 (1978). 106. Pang, T. H., and E. McLaughlin, "Supercritical Extraction of Aromatic Hydrocarbon Solids and Tar Sand Bitumens," Ind. Eng. Chem. Process Des. Dev., 34, 1027 (1985). 107. Panzner, F., S. R. M. Ellis and T. R. Bott, “Alternatives to Distillation,” I. Chem. E. Symp. Series, 54, 165 (1978). 108. Compton, L. E., "Supercritical-Fluid Extraction of Oil from Tar Sands," NASA Tech. Brief from JPL Invention Report, 6, 41 (1981). 109. Jacoby, R. H., "Phase Behavior of Heavy Oils with Application to Reservoir Recovery Processes," Heavy Crude Oil Recovery, Okandan, E., ed., NATO AS I Series, (1988). 110 Gearhart, J. A., “Solvent Treat Resids,” Hydrocarbon Processing, 59(5), 150(1980). . 111 . Nelson, S. R., and R. G. Roodman, “ROSE: The Energy Efficient Bottom of the Barrel Alternative,” Paper presented at the 1985 Spring AICHE Meeting, Houston, Texas (March 24-28, 1985). 112 . Corbett, R. W., and S. R. Nelson, “Kerr-McGee’s ROSE Process Expands on Proven Extraction Technology," Proc. Paper at the Third International Conference on Heavy Crude and Tar Sands, 1822 (1985) 113. Sprague, S., K. Washimi, and M. Ishibashi, “High Conversion Upgrading of Extra Heavy Oils by a New Process Combination,” Proc. Third International Conference on Heavy Crude and Tar Sands, 3, 1495, (1985). 114. Penning, R. T., A. G. Vickers, and B. R. Shah, “Extraction Upgrades Resid,” Hydrocarbon Processing, 61, 145 (1982). 493 115. Billion A., J. P. Peries, E. Fehr, and E. Lorenz, “SDA Key to Upgrading Heavy Crudes,” Oil and Gas J., 43 (Jan. 24, 1977). 116. Alberty, R. A., and C. A. Gehrig, “Standard Chemical Properties of Alkane Isomer Groups,” J. Phys. Chem. Ref. Data, 13, 1173 (1984). 117. Lee, S. T., R. H. Jacoby, W. H. Chen, and W. E. Culham, “Experimental and Theoretical Studies on the Fluid Properties Required for Simulation of Thermal Processes," SPE J., 37, 535 (1981). 118. Pederson, K. S., P. Thomassen, and A. A. Fredenslund, “SRK-EOS Calculations for Crude Oils,” Fluid Phase Equilibria, 14, 209 (1983). 119. Soave, G., “Equilibrium Constants for Modified Redlich-Kwong Equation of State,” Chem. Eng. Sci., 27, 1197 (1972). 120. Li, Y. K., L. X. Nghiem, and A. Siu, “Phase Behavior Computation for Reservoir Fluids: Effect of Pseudocomponent on Phase Diagrams and Simulation Results,” Paper presented at the Petroleum Soc. of CIM Annual Meeting, Calgary, Paper #84-35-19 (June 10-13, 1984). 121. Mehra, A. K, R. A. Heidemann, K. Aziz, and J. K. Donnelly, “A Statistical Approach for Combining Reservoir Fluids into Pseudocomponents for Compositional Model Studies," Paper presented at the 1982 SPE Annual Technical Conference and Exhibition, New Orleans, Paper #11201 (Sept. 26-29, 1982). 122. Colonomous, P., E. Ganzalez, and I. Rusinek, “A New Approach for Characterizing Oil Fractions, and for Selecting Pseudocomponents for Hydrocarbons,” Paper presented at the Petroleum Soc. of CIM Annual Meeting, Calgary, Paper #84-35-27 (June 10-13, 1984). 123. Montel, F., and P. L. Gousel, “A New Lumping Scheme of Analytical Data for Composition Studies,” Paper presented at the 1984 SPE Annual Technical Conference and Exhibition, Houston, Paper #13119 (Sept. 16­ 19, 1984). 124. Riazi, M. R., and T. E. Daubert, “Characterization Parameters for Petroleum Fraction," Ind. Eng. Chem. Res., 26, 755 (1987). 125. Twu, C. H., "Prediction of Thermodynamic properties of Normal Paraffin using only Normal Boiling Point," Fluid Phase Equilibria, 11, 65 (1983) 494 126. Reid, R. C., J. M. Prausnitz, and B. E. Poling, The Properties o f Gases and Liquids 4 ed., McGraw Hill Inc., New York, New York (1987). 127. Bowman, J. R., “Distillation of and Infinite Number of Components,” Ind. Eng. Chem., 41, 2004 (1949). 128. Edmister, W. C., “Improved Integral Technique for Petroleum Distillation Calculation,” Ind. Eng. Chem., 47, 1685 (1955). 129. Aris, R., and G. R. Gavalas, “On the Theory of Reaction in Continuous Mixtures,” Phil. Trans. Royal Soc., A 260, 351 (1966). 130. Hoffman, E. J., “Flash Calculations for Petroleum Fraction,” Chem. Eng. Sc/'., 23, 957 (1968). 131. Vrij, A., “Light Scattering of a Concentric Multicomponent System of Hard Spheres in Percus-Yevick Approximation," J. Chem. Phys., 69, 1742 (1978). 132. Blum, L., and G. Stell, “ Polydisperse Systems -1 : Polydisperse Fluids of Hard or Permeable Spheres,” J. Chem. Phys., 71, 42 (1979). 133. Dickinson, E., “Statistical Theory of Fluid Phase Equilibrium in a Conformal Polydisperse System,” J. C. S., Faraday II, 76, 1458 (1980). 134. Smith. E. R., and R. L. Rowlinson, “Phase Separation in Polydisperse Systems,” J.C.S., Faraday 11, 76, 1468 (1980). 135. Salacuse, J. J., and G. Stell, “Polydisperse Systems: Statistical Thermodynamics, with Application to Several Models Including Hard and Permeable Sphere,” J. Chem. Phys., 77, 3714 (1982). 136. Briano, J. G., “Classical and Statistical Thermodynamics of Continuous Mixtures,” Ph.D. Dissertation, University of Pennsylvania, Philadelphia (1983). 137. Gualtieri, J. A., J. M. Kincid, and G. J. Morrison, “Phase Equilibria in Polydisperse Fluids,” J. Chem. Phys., 77, 121 (1982). 138. Percus, J. K., and G. J. Yevick, “Analysis of Classical Statistical Mechanics by Means of Collective Coordinates,” Phys. Rev., 110, 1 (1958). 139. Ratzach, M. T., and H. Kehlen, “Continuous Thermodynamics of Complex Mixtures,” Fluid Phase Equilibria, 14, 225 (1983). 495 140. Ratzach, M. T., and H. Kehlen, “Continuous Thermodynamics of Complex Mixtures,” AICHEJ., 17, 1324 (1983). 141. Kehien, H., and M. T. Ratzach, “Liquid Phase Separation in Polymer Systems and Polymer Compatability by Continuous Thermodynamics,” Z Phys. Cheme, Leipzig, 264, 1153 (1983). 142. Ratzach, M.T., and H. Kehlen, “ Separate Treatment of Paraffins and Aromatics in Complex Hydrocarbon Mixtures by Continuous Thermodynamics,” Z Phys. Cheme, Leipzig, 265, 1049 (1984). 143. Ratzach, M.T., and H. Kehlen, “Equilibrium Flash Vaporization Curves by Continuous Thermodynamics,” Z Phys. Cheme, Leipzig, 266, 329 (1985). 144. Kehlen, H., and M. T. Ratzach, “Complex Multicomponent Distillation Calculations by Continuous Thermodynamics,” Chem. Eng. Sci., 42, 221 (1987). 145. Ratzach, M. T., H. Kehlen, and J. Schumann, “Flash Calculations for a Crude Oil by Continuous Thermodynamics,” Chem. Eng. Comm., 71, 113 (1988). 146. Cotterman, R. L., R. Bender, and J. M. Prausnitz, “Phase Equilibria for Mixtures Containing Very Many Compounds. Development and Application of Continuous Thermodynamics for Chemical Process Design,” Ind. Eng. Process Des. Dev., 2, 194 (1984). 147. Cotterman, R. L., and J. M. Prausnitz, “Flash Calculation for Continuous and Semicontinuous Mixtures Using Equation of State," Ind. Chem. Process Des. Dev., 24, 434 (1985). 148. Cotterman, R. L., D. Dimitrelis, and J. M. Prausnitz, “Design of Supercritical-Fluid Extraction Process Using Continuous Thermodynamics," Supercritical Fluid Technology, Penninger, J.M.L., M. Radosz, M. A. McHugh, and V. J. Krukonis, eds., Elsevier Science Publishers, B.V., Amsterdam, The Netherlands 107 (1985). 149. Willman, B. T., and A. S. Teja, “Continuous Thermodynamics of Phase Equilibria Using a Multivariate Distribution Function and an Equation of State,” AICHEJ., 32, 2067 (1986). 150. Behrens, R. A., and S. I. Sandler, “The Use of Semicontinuous Description to Model the C7+ Fraction in Equation of State Calculation," SPE Res. Eng., 3, 1041 (1988). 496 151. Schultz, G. V., “Uber Die Kinetik Der Kettenpolymerisationen," Z Phys. Cheme, Leipzig, 43, 25 (1939). 152. Walpole, R.E., and R. H. Myers, Probability and Statistics for Engineers and Scientists, Macmillan Inc., New York, New York 154 (1985). 153. Devore, J.L., Probability and Statistics for Engineering and the Sciences, Second Edition, Brooks/Cole Pub. Co., Monterey, California, 1987. 154. Hammings, R.W., Numerical Methods for Scientists and Engineers, 2 ed., McGrawHill, New York, New York (1973). 155. Patel, N. C., and A. S. Teja, “A New Cubic Equation of State for Fluids and Fluid Mixtures,” Chem. Eng. Sci., 37, 463 (1982). 156. Peng, D. Y., R. S. Wu, and J. P. Batycky, “Application of Continuous Thermodynamics to Oil Reservoir Fluid Systems Using an Equation of State,” AOSTRA J. Res., 3, 113 (1987). 157. Haynes, Jr. H. W., and M. A. Matthews, “Continuous-Mixture Vapor-Liquid Equilibria Computations Based on True Boiling Point Distillations,” Ind. Eng. Chem. Res., 30, 1911 (1991). . 158. Angelos, C. P., S. V. Bhagwat, and M. A. Matthews, "Measurement and Modeling of Phase Equilibria with Synthetic Multicomponent Mixtures,” Fluid Phase Equilibria, 72, 189 (1992). 159. Mani, K. C., M. A. Matthews, and H. W. Haynes, Jr., “Continuous Approach Optimizes Vapor-Liquid Equilibrium Calculation,” Oil & Gas J., 91, 76, February 15 (1993). 160. Chakravarthy, D. S., and M. A. Matthews, “An Equation-of-State-Based Reservoir Model Incorporating Continuous Thermodynamics," Ind. Eng. Chem. Res., 33, 1962 (1994). 161. Ying, X., R. Ye, and Y. Hu, “Phase Equilibria for Complex Mixtures. Continuous -Thermodynamics Method Based on Spline Fit,” Fluid Phase Equilibria, 53, 407 (1989). 162. Du, P. C., and G. A. Mansoori, “A Continuous Mixture Computational Algorithm for Reservoir Fluids Phase Behaviour,” Paper presented at the 56th California Regional Meeting of the Society of Petroleum Engineers held at Oakland, CA, SPE Paper #15082 (April 2-4, 1986). 497 163. Grossman, S. I., and W. R. Derrick, Advanced Engineering Mathematics, Harper & Row Pub., New York, New York (1988) 164. Kendall, M. G., and A. Stuart, The Advanced Theory o f Statistics, Vol. 1. Charles Griffin & Company, Ltd., London U. K, (1958). 165. Davis, P. J., and P. Rabinowitz, Methods o f Numerical Integration, 2 ed., Academic Press, New York, New York (1984). 166. Lapidus, L., Digital Computations for Chemical Engineers, McGraw Hill, New York New York (1962). 167. Abramowitz, M., and I. A. Stegun, Handbook o f Mathematical Functions, Dover Publications, New York, New York (1972). 168. Chapman, W. G., “Theory and Simulation of Associating Fluid Mixtures," Ph.D. Dissertation, Cornell University, Ithaca (1988). 169. Chapman, W.G., K. E. Gubbins, G. Jackson, and M. Radosz, "SAFT: Equation of State Model for Associating Fluids,” Fluid Phase Equilibria, 52, 31 (1989). 170. Yu, J. M., S. H. Huang, and M. Radosz, “Phase Behaviour of Reservoir Fluids: Supercritical Carbon Dioxide and Cold Lake Bitumen,” Fluid Phase Equilibria, 54, 429 (1989). 171. Huang, S. H., and M. Radosz, “Phase Behavior of Reservoir Fluids: SAFT Model of C02 and Bitumen Systems," Fluid Phase Equilibria, 70, 33 (1991). 172. Beret, S., and J. M. Prausnitz, “Perturbed Hard Chain Theory: An Equation of State for Fluid Containing Small or Large Molecules,” AICHE J., 21, 1123(1975). 173. Barker, J. A., and D. Hederson, “What is Liquid? Understanding the States of Matter,” Rev. Mod. Phys., 48, 587 (1976). 174. Carnahan, N. F., and K. E. Starling, “Intermolecular Replusions and the Equations of State for Fluids,” AICHE J., 48, 1184 (1972). 175. Prigogine, I., Molecular Theory o f Solutions, North-Holland, Amsterdam, Netherlands (1957). 498 176. Mehrotra, A. K., M. Sarkar, and W. Y. Svrcek, "Bitumen Density and Gas Solubility Predictions Using the Peng-Robinson Equation of State," AOSTRA J., 1(4), 215(1985). 177. Lu, B. C.-Y., W. K. Chung, Y. Adachi, and T. Boublik, “Correlations and Prediction of Solubilities of Gases in Athabasca Bitumen,” AOSTRA J., 2, 141 (1986). 178. Jamaluddin, A. K. M., N. E. Kalogerakis, and A Chakma, “Predictions of C 02 Solubilty and C 02 Saturated Liquid Density of Heavy Oil and Bitumens Using a Cubic Equation of State,” Fluid Phase Equilibria, 64, 22 (1991). 179. Haung, P. K., "Characterization of Thermodynamic Correlations for Undefined Hydrocarbon Mixtures," Ph.D. Dissertation, The Pennsylvania State University, Philadelphia (1977). 180. Bergman, D., "Prediction of Phase Behavior of Natural Gas Pipelines," Ph.D. Dissertation, University of Michigan, Ann Arbor (1976). 181. Merhotra, A. K., and W. Y. Svrcek, “Correlations for Properties of Bitumen Saturated with C 02 and N2 and Experiments with Combustion Gas Mixtures,” J. Can. Pet Technol., 21, 95, (1982). 182. Fu, F. T., V. R. Puttagunta, and G. Vilcsak, “Vapor-Liquid Equilibrium Properties for Pseudo-Binary Mixtures of C 02-Athabasca Bitumen and N2Athabasca Bitumen,” Paper presented at the Intl. Chem. Congress of Pacific Basin Soc., Honolulu, HA, (Dec., 16-21, 1984). 183. Lin, H. M., “Peng-Robinson Equation of State for Vapor-Liquid Equilibrium Calculations for Carbon Dioxide + Hydrocarbon Mixtures,” Fluid Phase Equilibria, 16, 151 (1984). 184. Lai, F. D., F. D. Otto, and A. E. Mather, “Solubility of CO, H2, and CO-H2 Mixtures in Athabasca Bitumen,” Proc. 34th Can. Chem. Eng. Conf., Quebec, Canada, 150 (Sept., 30 - Oct. 3, 1984). 185. Svrcek, W. Y., and A. K. Mehrotra, “Gas Solubilty, Viscosity and Density Measurements for Athabasca Bitumen,” J. Can. Pet. Techno., 21, 31 (1982). 186. Martin, J. J., “Cubic Equation of State - Which?," Ind. Eng. Chem. Fundam., 18, 81 (1979). 499 187. Mehrotra. A. K., and W. Y. Svrcek, “Correlations and Predictions of Gas Solubility in Cold Lake Bitumen,” Can. J. Chem. Eng., 66, 666 (1988). 188. Kokal, S. L., and S. G. Sayegh, “Gas Saturated Bitumen Density Prediction Using the Volume Translated Peng-Robinson Equation of State,” Report 1988-14, Petroleum Recovery institute, Calgary, Canada (1988). 189. Huang, S. H., and M. Radosz, "Phase Behavior of Reservoir Fluids II: Supercritical Carbon Dioxide and Bitumen Fractions," Fluid Phase Equilibria, 60, 81 (1990). 190. Huang, S. H., and M. Radosz, "Phase Behavior of Reservoir Fluids III: Molecular Lumping and Characterization," Fluid Phase Equilibria,” 66, 1 (1991). 191. AOSTRA: Syncrude Analytical Methods for Oil Sand and Bitumens, Syncrude Canada Limited, Edmonton, Alberta, Canada (1979). 192. Bukka, K., J. D. Miller, and A. G. Oblad, “Fractionation and Characterization of Utah Tar Sands Bitumens. Influence of Chemical Composition on Bitumen Viscosity," Energy & Fuels, 5, 333 (1991) 193. ASTM D2892-90, Distillation o f Crude Petroleum (15-Theoretical Plate Column) (1990). 194. ASTM D5236-92, Distillation of Heavy Hydrocarbon Mixtures (Vacuum Potstill Method) (1992). 195. ASTM D2887-89, Boiling Range Distribution o f Petroleum Fractions by Gas Chromatography (1989). 196. ASTM D5307-92, Determination of Boiling Point Distribution o f Crude Petroleum by Gas Chromatography (1992). 197. Worman and L. E. Green, “Simulated Distillation of High Boiling Petroleum Fractions," Anal. Chem., 37, 1620 (1965). 198. Neer, L., and M. D. Deo, “Simulated Distillation of Oils with a Wide Carbon Number Distribution,” J. Chromo. Sci., 33, 133 (1995). 199. American Petroleum Institute Project No. 44 Selected Values o f Properties of Hydrocarbons and Related Compounds, TRC Thermodynamic Table, (1972). 500 200. Starling, K.E., Fluid Thermodynamic Properties for Light Petroleum Systems, Gulf Publishing Co. Houston, Texas (1973). 201. Floyd, R.H., Personal Communication, The M.W. Kellogg Company, Houston, TX, March, 1996. 202. Johnston, K.P., “New Directions in Supercritical Fluid Science and Technology,” Supercritical Fluid Science and Technology, Johnston K.P., and Penninger, J.M.L., eds., American Chemical Society, Washington, DC (1989) 203. Nellesteyn, F.J., “The Colloidal Structure of Bitumen,” The Science of Petroleum, Oxford U. press, New York City, 4, 2750 (1938). 204. Swanson, J., J Chemical Physics, 46, 141 (1943). 205. Witherspoon P. A., and Z. A. Munir, Producers Monthly, 24, 20 (1960). 206. Dickie, J. P., and T. F. Yen, “Macrostructures of the Asphaltenic Fractions by Various Instrumental Methods,” Anal. Chemistry, 39, 1847, (1967). 207. Koots, J. A., and J. C. Speight, “Relation of Petroleum Resins to Asphaltenes,” Fuel, 54, 179 (1975). 208. Mansoori, G. A., T. S. Jiang, and S. Kawanaka, “Asphaltene Deposition and its Role in Petroleum Production and Processing,” Arab. J. Sci. and Eng., 13, 17 (1988). 209. Mitchell, D. L., and J. G. Speight, “The Solubility of Asphaltenes in Hydrocarbon Solvents,” Fuel, 52, 149 (1973). 210. Speight, J. G., R. B. Long, and T. D. Trowbridge, “Factors Influencing the Separation of Asphaltenes from Heavy Petroleum Feedstocks,” Fuel, 63(5), 616(1984). 211. Edmister, W. C., and D. H. Pollock, “Phase Relations for Petroleum Fraction,” Chem. Eng. Prog., 44, 905 (1948). 212. Beyers, W. H., CRC Handbook for Mathematical Sciences, 3 ed., CRC Press, West Palm Beach, Florida (1978). 213. CMGPROP 95.00, Computer Maintenance Group, Calgary, Canada (1995). 501 214. Watson, K. M., and E. F. Nelson, "Improved Methods for Approximating Critical and Thermal Properties of Petroleum Fraction,” Ind. Eng. Chem., 25, 880(1933).