UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effect of oil compatibility and resins/asphaltenes ratio on heat exchanger fouling of mixtures containing… Al-Atar, Eman 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2000-0009.pdf [ 5.05MB ]
Metadata
JSON: 831-1.0058530.json
JSON-LD: 831-1.0058530-ld.json
RDF/XML (Pretty): 831-1.0058530-rdf.xml
RDF/JSON: 831-1.0058530-rdf.json
Turtle: 831-1.0058530-turtle.txt
N-Triples: 831-1.0058530-rdf-ntriples.txt
Original Record: 831-1.0058530-source.json
Full Text
831-1.0058530-fulltext.txt
Citation
831-1.0058530.ris

Full Text

Effect of Oil Compatibility and Resins / Asphaltenes Ratio on Heat Exchanger Fouling of Mixtures Containing Heavy Oil By E M A N A L - A T A R B.A.Sc, The University of British Columbia, 1997 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF T H E REQUIREMENTS FOR T H E DEGREE OF MASTER OF APPLIED SCIENCE in T H E F A C U L T Y OF G R A D U A T E STUDIES DEPARTMENT OF CHEMICAL AND BIO-RESOURCE ENGINEERING We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH COLUMBIA February 2000 © Eman Al-Atar, 2000 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. 1 further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of ( I L m rJL 9 r \ g i r\QOJn The University of British Columbia ^ Vancouver, Canada Date DE-6 (2/88) A B S T R A C T Fouling o f heat transfer equipment due to unwanted deposition of solids during heating remains a major cost penalty in oi l refineries. Severe fouling is encountered during the processing o f asphaltene-containing oils, and with increased reliance on heavy oils the situation has been exacerbated. Petroleum oils can be separated by solvent fractionation into saturates, aromatics, resins and asphaltenes with the latter having the highest molecular weight. Asphaltene precipitation from oils depends on the concentration of solvent components such as resins and aromatics. The available literature suggests that resins stabilize asphaltenes, minimizing their tendency to flocculate. This work was undertaken to determine how the asphaltene-resin interactions affect fouling. Fouling of asphaltenes from a heavy oil in mixtures of fuel oi l and de-asphalted vacuum bottoms ( D A O ) was studied at asphaltene concentration o f 0.04 - 3.4 %, and resin concentrations of 3.1 - 4.9 %. Experiments were performed at a bulk temperature of 85 °C, fluid velocity o f 0.75 m/s and pressure of 410 kPa. Fluids under nitrogen were recirculated through an annular test section with initial surface temperature of 230 °C for periods up to 30 hours, and the fouling monitored by thermal measurement. The effects of concentration of heavy oil and de-asphalted oil are explored. High fouling rates were encountered at high pentane insolubles (asphaltene) concentrations. Fouling rates are correlated with the ratio o f resins / asphaltenes. A t a fixed D A O concentration, the fouling rate first increased, and then decreased as the H O concentration was raised from zero to 20 % and as the Re/As ratio decreased. The maximum initial fouling rate occurred at a ratio of = 2.5 and dropped to essentially zero for Re/As ratio > ii 5.8. The initial fouling rate, hot filtration insolubles concentrations and pentane insolubles concentrations were found to increase as DAO concentration was raised at a fixed Re/As ratio. This was somewhat unexpected. Pentane insolubles concentrations also increased as D A O concentration was increased at a constant asphaltene concentration, which suggests that it is not only the asphaltenes in the heavy oil that precipitate in fuel oil / D A O mixtures. The relationship of fouling to oil compatibility as determined by the method of Wiehe, was explored. Fouling rates of mixtures containing DAO did not correlate with the colloidal instability index. A fouling regime map indicated that low fouling rates were dependent on both the colloidal instability index and the resin/asphaltene ratio. Oil Co mpatibility Model predictions correlated well with the colloidal instability index and therefore were unable to predict the fouling behaviour of the mixtures. However, the Oil Compatibility Model was found to be very sensitive to small errors in titrations. Oil Compatibility Model titrations showed that the addition of D A O to a heavy oil sample resulted in asphaltene precipitation at a lower heptane concentration and required a higher toluene concentration in a toluene-heptane mixture to keep asphaltene in solution. This finding was consistent with the measured effect of DAO on fouling. iii Table of Contents Abstract ii Table of Contents iv List of Tables vi List of Figures viii Acknowledgment xi 1.0 INTRODUCTION 1 2.0 LITERATURE REVIEW 3 2.1 Heat Exchanger Fouling of Asphaltene-Containing Oils 4 2.1.1 Petroleum Oils and Solvent Fractionation 5 2.1.2 Petroleum Asphaltenes 6 2.1.3 Formation of Micelles, Colloids, and Flocculates by Petroleum Asphaltenes 11 2.1.4 Chemistry of Resins 14 2.1.5 Role of Resins in Asphaltenes Stabilization 16 2.2 Deposit Formation by Petroleum Asphaltenes 19 2.2.1 Mechanisms of Deposit Formation 20 2.2.2 Modeling of Deposit Formation 23 2.3 Modeling Oil Compatibility and it Relation to Deposit Formation 26 2.3.1 Colloidal Instability Index 26 2.3.2 Oil Compatibility Model 27 2.4 Aims and Objectives of Work 31 3.0 EXPERIMENTAL MATERIALS AND APPARATUS 32 3.1 Experimental Materials 3 2 3.1.1 Properties of Heavy Oil 32 3.1.2 Properties of De-Asphalted Oil 33 3.1.3 Properties of Fuel Oil 37 3.1.4 Properties of Test Solutions 3 7 3.2 Experimental Apparatus 41 3.2.1 Thermal Fouling Test Apparatus 41 3.2.2 The Annular Test Section 42 4.0 EXPERIMENTAL PROCEDURES 46 4.1 Procedure for Thermal Fouling Runs 46 4.2 Determination of Pentane Insolubles and Hot Filtration Insolubles 47 iv 4.3 Measurement of Test Fluid Properties 48 4.4 Procedure for Oil Compatibility Tests 49 4.4.1 Heptane Dilution Test 49 4.4.2 Toluene Equivalence Test 50 4.4.3 Nonsolvent Oil Dilution Test 51 4.4.4 Solvent Oil Equivalence Test 51 5.0 RESULTS AND DISCUSSION 53 5.1 Typical Thermal Fouling Run 53 5.2 Effect of Resins to Asphaltenes Ration on Heavy Oil Fouling 53 5.3 Effect of Resins to Asphaltenes Ratio on Hot Filtration and Pentane Insolubles 63 5.4 Deposit Characterization 72 5.5 Colloidal Instability Index 79 5.6 Oil Compatibility Model Relation to Asphaltenes Fouling 82 5.6.1 Heavy Oil as Reference Oil 82 5.6.1.1 Test Results 82 5.6.1.2 Model Prediction 84 5.6.2 Heavy Oil - DAO Blend as Reference Oil 85 5.6.2.1 Test Results 85 5.6.2.2 Model Prediction 88 6.0 CONCLUSIONS AND RECOMMENDATIONS 89 6.1 Conclusions 89 6.2 Recommendations 90 Abbreviations 92 Nomenclature 92 References 95 Appendices 100 Appendix A l : Summary of Fouling Runs 100 Appendix A2: Sample Calculations 101 Appendix A3: Reproducibility of Thermal Fouling Experiments 109 Appendix A4: Viscosity Data 112 V List of Tables Table 2.1: Analysis of Fractions of Cold Lake Vacuum Resid 5 Table 2.2: Molecular Weights and Average Molecular Formulae of Cold Lake Bitumen and its Fractions 8 Table 2.3: Yields of Asphaltenes Precipitated from Western Canadian Bitumen Using Various Solvents 10 Table 2.4: Resin Fractions for Cold Lake Heavy Oil 15 Table 2.5: Summary of Resin Fractions Analyses of Study Done by Hammami and Co-workers 19 Table 3.1: Properties of Cold Lake Heavy Oil 34 Table 3.2: Generalized Ranges for the Bulk Fractions in Crude Petroleum, Heavy Oil, and Residua 35 Table 3.3: Properties of De-asphalted Oil 36 Table 3.4: Properties of Fuel Oil 38 Table 3.5: Composition and Properties of Test Solutions 39 Table 5.1: Thermal Fouling Parameters for Experiments of 5 wt % DAO in HO/FO Mixtures at T b of 85 °C, T s o of 230 °C and U bof0.75m/s 56 Table 5.2: Thermal Fouling Parameters for Experiments of 10 wt % DAO in HO/FO Mixtures at T b of 85 °C, T s o of 230 °C and U b of 0.75 m/s 58 Table 5.3: Thermal Fouling Parameters for Experiments of 15 wt % DAO in HO/FO Mixtures at T b o f 85 °C, T s o of 230 °C and U b of 0.75 m/s 59 Table 5.4: Properties of Test Fluids 65 Table 5.5: Elemental Analysis of Hot Filtration Insolubles 71 Table 5.6 Chemical Characteristics of Probe Deposits of Some Runs 72 Table 5.7: Compositions of Test Fluids 81 vi Table 5.8: Oil Compatibility Test Results using HO as The Reference Oil 84 Table 5.9: Calculated Oil Compatibility Model Parameters Using HO as The Reference 85 Table 5.10: Oil Compatibility Model Prediction for Test Fluids 86 Table 5.11: Oil Compatibility Test Results Using HO-D AO as The Reference Oil 88 Table 5.12: Calculated Oil Compatibility Model Parameters Using HO-D AO Blend as Reference 88 Table A l . 1: Summary of Fouling Runs 100 Table A2.1: Modeling Values of Initial Fouling Rates of All Mixtures 105 Table A2.2: Average SARA Analysis of Working Fluids. 106 Table A3.1 Test of Reproducibility of Data 109 Table A4.1 Kinematic Viscosity of a Mixture of 10% DAO, 10% HO and 80% FO at Various Temperatures 112 vii List of Figures Figure 2.1: Clay-Gel Percolating Column 7 Figure 2.2: A Hypothetical Asphaltene Molecule 9 Figure 2.3: Asphaltene Micelle Formation 13 Figure 2.4: Structures of Resins 15 Figure 2.5: Physical Model of Petroleum 16 Figure 2.6: Dependence of Asphaltene Solubility on Temperature 21 Figure 2.7: Oil Compatibility Numbers for Souedie and Forties Crudes 30 Figure 3.1: Dynamic Behavior of 15% DAO - 10% HO - 75% FO 40 Figure 3.2: Viscosity of 10% DAO - 10% HO - 80% FO at Different Temperatures 40 Figure 3.3: Schematic of Fouling Apparatus 43 Figure 3.4: Heat Exchanger Fouling Probe 44 Figure 5.1: Surface Temperature and Heat Flux for a Typical Fouling Run 54 Figure 5.2: Overall Heat Transfer Coefficient and Thermal Resistance for a Typical Fouling Run 54 Figure 5.3: Fouling Resistance over Time of 5 wt % DAO in HO/FO Mixture 56 Figure 5.4: Fouling Resistance over Time of 10 wt % DAO in HO/FO Mixture 57 Figure 5.5: Fouling Resistance over Time of 15 wt % DAO in HO/FO Mixture 59 Figure 5.6: Relationship of Initial Fouling Rate with Calculated Asphaltene Content for 0, 5, 10 and 15 wt % DAO in HO/FO Mixture 61 Figure 5.7: Relationship of Initial Fouling Rate with Calculated Resins Content at Constant Asphaltenes Content 61 viii Figure 5.8: Relationship of Initial Fouling Rate with Re/As ratio for 0, 5, 10 and 15 wt % DAO in HO/FO Mixture 62 Figure 5.9: Relationship of Initial Fouling Rate with (Re + Ar)/As Content in Mixture for 0, 5, 10 and 15 wt % DAO in HO/FO Mixture 64 Figure 5.10: Relationship of Properties of Mixtures with Re/As ratio for all oil Mixtures 66 Figure 5.11: Pentane Insolubles Variation with Re/As Ratio for Various DAO Concentrations 67 Figure 5.12: Measured Pentane Insloluble Concentration Variation with Calculated Asphaltene Contents for Various DAO Concentrations 68 Figure 5.13: Measured Hot Filtration Insoluble Concentration Variation with Calculated Resins Contents for Various Asphaltene Concentrations 68 Figure 5.14: Hot Filtration Insolubles Variation with Re/As Ratio for Various DAO Concentrations 69 Figure 5.15: Initial Fouling Rate Dependence on Solids Concentrations (T b 85° C, T s o 230° C, V 0.75 m/s) 70 Figure 5.16: Relationship of % DAO and % HO to Hot Filtration Insolubles 71 Figure 5.17: Deposit Formed on Probe Surface for Run with 15 % DAO -3.5 % HO - 81.5 % FO with Low Fouling Rate. 73 Figure 5.18: Deposit Formed on Probe Surface for Run with 10 % D A O -10 % HO - 80 % FO with High Fouling Rate. 73 Figure 5.19: Close-Up of Deposit Formed on Probe Surface for Run with 10 % DAO - 10 % HO - 80 % FO with High Fouling Rate. 74 Figure 5.20: Deposit Characteristics Variation with DAO Content in Mixture 75 Figure 5.21: SEM Micrograph of Fouling Deposit for Run with 5 % DAO -5 % HO and 90 % FO 76 Figure 5.22: E D X Plot of Fouling Deposit for Run with 5 % DAO - 5 % HO and 90 % FO 76 ix Figure 5.23: Figure 5.24: Figure 5.25: Figure 5.26: Figure 5.27: Figure 5.28: Figure A3.1 Figure A3.2 Figure A4.1 SEM Micrograph of Fouling Deposit for Run with 15 % DAO -5 % HO and 80 % FO 77 E D X Plot of Fouling Deposit for Run with 15 % DAO - 5 % HO and 80 % FO 77 SEM Micrograph of Fouling Deposit for Run with 5 % DAO -15%HOand80%FO 78 E D X Plot of Fouling Deposit for Run with 5 % DAO - 15 % HO and 80 % FO 78 Fouling Regime Map 80 Relationship of Oil Compatibility Model Index to Colloidal Instability Index 87 Fouling Resistance over Time for A Repeat Run with 5% DAO -15% HO - 80% FO Oil Mixture 110 Fouling Resistance over Time for A Repeat Run with 15% DAO -10% HO - 75% FO Oil Mixture 111 Fitting Kinematic Viscosity Variation with Temperature of 10 % DAO, 10 % HO and 80 % HO Oil Mixture into a First Order Exponential Decay Function 112 A C K N O W L E D G E M E N T I wish to express my gratitude to Professor A. P. Watkinson, my research supervisor, for his guidance, encouragement, support and utmost patience during the course of this work without which this work would have not been possible. I would like to thank Dr. J. J. Dusseault and Dr. A. Uppal of Imperial Oil Ltd. for providing the oil samples, performing the HPLC SARA analysis on the samples and for their technical support. The financial support of Imperial Oil Ltd. is gratefully acknowledged. The support and helpful suggestions of Dr. S. Asomaning, B. Sundaram, Dr. I. Rose and Dr. D. Posarac are greatly appreciated. I would like to thank all members of the chemical engineering faculty, staff, workshop, stores and graduate students for their assistance. I wold like to dedicate this thesis to my family for their love and support. xi 1.0 INTRODUCTION Fouling is the deposition of unwanted materials on equipment surfaces such as heat exchangers while processing. The presence of these deposits represents a resistance to the transfer of heat and therefore reduces the efficiency of the particular heat exchanger. Therefore, when deposits accumulate, removing the deposits becomes very necessary to maintain desired process conditions. Fouling remains a major cost penalty in oil refineries. Bott [1995] reported that fouling costs in a typical US refinery in 1993 is about $ 20 - 30 million per year for processing 100000 barrels/day. This figure is based on extrapolation of results reported by Van Nostrand in 1981. A number of factors contribute to this cost of fouling such as increased capital investment, additional operating costs and loss of production. In order to make allowance for potential fouling, in the design stage the area for a given heat transfer is always increased. Operating costs result from cleaning the heat exchanger, and involve both labour costs and the costs of cleaning chemicals. Cleaning processes require shutdown of the heat exchanger, which result in severe cost penalties due to loss of production. Canadian petroleum resources include conventional light crude oils, heavy oils and oil sands. The latter two contain high percentages of heavy oil fractions such as asphaltenes, Speight [1991]. It is of great interest to the Canadian oil industry to process these heavier fractions at a minimal expense. Severe fouling is encountered during the processing of asphaltene-containing oils which increases the interest in understanding asphaltene fouling. 1 Chapter 1: Introduction 2 Petroleum oils may be characterized by solvent fractionation into saturates, aromatics, resins and asphaltenes. "Saturates" contain mainly paraffin and some olefin compounds having low boiling points and molecular weights. Aromatics include benzene-like compounds with higher boiling point and molecular weight compared to that of saturates. Resins and asphaltenes on the other hand are the heavier fractions of oils with high boiling points and molecular weights. Asphaltene precipitation from oils depends on the concentration of solvent components such as resins and aromatics. This thesis investigates the effect of varying oil composition on fouling of asphaltene-containing oils. The compatibility of oil mixtures and its relation to heat exchanger fouling is also studied. 2.0 L I T E R A T U R E R E V I E W Heat exchanger fouling has been divided into five primary categories including precipitation, particulate, chemical reaction, corrosion and biological fouling, Epstein [1983]. Precipitation fouling, sometimes referred to as scaling, may be caused by crystallization of dissolved inorganic salts present at the heat exchanger surface under supersaturation conditions, Hasson [1981]. Solidification fouling is another subdivision of this category which involves the freezing of a pure component, or of a high melting temperature component such as hydrocarbon wax. Particulate fouling is the accumulation of finely divided solids suspended in the process fluid on the heat transfer surface. In some cases, the equipment is run vertically to avoid sedimentation fouling that is caused by gravity. Chemical reaction fouling is defined as a deposition process in which a chemical reaction either forms the deposit directly on a surface, or is involved in forming foulants which become deposited, Watkinson [1992]. Reaction does not take place with the surface material itself. Corrosion fouling refers to the formation or accumulation of corrosion products on the heat transfer surface. Biological fouling involves the attachment of macro or micro-organisms to a heat transfer surface. Heat exchanger fouling is a complex process and in many practical situations, more than one type of fouling may be present. Petroleum fouling is an example of such a practical case in which many steps are involved. Therefore, this chapter will review some aspects of heat exchanger fouling while processing heavy oil fractions that will lead to the objectives of this work. 3 Chapter 2: Literature Review 4 2.1 Heat Exchanger Fouling of Asphaltenes-Containing Oils Crude oil heat exchanger fouling is a major problem facing the oil industry. Fouling occurs as a result of a combination of chemical reactions and physical changes that occur when crude oil is exposed to high metal surface temperatures in an exchanger. Analysis of these deposits indicates that they are composed primarily of infusable coke, asphaltenes, and inorganic materials. Many variables play a role in crude oil fouling in heat exchangers such as crude oil composition, inorganic contaminants, process conditions, and metal surface temperatures. Several mechanisms for fouling have been postulated including inorganic compounds deposition and corrosion of metal surfaces, oxygen induced polymerization, free radicals opening double bonds and initiating polymerization, and asphaltene precipitation, Murphy and Campbell [1992]. Although there may be cases where one or more other fouling mechanisms may predominate, both laboratory studies and analyses of actual exchanger deposits point to asphaltene precipitation and subsequent carbonization as the most significant mechanism. Further studies have found that the presence of asphaltene does not necessarily mean a crude oil will foul. Asphaltenes that are incompatible with the crude oil chemistry and composition have a much greater tendency to precipitate and foul. The unique chemistry of a particular crude oil, and the types and quantities of asphaltenes present, determine the potential for a particular crude to foul, Dickakian and Seay [1988]. Therefore, it is very important to understand the chemistry and the compatibility of a certain crude to be able to predict its tendency to foul. Chapter 2: Literature Review 5 The following sections will be devoted to understanding the chemistry of asphaltenes and the role of other constituents in keeping asphaltenes in solution and therefore preventing fouling. 2.1.1 Petroleum Oils and Solvent Fractionation Petroleum consists of hydrocarbon compounds with a wide rage of boiling points and carbon numbers, and other heteroatomic organic compounds containing nitrogen, sulfur and oxygen as well as heavy metals such as vanadium and nickel. Oils are classified based on their viscosities, densities and API gravities into light crude oils (viscosity < 100 mPa.s, density < 934 kg/m3 and API > 20°), heavy crude oil (viscosity 100-10,000 mPa.s, density 934-1000 kg/m3 and API > 10° - 20°), and tar sand bitumen (viscosity > 10,000 mPa.s, density > 1000 kg/m3 and API < 10°), Speight [1991]. Petroleum oils can be separated by solvent fractionation into four constituents namely saturates, aromatics, resins and asphaltenes. Table 2.1 lists the properties of these fractions given by Wiehe [1999] for a western Canadian vacuum resid. Asphaltenes have the highest molecular weight being the heaviest fraction among the four constituents and are commonly defined as the portion of petroleum which is insoluble in low-boiling liquid hydrocarbon alkanes but soluble in benzene, Ferworn and co-workers [1993]. Table 2.1 Analysis of Fractions of Cold Lake Vacuum Resid, Wiehe [19991 Fraction Yield wt % C Wt% H wt% H/C Atomic S ' Wt% N wt% VPO M W Saturates 18 84.54 12.31 1.73 2.74 0.03 920 Aromatics 17 81.87 10.00 1.46 5.56 0.12 613 Resins 40 82.08 9.50 1.38 6.09 0.77 986 Asphaltenes 25 81.93 7.94 1.15 7.50 1.15 2980 Chapter 2: Literature Review 6 Solvent fractionation is used for classifying oil into the hydrocarbon types of polar compounds, aromatics and saturates, and recovery of representative fractions of these types. Asphaltenes and resins comprise the polar fraction of petroleum. The A S T M D 2007-93 is a standard solvent fractionation method used for samples of initial boiling point of at least 260°C. Asphaltene fraction is separated from the sample by precipitation using n-pentane at a ratio of 1:40. Precipitated asphaltenes are removed by filtration. The oil sample, diluted with n-pentane, is then charged to a glass percolation column containing Attapulgus clay in the upper section and silica gel in the lower section. The saturate fraction, on percolation in a n-pentane eluent, is not adsorbed on either the clay or silica gel and therefore collects at the bottom of the columns. The resin fraction is adsorbed on the clay and subsequently desorbed with a mixture of toluene and acetone. Aromatics, on percolation, pass through the adsorbent clay but adsorb on the silica gel and are later desorbed by recirculation of toluene. Solvents are evaporated off the collected samples and the oil fractions are recovered. This procedure requires large solvent quantities and therefore, is unfeasible for use to recover large samples of oil fractions. The apparatus used in this test method is shown in Figure 2.1. 2.1.2 Petroleum Asphaltenes Asphaltenes are dark brown amorphous solids which consist of highly polydisperse macromolecules containing a broad distribution of polar groups in their structure. The published molar mass data for petroleum asphaltenes range from 500 to 500,000 g/mol, Long [1981]. Current research suggests a much narrower range of 1000 to 10,000, Thawer et al. [1990]. The measured molar mass is dependent on the source of the crude oil and the type of solvent used to precipitate the asphaltenes. Vapor pressure Chapter 2: Literature Review 7 Figure 2.1: Clay-Gel Percolating Column, A S T M D 2007-93 osmometry (VPO) is one of the most common methods of determining the molar mass of asphaltenes. The color of dissolved asphaltenes in benzene is deep red at low concentrations, Kawanaka and co-workers [1991], On heating to temperatures above 400°C, asphaltene molecules decompose forming carbonacious coke and volatile products. The complexity of asphaltene fractions has made it very challenging to determine a definite structure; however, efforts have been made to describe asphaltenes in terms of Chapter 2: Literature Review 8 chemical structure or elemental analysis for the past six decades. An example of these attempts is the work carried out by Suzuki and co-workers [1982], in which they were able to obtain chemical structure of tar-sand bitumen. Table 2.2 represents the molecular weights and average molecular formulae of Cold Lake bitumen and its fractions. Table 2.2 Molecular Weights and Average Molecular Formulae of Cold Lake Bitumen and its Fractions, Suzuki et al. [19821 Bitumen and Fractions wt% M W (VPO) H / C Atomic Average Molecular Formula Cold Lake Bitumen - 500 1.55 C34.5H53.5N0.11 S0.72O0.47 Saturate fr. 30.3 331 1.83 C24.0H44.0 Aromatic fr. 40.6 517 1.46 C34.9H52.3N0.09S 1.04 Resin fr. 13.5 1010 1.38 C68.8H94.9N0.95S 1.7601.18 Asphaltene 15.6 2030 1.19 C 139H165N2.33 S4.74O0.6O More recent study was carried out by Strausz [1992] and co-workers where they were able to obtain a hypothetical asphaltene molecule as shown in Figure 2.2. This hypothetical molecule was obtained using Athabasca asphaltenes and incorporates in its structure condensed aromatic clusters with side chains and heteroatoms. It has an elemental formula of C420FI496N6S14O4V, and a H/C atomic ratio of 1.18. Chapter 2: Literature Review 9 Figure 2.2: A Hypothetical Asphaltene Molecule, Strausz et al. [19921 Asphaltenes solubility varies dramatically in different solvents. These differences can be explained considering the solvent power of the precipitating liquids, which can be related to molecular properties using the solubility parameter given by Hildebrand and co-workers [1970]. The solubility parameter of nonpolar solvents can be related to the heat of vaporization AH V and the molar volume V, / v \ 1 / 2 A H - R T V (2.1) R is the gas constant and T is the absolute temperature. Rogel [1998] classified asphaltenes according to their solubility parameters as being highly soluble if they are in the 17.0 - 22.5 M P a 0 5 range, fairly soluble within the range 22.5 - 25.5 M P a 0 5 , and difficult to dissolve beyond 25.5 MPa 0' 5. The solubility parameter of Cold Lake Chapter 2: Literature Review 10 asphaltenes was reported, by Rogel [1997], to be 24.3 MPa 0' 5 lying in the fairly soluble range. The yield of precipitate of each solvent depends on the difference between the solubility parameter of the asphaltenes and the solvent. The solubility parameters and the asphaltene precipitate yield at 21°C for a variety of solvents are presented in Table 2.3. It is apparent that asphaltenes are soluble in hydrocarbon solvents with solubility parameters greater than or equal to 8.4 (cal/cm3)0'5, while they precipitate in solvents with solubility parameters less than or equal to 8.2 (cal/cm3)05. In general, asphaltenes are more soluble in aromatics than straight and branched chain paraffins. Table 2.3 Yields of Asphaltenes Precipitated from Western Canadian Bitumen Using Various Solvents fSpeight, 19911 Hydrocarbon Solvent Solubility Parameter Precipitate (wt. % Asphaltenes) (cal/ml)0 5 (MPa) 0 5 Isopentane 6.8 13.9 17.6 N-Pentane 7.0 14.3 16.9 Isohexane 7.1 14.5 15.3 N-Hexane 7.3 14.9 13.5 Isoheptane 7.2 14.7 12.8 N-Heptane 7.5 15.3 11.4 Isodecane 7.6 15.6 9.8 N-Decane 7.7 15.8 9.0 Cyclopentane 8.2 16.8 1.0 Cyclohexane 8.2 16.8 0.7 Benzene 9.2 18.8 0 Toluene 8.9 18.2 0 Xylene 8.8 18.0 0 Isopropylbenzene 8.6 17.6 0 Isobutylbenzene 8.4 17.2 0 Chapter 2: Literature Review 11 2.1.3 Formation of Micelles, Colloids, and Flocculates by Petroleum Asphaltenes Asphaltene particles are believed to exist in oil partly dissolved and partly in colloidal and/or micellar form. Whether the asphaltene particles are dissolved in crude oil, in colloidal state or in micellar form, depends to a large extent, on the presence of other species (saturates, aromatics, resins) in the crude oil. The existence of various states of asphaltenes in crude oil has been extensively discussed in literature, Yen [1974]; Mansoori [1996]. Small asphaltene particles can be dissolved in a petroleum fluid, whereas relatively large asphaltene particles may flocculate out of the solution and then can form colloids in the presence of excess amounts of resins and hydrocarbon saturates. Flocculation of asphaltene in paraffinic crude oils is known to be irreversible. Asphaltene and its flocculates are known to be surface-active agents. The flocculated asphaltene will precipitate out of the solution unless there is enough resins in the solution so that they can cover the surface of asphaltene particles by adsorption and form colloids. Asphaltene flocculates will also precipitate upon any actions of a chemical, electrical, or mechanical nature (such as the addition of a n-alkane) that would upset the colloidal balance of the flocculates. Various investigators have established the existence of asphaltene micelles when an excess of aromatic hydrocarbons is present in a crude oil, Pfeiffer and Saal [1940]; Dickie and Yen [1967]; Galtsev and co-workers [1995]; Mansoori [1996]. Several investigators have performed experimental measurements of critical micelle concentration for solutions of asphaltene in aromatic solvents, Sheu [1996]; Andersen and Birdi [1991]; Ravey and co-workers [1988]. Furthermore, the phenomenon of self-Chapter 2: Literature Review 12 association in asphaltene/toluene systems has been confirmed through measurements of surface tension, Sheu [1996]. Sheu has shown that at low concentrations, below the critical micelle concentration (CMC), the asphaltenes in solution are in a molecular state, whereas, above the CMC, asphaltene micelle formation occurs in a manner similar to that in surfactant systems. However, Andersen and Speight [1993], consider that there is an alternative method of data interpretation that will add another dimension to the determination of the critical micelle concentration. This alternative analysis is based on the Gibbs excess adsorption equation: - C ^ y _ a RT dC a where T a is the Gibbs surface excess, C a the concentration of compound a, and y the surface tension of the solution. According to this equation, the determination of a critical micelle concentration would be better served by an examination of surface tension y versus In C a . Espinat and Ravey [1993] with the use of scattering techniques have shown that the best model to describe the morphology of asphaltene micelle in solution is a disk. However, several other experimental investigations have shown that asphaltenes could be of spherical-like, cylindrical-like, or disk-like form, Ravey and co-workers [1988]. All these investigations are indicative of the fact that asphaltene particles may self-associate, but not flocculate, and form micelles in the presence of excess amounts of aromatic hydrocarbons as shown in Figure 2.3, Lian et al. [1994]. Chapter 2: Literature Review 13 1 MOHOMERICSHEET }-2asa 2 - $ nm REVERSED MICELLE i; HARTLEY MECFJXE SUP5RR MICEULE GIANT SUPER MTCELLE 200-2000 ran 10-20 wa MULTT-LAMi.Lr.-AR VESICLE 1000 - iO,OGO am UK ft? FLOC 20,000 run :w -GEL l O O M t t M Figure 2.3 Asphaltene Micel le Formation, L i a n et al. [19941. While the mechanisms of asphaltene flocculation and colloid formation are relatively well understood and modeled [Mansoori, 1996], the phase behavior of asphaltene micelle formation is not well characterized. In many cases, in recent years, micelle formation is confused with the asphaltene colloids. Despite the experimental Chapter 2: Literature Review 14 evidence on the micellization of asphaltenes, little or no theoretical and modeling research has been performed to explain and quantify this phenomenon. The process of colloid formation of asphaltenes with resins follows an irreversible process of flocculation, Leontaritis and Mansoori [1988]; Park and Mansoori [1988], Kawanaka co-workers [1988]. When asphaltenes form micelles, a reversible self-association process is recognized in which resins have no role. 2.1.4 Chemistry of Resins Resins are defined as the polar fraction of petroleum that is soluble in n-alkanes and aromatic solvents, and insoluble in ethyl acetate. Table 2.1 shows that the resin fraction has properties intermediate between those of the asphaltene and the aromatic and saturates fractions. There is a notable decrease in the H/C atomic ratio of the asphaltenes relative to that of resins which indicates that aromatization is more advanced in the asphaltenes than in the resins. It also indicates that if the asphaltenes are maturation products of the resins, one of the maturation processes involves aromatization of the non-aromatic portion of resins, Koots and Speight [1975]. Resins typically contain the heteroatoms nitrogen, sulfur and oxygen in moderate amounts, Clark and Pruden [1997], and can be found to be neutral, basic or acidic as shown by Strausz and Rubinstein [1980] for Cold Lake heavy oil (Table 2.4). Initial postulates of resin structure invoked the concept of long paraffinic chains with naphthenic rings interspersed throughout as shown in Figure 2.4. Other structures used the idea of condensed aromatic and naphthenic ring systems and allowed the interspersion of heteroatoms throughout the molecule, Speight [ 1991 ]. The average Chapter 2: Literature Review 15 molecular formula of resins as found by Suzuki and co-workers [1982] in Cold Lake Bitumen is given in Table 2.2. Table 2.4 Resin Fractions for Cold Lake H ^ v y Oil, Strausz anH Rnbinste.n HMOl Component Cold Lake Heavy Oil Hydrocarbons: Saturates Aromatic 40 21 19 Asphaltenes 16 Resins: Acidic Basic Neutral 44 15 7 22 CH, {CH )-CM-l C H , "C?i, cm, Cl-i .£ *t CM •3 CH... CH. C M , CH 5 CH: 3 C i , 3 C H 3 Figure 2.4 Structures nf Resins. Speight 11991] Chapter 2: Literature Review 16 2.1.5 Role of Resins in Asphaltenes Stabilization The activity of resins to prevent asphaltene flocculation is still not well known. Asphaltene molecules (micelles) are believed to be surrounded by resins that act as peptizing agents as shown by Wiehe [1997] in Figure 2.5. The resins maintain the asphaltenes in a colloidal dispersion (as opposed to a solution) within the crude oil. The peptization model was first proposed by Pfeiffer and Saal [1940]. This concept is often repeated, Dickie and Yen [1967], Koots and Speight [1975], Speight [1982], Leontaritis and co-workers [1988], Wiehe and Liang [1996] and Barre [1997]. Experimental evidence, Lichaa [1977], suggests that for an oil mixture there is a critical concentration of resins below which the asphaltene flocculates may precipitate and above which they cannot precipitate. s a a a s s a R R R a s s a IR'AA R a s s a B A A ' R . a s s a R R R a s s a ,a.as s s A s Asphaltenes R s: Restns a - Aromatics s = Saturates Figure 2.5 Physical Model of Petroleum, Wiehe [1996]. The resins are attracted to the asphaltene micelles through their end group. This attraction is a result of both hydrogen bonding through the heteroatoms and dipole-dipole interactions arising from the high polarities of the resin and asphaltene molecules. The Chapter 2: Literature Review 17 paraffinic component of the resin molecule acts as a tail making the transition to the relatively non-polar bulk of the oil where individual molecules also exist in true solution, Hunt [1996]. Koots and Speight [1975], studied the chemical and structural analyses of a series of petroleum resins and concluded that a crude oil is a complex system whereby each constituent fraction depends upon others for complete mobility; and it is not necessarily achieved by interchanging fractions from one crude oil to another. It is presumed that the resins associate with the asphaltenes in the manner of an electron donor-acceptor system and that there could well be several points of structural similarity between the asphaltenes and resins. In that case, this may well have an adverse effect on the ability of resins to associate with asphaltenes from any other crude oil. Lian and co-workers [1994] reached another interesting conclusion stating that the resins are better solubilizers of asphaltenes from the same crude oil. Extensive research is directed to evaluate the asphaltene-solvating power of various nonconventional solvents for asphaltenes and the effect of adding resin-containing compounds on the stability of asphaltenes in the oil phase. Jamaluddin et al. [ 1996] systematically evaluated the asphaltene-solvating power of deasphalted oil (DAO) using a light-scattering technique. DAO was prepared using Lindbergh oil and asphaltenes were precipitated using n-pentane and a rotary evaporator was used to remove the solvents. Asphaltenes were removed and dried in an oven then added at a concentration of 4.8 % to 25 g of DAO. N-pentane was slowly added to the mixture and the onset of asphaltene precipitation was determined using a light scattering technique. Experimental results suggested that DAO is a strong asphaltene solvent presumably Chapter 2: Literature Review 18 because of its native resin and aromatics contents. In general it was concluded that a saturates content of less than 30 - 35 wt. % as determined by SARA analysis represents a good solvent. Work of Clarke and Pruden [1997], was carried out to test the ability of a number of aromatic compounds, chosen for their similarity to resins, to delay the onset of asphaltene precipitation. Indole and quinoline were used in their study to represent the non-basic and basic nitrogen that was found in the resin fraction. Quinoline also contains an aromatic core. These two compounds were added separately to a sample of Cold Lake bitumen and the onset of asphaltene precipitation was detected using heat transfer analysis. Results indicated that quinoline was effective in delaying the onset of asphaltene precipitation while indole enhanced the precipitation of asphaltenes. Hammami and co-workers [1998] conducted a study to evaluate the effect of different resin fractions on the onset of asphaltene precipitation. They separated the resin fraction of a North Sea stock tank oil sample and classified it into resin (1) and resin (2) fractions. They further classified the resin fractions in terms of their neutral, basic, pyrrolic and acidic components as shown in Table 2.5. The onset of asphaltene precipitation of a North Sea stock tank oil sample was determined upon addition of resins (1) and (2) fractions using light transmission technique. Experimental results showed that the addition of resins (2) fraction increased the stability of the asphaltenes by increasing the minimum n-pentane concentration required to induce asphaltenes precipitation. However, the addition of resins (1) fraction, was found to have no obvious effect on asphaltenes stability over the range of concentrations tested. These results imply that particular components or group of components present in the resins (2) Chapter 2: Literature Review 19 fraction are responsible for stabilizing asphaltenes in the mixture. As can seen in Table 2.5, resins (2) fraction is more polar in nature than resins (1). In addition, Resin (2) fraction is more basic than resin (1) fraction which supports the results obtained in the study by Clarke and Pruden [1997] where basic quinoline was effective in delaying asphaltene precipitation. Table 2.5 Summary of Resin Fractions Analyses of Study Done by Hammami and co-workers [19981. Fraction Aromatics Polars Basic Others Resins (1) 63.3 36.7 1.2 98.8 Resins (2) 48.9 51.1 10.3 89.7 Storm and co-workers [1998], performed a study where they observed a rise in insoluble particles upon heating a heptane soluble oil sample (DAO) to 285°C. This indicated that not all micelle-forming molecules are in the asphaltenes. They also found that the presence of asphaltic (asphaltenes plus resins) micelles in the crude appears to accelerate the conversion of pentane insolubles (asphaltenes and resins) to heptane insolubles (asphaltenes) during the induction period to form coke as given by Wiehe's model for coke formation, [1993]. 2.2 Deposit Formation by Petroleum Asphaltenes Asphaltene deposit formation is a complex process and it has been a great challenge to try to postulate the mechanism of the process and develop a model to predict it. Asphaltene precipitation is described in detail by Asomaning, [1997]. It has been found that asphaltene precipitation during crude oil production is a physical process; however, it is not yet clear if deposit formation on heat transfer surfaces is a physical Chapter 2: Literature Review 20 process, a chemical reaction, or a combination of both. There are many factors that play a role in controlling deposit formation on heat transfer surface such as velocity, bulk temperature and surface temperature, which makes model development a very challenging task. The following sections will describe commonly developed mechanisms and models for asphaltene deposit formation on heat exchanger surfaces. 2.2.1 Mechanisms of Deposit Formation Deposit formation mechanisms proposed by Dickakian and Seay [1988], and Lambourn and Durrieu [1983] are based on the incompatibility between asphaltenes and other components of crude oil and therefore suggested to be a physical process. Dickakian and Seay summarized the mechanism in the following steps: 1. Precipitation of asphaltenes initiated due to incompatibility between asphaltenes and the oil. 2. Adherence of precipitated asphaltenes to the hot heat transfer surface. 3. Coking of asphaltenes on the heat transfer surface. This mechanism suggests that the critical step in this process is the first step of asphaltene precipitation. The precipitation step is linked to the solubility of asphaltenes in the oil. However, the mechanism does not state the exact method by which the asphaltene precipitate or asphaltene micelle gets destroyed. The mechanism suggests the possible role of chemical reaction in adhesion and coke formation stages; however, this role is minor compared to the physical process of asphaltene precipitation. Lambourn and Durrieu studied fouling in crude oil heat exchangers and found that asphaltene precipitation is the major cause of fouling. They performed microscopic Chapter 2: Literature Review 21 examination of colloidal asphaltenes in the crude. These studies revealed that the asphaltenes, upon precipitation, coated droplets of water left in the crude oil after desalting and formed an emulsion. The emulsion incorporated particulates, mainly oxides and sulphides of iron, to form stable entities that were insoluble in the crude oils. These emulsions deposited on the heat transfer surface where it aged to form coke. This proposed mechanism requires a threshold concentration of asphaltenes of 1.3 wt %. The study performed by Lambourn and Durrieu revealed another interesting behavior for asphaltenes at elevated temperatures. The study showed that suspended asphaltenes have a complex relationship with the bulk temperature, that asphaltenes dissolve in the temperature range 100-140 °C, but these dissolved asphaltenes re-100 203 330 Figure 2.6 Dependence of Asphaltene Solubility on Temperature TLambourn and Durrieu, 19831. precipitate when the temperature is raised above 200 °C as shown in Figure 2.6. Chapter 2: Literature Review 22 The mechanism proposed by Lambourn and Durrieu included an asphaltene precipitation step followed by an interaction of the asphaltenes with water and particulates to form an emulsion which deposits on the heated surface to age and form coke-like materials. These mechanisms suggest that asphaltene fouling of heat exchangers is mainly a physical process, however Eaton and Lux [1984] proposed that fouling by asphaltenes is principally via chemical reaction. They have suggested that saturated hydrocarbons convert to unsaturated hydrocarbons, organic acids, resins and asphaltenes that are finally converted to coke as follows: Inorganic Acids 0 2 Saturated hydrocarbon > Unsaturated hydrocarbon > Organic Metals Wall acids > Resins and Asphaltenes > Coke-like deposits. A A It is widely accepted that the quantity of asphaltenes in virgin crude oil is relatively low and it increases upon processing and exposure to high temperatures as resins and aromatics are converted to asphaltenes, Blazek and Sebor [1993]. This fact supports the mechanism proposed by Eaton and Lux. There are two processes that may be responsible for this increase in asphaltene concentrations. The first is the hydrocracking of maltenes (saturates, aromatics and resins) to asphaltenes as supported by Blazek and Sebor [1993] and Savage and Klein [1987]. The second process is the dealkylation of resins and aromatics followed by condensation reactions that result in the formation of asphaltenes of varying molecular weights, and the production of lower molecular weight compounds. Chapter 2: Literature Review 23 Mechanisms proposed in the literature for asphaltene fouling are not consistent, which suggests that additional efforts are necessary to understand and control the fouling process. 2.2.2 Mode l ing of Deposit Format ion Organic fluid fouling of heat exchangers has been the focus of many researchers over the years and there have been many attempts to model this behavior to enable its accurate prediction. However these models are based on many assumptions, some of which are not valid in reality, that limit their predictive accuracy. The modeling of asphaltene fouling has been a greater challenge due to numerous reasons such as the high complexity of asphaltenes, inconsistency in the properties of asphaltenes from different oil samples, variation in asphaltene precipitation detection methods encountered by different researchers, and the lack of accurate thermodynamic properties for asphaltenes. These difficulties have made it hard to develop an accurate model for asphaltene fouling under conditions encountered in heat exchangers. However, there are several models proposed in the literature for chemical reaction and particulate fouling of heat exchangers. There have also been attempts to model asphaltene precipitation, mostly under ambient temperatures and high pressures, but there has been no attempt to link such models, or precipitation data to thermal fouling. This section will refer to some heat exchanger fouling and asphaltene precipitation models. A simple model was developed by Kern and Seaton [1959] to predict asymptotic fouling. This model consists of a deposition term which is a function of the concentration of foulants and the mass flow rate, and a removal term which is a function of the deposit Chapter 2: Literature Review 24 thickness and the wall shear stress. The model equation beyond the induction period is given by: m = m (l-exp(-bt)) (2.3) where m is the mass of deposit per unit surface area (kg/m2), m* is the asymptotic mass of deposit per unit area and "b" is the time constant taken to be the time when "m" reaches 63 % of its asymptotic value. The thermal fouling resistance is related to the mass of deposit per unit surface area and the deposit thickness as follows: R f = ~ = — (2.4) k f p f k f where x is the deposit thickness, pf is the density of the foulant and kf is the thermal conductivity of the foulant. Assuming that the density and the thermal conductivity are constant with time, the fouling resistance is given by R f =R/(l-exp(-bt)) (2.5) where Rf* is the asymptotic fouling resistance. The initial fouling rate is given by R f = b x R f * (2.6) Taborek [1982] and Pinhero [1981] have modified the basic Kern and Seaton model to incorporate other terms such as the shear strength of the deposit. Watkinson and Epstein [1969] introduced an Arrhenius type sticking probability while Paterson and Fryer [1988] introduced reaction kinetics to heat exchanger fouling modeling. Crittenden et al. [1987] presented a general model that took into account the transport of fouling Chapter 2: Literature Review 25 precursors as well as chemical reaction. Epstein [1994] developed a modified version of the Crittenden model by treating the attachment to the surface as a process in series with mass transfer. This modified model is based on the assumption that the reaction and attachment constants are proportional to the residence time of the fluid at the surface. Therefore, the longer the fluid stayed at the surface the bigger the chance that reaction would occur. Zhang and Watkinson [1991] and Panchal and Watkinson [1993] modeled autoxidation reaction fouling by assuming a two step process resulting in foulant generation. The first step involves the reaction of a soluble reactant to form a sparingly soluble foulant precursor which then reacts to form an insoluble foulant. All the models described above were able to describe a certain type of fouling mechanism with varying degrees of success. There have been great advancements in modeling heat exchanger fouling over the years; however there is still a need to improve the prediction accuracy of these models and enable their application more to the general case. Asphaltene precipitation models proposed in the literature are based on the thermodynamic phase behaviour of asphaltenes and the use of equations of state to predict the volume fractions of asphaltenes in different phases. Models available are mostly developed for processes occurring during oil production, recovery and transport with conditions different from those in heat exchangers. Hirshberg et al. [1984] approached asphaltene precipitation by using a bulk phase equation of state to describe asphaltene solubility, neglecting its colloid nature assuming it is a homogeneous solid. A thermodynamic colloid model of asphaltene precipitation was proposed by Leontaritis Chapter 2: Literature Review 26 and Mansoori [1987]. This model does not explicitly deal with the dependence of the micellization process on the characteristics of the micelles. These models are based on many assumptions that are not valid in reality, such as the elimination of asphaltenes micellization and the reversibility of the process of asphaltene precipitation. In addition, the high complexity of some of these models limits their use due to the lack of reliable thermodynamic properties for asphaltenes. However, these models provide the first steps in future attempts in predicting heat exchanger fouling of asphaltene-containing oils. 2.3 Modeling Oil Compatibility and Its Relation to Deposit Formation The problems associated with crude oil fouling have made it desirable to develop techniques that will enable the prediction of the fouling behavior of a certain oil mixture prior to processing. It has been shown in previous sections that crude oil fouling is caused mainly by asphaltenes that are incompatible with the crude oil chemistry. Therefore, attempts have been made to predict the instability of oil blends. This section will discuss some oil compatibility models used for stability prediction. 2.3.1 Colloidal Instability Index The stability of crude oils has been expressed using the colloidal instability index (CH) by Gastel, [1971]. The CII bases instability or incompatibility of the petroleum on its composition of oil fractions, saturates, aromatics, resins and asphaltenes. The index is an expression of the colloidal nature of petroleum fractions and is a ratio of the asphaltenes and saturates which precipitate asphaltenes, to the sum of the aromatics and resins that peptize asphaltenes as follows: Chapter 2: Literature Review 27 CII = (Saturates + Asphaltenes) (Aromatics + Resins) (2.3) At fixed temperature and velocity, the fouling rate of heavy oil-fuel oil blends including mixtures with added pentane and xylene was previously correlated by Asomaning and Watkinson [1997] via the colloidal instability index. In their experiments, concentrations of asphaltene varied between 1.4 and 3.7 wt. %, resins between 2.2 and 4.4 %, whereas saturates were 30 - 67 % and aromatics 29 - 66 %. The ratio of Resins/Asphaltenes covered the narrow range of 1.2-2.6. It is evident that the colloidal instability index alone can not predict fouling over a full range of composition, since for a given sum of the peptizing agents (Resins + Aromatics), one should expect greatly different behavior if either the one or the other of the two terms in the numerator, the saturates or the asphaltenes, went to zero. Furthermore, low concentrations (ca. 100 ppm) of amphiphiles can greatly affect asphaltene solubility, and would not be reflected in this stability ratio. Hence, additional factors beyond the CII will be necessary to characterize the fouling potential of asphaltene-containing systems. 2.3.2 Oil Compatibility Model A solubility parameter based model has been developed by Wiehe [1999] to determine the correct order and proportions of blending petroleum oils to prevent rapid fouling and coking from the precipitation of asphaltenes. The basic hypothesis of the Oil Compatibility Model (OCM) is that the asphaltene/resin dispersion has the same flocculation solubility parameter, whether the oil is blended with nonpolar liquids or other oils. Chapter 2: Literature Review 28 Model parameters include the insolubility number, IN, which measures degree of insolubility of the asphaltenes present in the oil, and the solubility blending number, S B N , which measures the solvency of the oil for asphaltenes. They are defined as follows, I N = I O O I L Z H ( 2 4 ) IP T -0 H ) S b n = 1 0 O | ° " " 5 h ) (2.5) where 5f is the flocculation solubility parameter, 8H is the solubility parameter of n-heptane, 8T is the solubility parameter of toluene and 50ii is the solubility parameter of the oil. Therefore, if the oil is completely soluble in n-heptane and thus, contains no asphaltenes, the insolubility number is 0 but if the asphaltene-resin dispersion is barely soluble in toluene, the insolubility number is 100. Likewise, an oil that is as poor a solvent as n-heptane has a solubility blending number of 0 and an oil that is as good a solvent as toluene has solubility blending number of 100. The model parameters are calculated as follows, T E I N - r -. ( 2 6 ) 25p C — T BN A N 1 + — 5 (2.7) where T E is the minimum percentage of toluene required in a toluene-heptane mixture to keep asphaltenes in solution at a concentration of 2 grams of oil and 10 ml of test liquid (Toluene Equivalence Test), V H is the maximum volume in ml of n-heptane that can be added to 5 ml of oil without precipitating asphaltenes (Heptane Dilution Test) and p is the density of the oil in g/L. Chapter 2: Literature Review 29 The solubility blending number of a mixture of oils from the mixing rule for solubility parameters is the volumetric average calculated as follows, _ V , S N > . „ + V T S D M , + V 5 S „ M 7 + ... 3 BNmix C M ° B N 1 ^ V 2 ° B N 2 ^ Y 3 | J B N 3 ^ ••• fn Q\ V l + V 2 + V 3 + • • The compatibility criterion for a mixture of oils can be defined as, ^BNmix > ^Nmax (2-9) The nonsolvent oil dilution test is used for a sample oil that is a nonsolvent for asphaltenes but contains no asphaltenes itself. For such oil the insolubility index is set to 0 and the solubility blending number is calculated as follows, S-ro [ V N S D — V H ] 'NSO _ r „ i (210) c -v V N S D 1 + ^ 5 where STO is the solubility blending number for the test oil, VNSD is the maximum volume in ml of nonsolvent oil that can be blended with 5 ml of test oil without precipitating asphaltenes from the nonsolvent dilution test, and V R is the maximum volume in ml of n-heptane that can be blended with 5 ml of test oil without precipitating asphaltenes from the heptane dilution test. The solvent oil equivalence test is used for a sample oil that is a solvent for asphaltenes but contains no asphaltenes itself. The insolubility number is also set to 0, and solubility blending number is calculated as follows, S s o =100 T E SOE (2.11) Chapter 2: Literature Review 30 where T E is the toluene equivalence of the test oil and SOE is the minimum percentage of sample oil required in a sample oil-heptane mixture to keep asphaltenes in solution at a concentration of 2 grams of test oil and 10 ml of test liquid. An example is shown in Figure 2.7 given by Wiehe [1999] to explain this model. Figure 2.7 shows the insolubility number for a mixture of Souedie and Forties crudes It shows that for a compatible mixture, the volume percentage of Forties crude must be less than 67 % to obtain an insolubility number for the mixture lower than that of Souedie crude. Blends of Souedie arid Forties 1 0 O T A $ 0 -3 8 0 -mi 7 0 -S 6 0 * S S O & ® 4 0 3 0 -I 2 0 -3 1 0 -0 » 0 -F o r t i e s ; S B K = 2 7 , X N = 11 S o u e d i e : 8 M * 6 3 , 1 N « 3 9 I N « 3 9 f o r S o u e d i e • -J F 0 10 20 30 40 W ^ Volume % Forties Figure 2.7: Oil Compatibility Numbers for Souedie and Forties Crudes Wj^Hp M9991. " Chapter 2: Literature Review 31 2.4 Aims and Objectives of Work The role of resins in asphaltene stability in crude oil has been documented in the literature. It is of interest to examine the role of resins on the fouling of asphaltene-containing oils. The objectives of this work include the following: 1. Investigate the effect of varying the resins to asphaltenes ratio on the asphaltene fouling rate, hot filtration insolubles present in the sample at bulk temperature and pentane insolubles. 2. Study the effect on asphaltene fouling of adding de-asphalted oil to heavy oil blends. 3. Investigate the asphaltene-crude oil incompatibility and its relation to asphaltene fouling. This will aim at investigating oil compatibility models and their ability to predict fouling behavior. 4. Characterize deposits formed at heat exchanger surface and investigate their relations to the insoluble species in the fluid. 3.0 E X P E R I M E N T A L M A T E R I A L S AND A P P A R A T U S This chapter describes the fluids used in this study and the properties of test solutions. It also describes the experimental apparatus used to carry out the thermal fouling experiments. 3.1 Experimental Materials Cold Lake heavy oil (HO) supplied by Imperial Oil Resources Ltd. was used as the source for asphaltenes. Fuel oil (FO) taken from a crude unit vacuum-top-side cut supplied by Chevron Canada Ltd., was used as the carrier solvent. A sample of de-asphalted vacuum bottoms (DAO), made available by Imperial Oil Ltd., was added to the mixture to vary the resins/asphaltenes ratio. 3.1.1 Properties of Heavy Oil Cold Lake heavy oil is a black viscous liquid with high density and low API gravity compared to crude oil. It contains a significant fraction of high molecular weight hydrocarbons with carbon number greater than C25. It has an asphaltenes content of about 16.6 %, which is much higher than that of conventional crude oils. Heavy oil is known to have a sulphur content higher than 2 %, which gives it a rotten-egg smell, and it also includes heteroatoms and heavy metals. Properties of heavy oil used in this work are shown in Table 3.1. Saturates, aromatics, resins and asphaltenes contents of test oils used in this study were based on an average of two analyses done at two different research centers to ensure accuracy. One analysis was performed at the Imperial Oil Research Laboratory in Sarnia, Ontario, using a high precision liquid chromatography method designed for oils with a 32 Chapter 3: Experimental Materials and Apparatus 33 boiling point greater than 300°C. This method separates the oil sample into saturates, aromatics and polars. The total asphaltenes content was found by n-pentane precipitation and the resins content was found by the difference between the polar and asphaltene contents. Another analysis was performed at the National Center for Upgrading Technology according to A S T M D 2007M dividing the oil sample into the four oil solvent fraction constituents. Results of both analyses were similar; therefore, the average was used in this study. Results obtained using the A S T M D 2007M showed higher polar content and lower aromatics content than that of the H.P.L.C. SARA analysis shows that heavy oil has high aromatics content of 50 % followed by saturates at 23 %, asphaltenes at 17 % and resins at 10 %. Elemental analysis of heavy oil was carried out by Canadian Mircroanalytical Service Ltd. of Delta, B.C. Results of elemental analysis indicate high sulphur content of 4.51 %, and low H/C atomic ratio of 1.57. SARA analysis results obtained in this study are consistent with values reported in literature by Speight (1991) as shown in Table 3.2. These literature values are generalized ranges for the bulk fractions in crude petroleum, heavy oil, and residua. 3.1.2 Properties of De-asphalted Oil Amounts of resins needed in this study made it impractical to recover a pure resin stream from the heavy oil. De-asphalted oil was therefore used as the source of natural resins to vary the Re/As ratio of mixtures. It is a sample of de-asphalted vacuum bottoms that was made available by Imperial Oil. Oil de-asphalting is a common process in refineries where propane is used to precipitate asphaltenes out of the oil. Propane is then separated from the oil in evaporators heated by steam (Speight 1991). Chapter 3: Experimental Materials and Apparatus 34 Table 3.1: Properties of Cold Lake Heavy Oil Test Description Value SARA Analysis (HPLC) Saturates (wt. %) 21.94 1 Ring Aromatics (wt. %) 12.4 2 Ring Aromatics (wt. %) 14.4 3 Ring Aromatics (wt. %) 8.9 4 Ring Aromatics (wt. %) 18.4 Aromatics (total wt. %) 54.1 Resins (wt. %) 8.4 Asphaltenes (wt. %) 15.6 SARA Analysis (ASTM D 2007M ) Saturates (wt. %) 24.37 Aromatics (wt. %) 45.58 Resins (wt. %) 12.39 Asphaltenes (wt. %) 17.66 Average SARA Analysis Saturates (wt. %) 23.1 Aromatics (wt. %) 49.8 Resins (wt. %) 10.4 Asphaltenes (wt. %) 16.6 Elemental Analysis Carbon 80.27 Hydrogen 10.52 Nitrogen 0.41 Sulphur 4.51 H/C atomic ratio 1.57 Specific Gravity (@15°C) 1.038 API Gravity 10.1 Kinematic Viscosity (@ 80°C, m2/s) 4.25E-03 350-525°C 23.75 % 525°C + 76.25 % ~1 Chapter 3: Experimental Materials and Apparatus 35 Table 3.2: Generalized Ranges for the Bulk Fractions in Crude Petroleum, Heavy Oil, and Residua. Speight H9911 Range of Composition (wt/wt%) Asphaltenes Resins Oils Carbon Residue (wt/wt%) Petroleum <0.1-12 3-22 67-97 0.2-10.0 Heavy oil 11-45 14-39 24-64 10.0-22.0 Residua 11-29 29-39 ?-49 10.0-32.0 De-asphalted oil is a brown viscous semi-solid liquid at room temperature. It has a smell similar to that of heavy oil that is caused by its sulphur content. Table 3.3 shows some of the properties of DAO. SARA analyses results obtained by the two methods gave almost identical results for DAO. It was found that DAO has a high aromatics content of 68.45 %, followed by a saturates content of 20.72 % and a resins content of 10.08 %. It has an asphaltene content of 0.76 %. The DAO and HO contain roughly the same amount of saturates (22 %), and resins (10 %), but strikingly different asphaltenes levels of 17 % for HO and < 1 % for the DAO. Therefore DAO is used to vary the Re/As ratio of the mixture. However, it should be noted that adding DAO to the mixture increases total aromatics and saturates content of the mixture since they make up about 90 % of DAO. DAO has a sulphur content of 3.5 % compared to 4.5 % for heavy oil. DAO and HO have similar H/C ratio of 1.57 and 1.58, respectively, indicating that both oils may contain heavy fractions of similar compositions. Chapter 3: Experimental Materials and Apparatus Table 3.3: Properties of De-asphalted Oil Test Description Value SARA Analysis (HPLC) Saturates (wt. %) 20.48 1 Ring Aromatics (wt. %) 12.9 2 Ring Aromatics (wt. %) 17.5 3 Ring Aromatics (wt. %) 15.8 4 Ring Aromatics (wt. %) 22.4 Aromatics (total wt. %) 68.6 Resins (wt. %) 10.1 Asphaltenes (wt. %) 0.8 SARA Analysis (ASTM D 2007M) Saturates (wt. %) 20.93 Aromatics (wt. %) 68.3 Resins (wt. %) 10.05 Asphaltenes (wt. %) 0.72 Average SARA Analysis Saturates (wt. %) 20.7 Aromatics (wt. %) 68.5 Resins (wt. %) 10.0 Asphaltenes (wt. %) 0.8 Elemental Analysis Carbon 86.71 Hydrogen 11.15 Nitrogen 0.28 Sulphur 3.54 H/C atomic ratio 1.58 Chapter 3: Experimental Materials and Apparatus 37 3.1.3 Properties of Fuel Oil Fuel oil is used as a diluent for heavy oil and de-asphalted oil to reduce the viscosity of the mixture to make pumping of the test fluid easier. It is also used to vary the composition of test fluid and therefore test for a wider range of Re/As ratio. Fuel oil is yellowish brown in colour. It is semi-solid similar to paraffin wax at room temperature, however, upon heating it turns into a viscous fluid similar to lube oil. Table 3.4 shows the properties of fuel oil. Fuel oil is very high in saturates content followed by aromatics. Resin content is low, at 2.7 %, and there are only trace amounts of asphaltenes present. Its sulphur content is much less than that of heavy oil. However, the H/C atomic ratio is higher in fuel oil since it is a lighter oil than heavy oil. SARA analysis results using A S T M D2007M method showed higher aromatics content and lower polars content compared to that of H P . L . C . 3.1.4 Properties of Test Solutions The compositions of the mixtures for the thermal fouling runs are shown in Table 3.5. The table also includes some of the properties of these test solutions such as kinematic viscosity and density. Results show that increasing heavy oil and DAO concentrations increases the viscosity and density of the test solution. Viscosity measurements are obtained using the Haake V T 500 Rotovisco which gives information about the dynamic behavior of the fluid. Test fluids are found to exhibit Newtonian behaviour when the shear stress is plotted against the shear rate. Results for 15% DAO - 10% HO - 75% FO are shown in Figure 3.1. Chapter 3: Experimental Materials and Apparatus 38 Table 3.4: Properties of Fuel Oil Test Description Value SARA Analysis (HPLC) Saturates (wt. %) 70.88 1 Ring Aromatics (wt. %) 12.4 2 Ring Aromatics (wt. %) 7.2 3 Ring Aromatics (wt. %) 3.7 4 Ring Aromatics (wt. %) 2.2 Aromatics (total wt. %) 25.5 Resins (wt. %) 3.7 Asphaltenes (wt. %) Trace SARA Analysis (ASTM D 2007M) Saturates (wt. %) 68.34 Aromatics (wt. %) 29.92 Resins (wt. %) 1.74 Asphaltenes (wt. %) Trace Average SARA Analysis Saturates (wt. %) 69.6 Aromatics (wt. %) 27.7 Resins (wt. %) 2.7 Asphaltenes (wt. %) Trace Elemental Analysis Carbon 86.41 Hydrogen 12.76 Nitrogen 0.21 Sulphur 0.56 H / C atomic ratio 1.77 Density (@ 25°C, kg/m 3) 851 Kinematic Viscosity (@ 85°C, m 2/s) 2.15E-06 Chapter 3: Experimental Materials and Apparatus 39 Table 3.5: Composition and Properties of Test Solutions Weight Percent Kinematic Viscosity at 85°C x 10 6 (m2/s) Density at 85°C (kg/m 3) Kinematic Viscosity at 158°C x 10 6 (m2/s) 0% D A O - 5% H O - 95% F O 5.38 848 3.93 5% D A O - 0% H O - 95% F O 6.15 847 4.39 5% D A O - 5% H O - 90% F O 5.92 850 4.25 5% D A O - 10% H O - 85% F O 7.81 858 5.33 5% D A O - 15% H O - 80% F O 7.50 863 5.16 5% D A O - 20% H O - 75% F O 8.60 867 5.76 10% D A O - 0% H O - 90% F O 6.40 848 4.54 10% D A O - 3% H O - 87% F O 6.94 852 4.85 10% D A O - 5% H O - 85% F O 7.11 855 4.94 10% D A O - 10% H O - 80% F O 7.40 861 5.11 15% D A O - 2% H O - 83% F O 7.60 851 5.22 15% D A O - 3.5% H O - 87% F O 7.70 855 5.27 15% D A O - 5% H O - 80% F O 8.16 858 5.53 15% D A O - 10% H O - 75% F O 10.2 860 6.61 15% D A O - 15% H O - 70% F O 9.60 865 6.30 * Estimated at the film temperature using Puttagunta equation The kinematic viscosity of 10% D A O - 10% H O - 80% F O was measured at different temperatures and the results are plotted in Figure 3.2. The viscosity exponentially decreased with increasing temperature. Viscosity values at each temperature along with an exponential fit to the data is given in Appendix A 4 . Chapter 3: Experimental Materials and Apparatus 40 10 200 — I — 400 600 800 1000 Shear Rate (1/s) Figure 3.1: Dynamic Behavior of 15% D A O - 10% H O - 75% F O at 85°C Temperature (°C) Figure 3.2: Viscosity of 10% D A O - 10% H O - 80% F O at Different Temperatures Chapter 3: Experimental Materials and Apparatus 41 3.2 Experimental Apparatus 3.2.1 Thermal Fouling Test Apparatus Fouling runs were carried out in an existing UBC fouling loop, which is equipped with an electrically heated annular probe supplied by Ashland Chemical Company, Drew Divisions. The fouling loop is a recirculation system which consists of a supply tank for holding the liquid, a centrifugal pump, an orifice meter for measuring flow rates, the annular fouling probe, two pressure relief valves and a host of regulating valves. A schematic of the apparatus is shown in Figure 3.3. All surfaces in contact with the liquid other than the pump, were constructed from stainless steel. Liquid is pumped by a 2.2 kW centrifugal pump from a 9.45 L holding tank through the flow control valve, the orifice and the fouling probe before being returned to the tank. The fluid flows through two mixing chambers (MC) one upstream and another downstream of the test section where thermocouples measure the entry and exit bulk temperatures respectively. Most experiments were performed at a constant holding tank pressure of 411 kPa (45 psig). The flow rate is set using a control valve and it is measured using an orifice meter. The pressure drop across the orifice meter is obtained and the following equation is used to calculate the flow rate, (3.1) where V is the volume flow rate in m3/s, Ca is the orifice discharge coefficient found by Asomaning [1990], A o r is the cross sectional area of the orifice in m 2, A P is the pressure Chapter 3: Experimental Materials and Apparatus 42 drop across the orifice meter in Pa, p is the density of the fluid at bulk temperature in kg/m3, P = di/d2 where di and 62 are the diameters of the orifice plate and pipe respectively. 3.2.2 Annular Test Section The annular test section consists of the probe and an outer annular assembly. The design of the probe is shown in Figure 3.4. Liquid flows upwards through the annulus between the metal core and the outside wall. The heated section consists of a 32 ohm nichrome electric heater embedded in a ceramic matrix and sheathed with a stainless steel tube. The length of the heated section is 0.102 m. The surface temperature (Ts) is calculated from the temperatures, T m , measured by four thermocouples embedded in the sheath, a distance xs, below the surface, using the calibrations supplied by the supplier and the following relationship T s = T m - - ^ q (3.2) A m e t where q = Q/A = (power input to the probe)/(probe heat transfer area). The diameter of the probe and outer annulus are 0.011 m and 0.0254 m respectively. The probe is designed for a maximum power input of 1920 Watts. It operates at a constant heat flux with time, hence the fluid/deposit interface is assumed to be at a constant temperature. The reciprocal of clean overall heat transfer coefficient is calculated as follows Chapter 3: Experimental Materials and Apparatus 43 Air Rotameter RBPC Cooling -Water out STT Immersion Heater h - f X r -Supply Tank Bypass Valve Hx3-SS Q DPM Annular Test Section MC L J - " T b , i n Orifice Plate SS - Syringe Sampling (Septum) DPM RBPC - Rotameter Back Pressure Control p V - Outlet for Venting Gas STT MC - Mixing Chamber TC T b - Bulk Temperature Thermocouples - Pressure Gauge Figure 3.3 Schematic of Fouling Apparatus Chapter 3: Experimental Materials and Apparatus 44 To Datalogger To 240V Power Supply 294 mm Fi2ure 3.4: Heat Exchanger Fouline Probe The fouling reciprocal overall heat transfer coefficient is give by v s b ' t U(t) (3.4) Chapter 3: Experimental Materials and Apparatus 45 The fouling resistance could then be obtained using the expression I T - T . 1 - I T - T , I R f ( t ) = J — b / t V s b *> (3.5) f U(t) U(0) q V ' This fouling resistance was plotted against time to determine the thermal fouling profiles. The fouling rate can then be calculated as dR f _ d dt dt f 1 ^ v U ( t ) , (3.6) In some cases where the fouling rate exhibited an asymptotic behaviour the initial fouling rates were determined from fits of the fouling data to the Kern-Seaton equation, for times beyond the induction period R f = R f * ( l - e x p ( - b t ) ) (3.7) Whence R f o =bxR* f 4.0 E X P E R I M E N T A L P R O C E D U R E S This chapter will explain the procedures followed for thermal fouling runs, test fluid properties and oil compatibility tests. 4.1 Procedure for Thermal Foul ing Runs Cold Lake heavy oil was received from Imperial Oil Resources Ltd. in 5 gallon containers. The contents of the containers were all mixed together prior to the fouling runs to ensure uniformity. The bottom third of each container was first discarded to avoid the water that was found at the bottom of the container being mixed with the oil sample. This research is built on methodology developed previously by Asomaning (1997). Heavy Oil is weighed in a flask and heated on a hot plate to 30 - 40°C. De-asphated oil is added to the flask and mixed while heating. Fuel oil is added gradually to the flask to dissolve the heavy oil and de-asphalted oil. The mixture is then mixed thoroughly to ensure a uniform test solution. After filling the supply tank, the test fluid is purged with nitrogen for one hour under 410 kPa absolute pressure to eliminate any dissolved oxygen. After purging, the control panel is activated along with the pump and tank heater. If pressure in the system rises due to the presence of volatiles upon heating, the system is vented to achieve the desired pressure. Once the desired pressure is reached, the vent is closed completely for the experiment. When the bulk temperature reaches 35°C, the flow control valve is set to achieve the desired bulk velocity. Before starting the probe heater, a sample of the test fluid is taken for analysis. The power to the heat transfer probe is adjusted to achieve the desired initial surface temperature of 230°C. 46 Chapter 4: Experimental Procedure 47 The datalogger and the PC are activated to record system variables at a scanning interval of 10 minutes. A small flow of cooling water is used to maintain a constant bulk temperature throughout the run. The end of the experiment is identified when the maximum probe temperature is reached or a constant surface temperature is obtained. The power to the probe and the pump are switched off. A sample of test fluid is taken once again for analysis and the cooling water flow rate is increased to cool the system. The pressure of the system is released and the fluid is drained. The probe is taken out of the test section and rinsed with varsol, a paint thinner sold by Esso Chemicals, to dissolve the oil, followed by acetone to evaporate varsol. The probe is photographed and the deposits are removed mechanically, with care, and stored for further analysis. The rig is cleaned by pumping 9 liters of varsol for 30 minutes. This is followed by 8 liters of a 50-50 % acetone-toluene mixture that is circulated for only 15 minutes due to the corrosiveness of acetone and the toxicity of toluene. Finally, the rig is rinsed with 6 liters of acetone to help evaporate the cleaning solvents. The probe is cleaned with varsol and acetone prior to each run. 4,2 Determination of Pentane Insolubles and Hot Filtration Insolubles The test sample is analyzed for pentane insolubles, filtration insolubles, and viscosity at the beginning and end of each run. Pentane insolubles content is determined by measuring 2 ml of test sample into a flask and adding 80 ml of n-pentane. In the case of heavy oil, an equal volume of benzene is added to the sample prior to the addition of n-pentane. The sample is well mixed and placed in the dark for several hours with intermittent shaking. The content of the flask is then filtered at room temperature using a Chapter 4: Experimental Procedure 48 3 micron Millipore filter paper and a Millipore filter funnel holding 47 mm diameter filters. The precipitate is washed with n-pentane until the filtrate is clear. The filter with the precipitate is placed in an aluminum dish and dried over night in an oven at 100°C. Hot filtration insolubles are determined at the bulk temperature. The sample is heated in a beaker on a hot plate to 85°C. Ten ml of sample are measured into a graduated cylinder and filtered through a pre-weighed 3 micron membrane filter that has been washed with n-pentane. After filtration, the precipitate is washed with n-pentane until the filtrate is clear. The filter and the precipitate are placed in an aluminum dish and dried over night in an oven at 100°C. 4.3 Measurement of Test Fluid Properties The density and viscosity of mixtures are measured prior to each run. These properties are required to set the desired flow rate for each run. The viscosity of mixtures is determined using the Haake V T 500 Rotovisko, a computer controlled rotary viscometer. A N V sensor system consisting of a cup and a bell-shaped rotor is used. The sample is introduced into the space between the coaxial concentric cylinders. The outer hollow cylinder is stationary while the inner solid bob rotates. The bob is motor driven and its torque is measured by a force sensor. The viscosity of the fluid is a measure of the resistance of the fluid to the rotation of the inner bob. The connected computer uses the Haake V T 500 software to record the data and compute the variables. The shear stress of Newtonian fluids is directly proportional to the shear rate in the absence of turbulence. In this case, the absolute viscosity is given by Newton's equation: (4.1) Chapter 4: Experimental Procedure 49 where x is the shear stress and D is the shear rate. The densities of the mixtures are measured using a density bottle. The density bottle is rinsed with a solvent, dried and weighed. The sample is heated to the bulk temperature and charged to the density bottle. The bottle is then weighed and the density is determined as the ratio of the mass to the volume of the sample. 4.4 Procedure for Oil Compatibility Tests Oil compatibility tests were developed by Exxon Ltd. Detailed procedures were supplied by Wiehe (1999). The four tests outlined below permit one to determine compatibility of oils upon mixing. 4.4.1 Heptane Dilution Test This test method is used to determine the onset of precipitation during heptane dilution of oils that contain heptane insolubles (asphaltenes). N-heptane is added to 5 ml of sample until insoluble asphaltenes are detected. In case of viscous oils, the sample is heated to 70°C for half an hour at least with intermediate shaking using an ultrasonic bath to assure that the sample is well mixed. The point of precipitation is detected using a spot test, where a drop of the sample is placed on a 5 micron filter membrane using a medical dropper and examined. The point of asphaltene precipitation is taken where a definite ring of darker color is observed in the center of the spot with a lighter surrounding color. The point is always tested with a further dilution. If asphaltenes were present, the color change gets more definite. In the first trial, n-heptane is added in 5 ml increments until a definite change in the spot is observed. If asphaltenes are detected, the test is repeated with an amount of n-Chapter 4: Experimental Procedure 50 heptane equal to 4 ml less than the amount when asphaltenes were first detected in the first trial. In the second trial, n-heptane is added in 1 ml increments until asphaltenes are detected and confirmed with an extra dilution. The test is then repeated for a third time with an amount of n-heptane equal to 0.8 ml less than the amount when asphaltenes were first detected in the second trial. n-Heptane is added in 0.2 ml increments in this case until asphaltene detection is confirmed with an extra dilution. The heptane dilution is taken as the average of the total ml of n-heptane added when asphaltenes were first detected in the third trial and the highest total ml of n-heptane added without detecting asphaltenes. The results are reported as the volume of n-heptane in ml. added to 5 ml. of oil at the point just before insoluble asphaltenes first appear, V H . The range of the heptane dilution is from 0 to 25 ml. If no insolubles are detected at a heptane dilution of 25 ml., the oil is declared heptane soluble. 4.4.2 Toluene Equivalence Test This test method is used to determine the toluene equivalence of oils that contain heptane insolubles (asphaltenes). Ten ml. of a mixture of toluene and n-heptane are added to 2 grams of the sample oil and the presence or not of insoluble asphaltenes is detected using the spot test. In the case of viscous oils, the sample is heated to 7 0 ° C for half an hour at least with intermediate shaking using an ultrasonic bath to assure that the sample is well mixed. If insoluble asphaltenes are detected, the test is repeated but with a higher percentage of toluene in the n-heptane-toluene mixture. On the other hand, if no insoluble asphaltenes are detected in the first test, the second test is done with a lower percentage of toluene. This procedure is continued until the minimum percentage of toluene in the n-heptane-toluene mixture to keep asphaltenes in solution is determined to Chapter 4: Experimental Procedure 51 the desired accuracy. This minimum percentage of toluene is reported as the toluene equivalence, TE. The range of the toluene equivalence is from 0 to 100. If no insolubles are detected at a toluene equivalence of 0, the oil is declared heptane soluble. 4.4.3 Nonsolvent Oil Dilution Test This test method is used for an oil sample that is a nonsolvent for asphaltenes but contains no asphaltenes itself. In the present case, this is done on the fuel oil. The nonsolvent oil is blended with an asphaltene containing test oil that has been previously subjected to the Toluene Equivalence and Heptane Dilution tests. Heavy oil is used as the test oil in this case. Basically, for the Nonsolvent Oil Dilution test, the sample oil (or nonsolvent oil) is added to the test oil until insoluble asphaltenes are detected. The test is repeated to obtain the desired accuracy. The results are reported as the volume of nonsolvent oil in ml. added to 5 ml. of test oil at the point just before insoluble asphaltenes first appear, VNSD- The range of the nonsolvent oil dilution is from 0 to 25 ml. 4.4.4 Solvent Oil Equivalence Test This test method is used for an oil sample that is solvent for asphaltenes but contains no asphaltenes (soluble in heptane). In the present case, this is applied to the DAO. A test oil containing asphaltenes is needed that has been previously put through the Toluene Equivalence and Heptane Dilution tests. In this case, heavy oil is used as the test oil. Basically, for the Solvent Oil Equivalence test, toluene in the Toluene Equivalence Test is replaced by the sample oil (or solvent oil) and the equivalence test is rerun on the test oil. That is the ratio of solvent oil to n-heptane is varied and mixed with the test oil in the ratio of 10 ml. of solvent oil-heptane mixture per 2 grams of test oil. Chapter 4: Experimental Procedure 52 The minimum volume percent of solvent oil in the solvent oil-heptane mixture to keep the asphaltenes in solution is the Solvent Oil Equivalence, SOE. The range of values is from 0 to 100. 5.0 RESULTS AND DISCUSSION Experiments were carried out in this research to examine the effect of resins to asphaltenes ratio on heat exchanger fouling of heavy oil. A sample of de-asphalted oil was used to vary the Re/As ratio as it was difficult to separate large quantities of resins required for this study. Blends of varying concentrations of HO, DAO and FO were used in thermal fouling runs at similar conditions to examine their fouling behaviour. The results along with the discussion of these results will be presented in this chapter. 5.1 Typical Thermal Fouling Run Fouling runs are carried out for periods up to 30 hours. Examples of typical results obtained in a thermal fouling run are shown in Figures 5.1 and 5.2. The surface temperature starts rising as the deposit builds up beyond the induction period. The heat flux drops slightly in the beginning of the experiment, but it stays almost constant over the course of the experiment. The overall heat transfer coefficient decreases as deposit builds up on-heat exchanger surface and the fouling resistance increases correspondingly. Results shown in Figures 5.1 and 5.2 are for the experiment carried out with 5 % DAO -15 % HO and 80 % FO. A test of reproducibility is given in Appendix A3. 5.2 Effect of Resins to Asphaltenes Ratio on Heavy Oil Fouling The role of resins on asphaltene stability is not very well understood. There are views in support of the idea that resins keep asphaltenes in solution and form a stable mixture. However, others noted that not all types of resins have such an effect and it is important that the resins come from the same crude as the asphaltenes. Therefore, the effect of resins to asphaltenes ratio on the stability of asphaltenes in solution is examined 53 Chapter 5: Results and Discussion 54 Figure 5.1: Surface Temperature and Heat Flux for a Typical Fouling Run Figure 5.2: Overall Heat Transfer Coefficient and Thermal Resistance for a Typical Fouling Run Chapter 5: Results and Discussion 55 in this work through investigating the fouling behaviour of oil samples of varying resins to asphaltenes ratio. Blends of HO/FO/DAO were prepared with 5 wt % of DAO and different amounts of HO to test the effect of HO concentration and resins/asphaltenes ratio on the fouling rate. The fouling resistance over time for these runs is shown in Figure 5.3. There appears to be an induction period of 2 - 5 h, followed by fouling with a falling rate with time. For zero percent heavy oil, fouling is negligible. Thus on its own DAO does not cause fouling. As the HO concentration was raised from 0 to 20 %, the extent of fouling increased. The corresponding asphaltene concentration rose from 0.04 % to 3.36 %, and the Re/As ratio decreased from 81.3 to 1.37. The run at 10 % HO showed a shorter induction time, and had the most rapid rise in fouling resistance with time. At 15 % HO and 20 % HO, fouling behaviour was similar and the extent of fouling exceeded that at 10 % HO after about 14 hours, although the rate was lower. Table 5.1 lists the thermal fouling parameters for the series of experiments of 5 wt % DAO in HO/FO mixtures. The reciprocal of the clean overall heat transfer coefficient decreased with increasing percentage of HO in mixture. The final fouling resistance was consistent with the initial fouling rate. The initial fouling rates of these runs were determined by fitting the data to the Kern-Seaton [1959] asymptotic fouling model except for the run with 10 % HO where the slope was used to obtain the initial fouling rate. A program in M A T L A B was prepared using Marquardt's method to fit experimental data to the asymptotic fouling model. The parameters obtained along with the standard deviation of the fit can be found in Appendix A2. Chapter 5: Results and Discussion 56 -0.04 - | 1 1 1 1 • 1 • 1 > 1 r 1 0 5 10 15 20 25 30 Time (h) Figure 5.3: Fouling Resistance over Time of 5 wt % D A O in H O / F O Mixture Table 5.1: Thermal Fouling Parameters for Experiments of 5 wt % D A O in H O / F O Mixtures at T h of 85 °C, T«„ of 230 ° C and Uh of 0.75 m/s Heavy Oil (wt %) Re/As Heat Flux (kW/m2) Initial Fouling Rate (m 2 K/k\\h) 1/U„ (m 2K/k\V) Range of Initial Rate Calculation (h) Final R f (m 2K/k\V) 0 81.3 250 Neg. Fouling 0.57 0-26 0.008 5 4.0 366 0.020 0.385 3 - 22.5 0.164 10 2.3 365 0.039 0.395 1.1-6.5 0.272 15 1.7 411 0.026 0.354 3 - 19 0.305 20 1.4 451 0.026 0.301 3 - 16.5 0.269 Figure 5.4 shows the fouling resistance over time for 10 wt % DAO with different HO (and hence asphaltene) concentrations. Fouling was negligible in the absence of HO. It is evident in this case as well, that the extent of fouling increased with increasing HO Chapter 5: Results and Discussion 57 or asphaltene concentration, which corresponded to decreasing Re/As ratio. As in Figure 5.3, fouling was rapid and severe, reaching Rf values of 0.1-0.2 m 2K/kW in less than ten hours. Thermal fouling parameters of these runs are listed in Table 5.2. Heat flux increased while the reciprocal of the clean overall heat transfer coefficient decreased with increasing percentage heavy oil. Initial fouling rates were estimated using the slope of the curve in these runs except for the 10 % HO where the fouling model was used to obtain the initial fouling rate. Figure 5.4: Fouling Resistance over Time of 10 wt % D A O in H O / F O Mixture A series of experiments was performed at 15 wt % of DAO in which the effect of addition of low concentrations of HO was also examined. Results shown in Figure 5.5, indicate a dramatic increase in fouling as the HO concentration was increased from 3.5 to Chapter 5: Results and Discussion 58 5 wt %, which corresponded to a decrease in Re/As ratio from 5.9 to 4.5. As in Figure 5.3, fouling was most rapid at the intermediate concentration of 10 % HO. Table 5.2: Thermal Fouling Parameters for Experiments of 10 wt % D A O in H O / F O Mixtures at T h of 85 ° C T.» of 230 °C and U h of 0.75 m/s Heavy Oil (wt %) Re/As Heat Flux (kW/m2) Initial Fouling Rate (m 2K/kWh) 1/Uo (m 2K/k\V) Range of Initial Rate Calculation (h) Final R f (m 2K/kW) 0 45.5 243 Neg. Fouling 0.587 0-21 0.0049 3 6.4 238 Neg. Fouling 0.596 2-25 0.007 5 4.2 353 0.025 0.392 0- 18 0.183 10 2.4 452 0.045 0.324 1 - 11 0.279 Parameters of the thermal fouling runs for 15 % DAO series runs are available in Table 5.3. Results show that as percentage of heavy oil increased the heat flux required to achieve the desired initial surface temperature is increased. The reciprocal of the clean overall heat transfer coefficient decreased as heavy oil concentration increased in the mixture. The final fouling resistance values are consistent with the initial fouling rate calculated. The 5 % HO and 15 % curves were fitted to the Kern-Seaton model to obtain the initial fouling rate while the slope of the fouling resistance versus time plot was used for the 10 % HO concentration curve. Chapter 5: Results and Discussion 59 Figure 5.5: Fouling Resistance over Time of 15 wt % DAO in HO/FO Mixture Table 5.3: Thermal Fouling Parameters for Experiments of 15 wt % DAO in HO/FO Mixtures at T h of 85 °C, T«n of 230 °C and U h of 0.75 m/s Heavy Oil (wt %) Re/As Heat Flux (kW/m2) Initial Fouling Rate (m2K/k\\h) 1/U„ (m2K/k\V) Range of Initial Rate Calculation (h) Final R f (m2K/kW) 2 8.9 237 Neg. Fouling 0.587 0-27 0.017 3.5 5.9 237 Neg. Fouling 0.600 0-27 0.020 5 4.5 389 0.036 0.353 0-19 0.257 10 2.6 431 0.054 0.317 1 -6.3 0.305 15 1.9 516 0.039 0.231 0-11 0.271 Initial fouling rates for 0 % DAO, obtained by Asomaning [1997], 5 % DAO, 10 % DAO and 15 % DAO mixtures were plotted against their calculated asphaltene content Chapter 5: Results and Discussion 60 as shown in Figure 5.6. The graph indicates that for each feed mixture, the initial fouling rate increases with asphaltene content, reaches a maximum, and then decreases. The value of the maximum rate increases as the DAO and hence resin content in the mixture increases for a given percentage of asphaltene. The maximum occurs at an asphatlene content of = 1.7 % corresponding to a Re/As ratio of = 2.5. At asphaltene contents below 0.75 %, negligible fouling is observed. The initial fouling rates of mixtures containing similar asphaltene contents were plotted against their resins content in Figure 5.7. As the resin concentration is increased, higher fouling rates are encountered when it was expected that fouling would decrease based on available literature. Figure 5.7 clearly indicates that at a given asphaltene concentration, fouling rates increase with the resins content in this system. This suggests that there is some interaction between the resins and the asphaltenes which promotes fouling in this case. The effect of resin and asphaltene contents on the initial fouling rates of mixtures was examined further by plotting the initial fouling rates of the mixtures against Re/As ratios in Figure 5.8. For a given percentage DAO, initial fouling rates go through a maximum with increasing Re/As ratios. At Re/As < 2.5, the initial fouling rate increased as the Re/As ratio increased. However, for values > 2.5 the initial fouling rate decreased as the Re/As ratio increased with a dramatic decrease in the initial fouling rate at Re/As > 4. For mixtures containing DAO, fouling rates were essentially zero at Re/As > 6. Chapter 5: Results and Discussion 61 0.060 0.055 - • 0 % DAO 1 Re/As-2.5 0.050 - 9 5 % DAO /'j 0.045 - - - A 10% DAO / t 0.040 - ••• 15% DAO /I9 0.035 - y / j 0.030 - / / ' ' , * / 0.025 - / / \ e 0.020 - )"/ \ 0.015 - \ 0.010- :U 0.005 -0.000 --0.005 -// 8Sk-— i 1 1 1 1 1 — -1 1 T I 1 | • | — i 1 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Calculated Asphaltene Content (wt. %) 4.0 Figure 5.6: Relationship of Initial Fouling Rate with Calculated Asphaltene Content for 0. 5.10 and 15 wt % D A O in H O / F O Mixture 0.010 -\ 1 1 i 1 • 1 > 1 ' r 2.5 3.0 3.5 4.0 4.5 5.0 Calculated Resins Content (wt. %) Figure 5.7: Relationship of Initial Fouling Rate with Calculated Resins Content at Constant Asphaltenes Content Chapter 5: Results and Discussion 62 0.060 R e / A s Figure 5.8: Relationship of Initial Fouling Rate with Re/As ratio for 0, 5, 10 and 15 wt % D A O in H O / F O Mixture Results in Figure 5.8 showed that at a constant Re/As ratio, the initial fouling rate increased as the DAO concentration increased in the mixture. This indicates that the addition of DAO caused incompatibility in the mixture and therefore increased the fouling rate. As noted previously the presence of DAO did not cause fouling in the absence of heavy oil. Examination of Figure 5.8 reveals the complexity of the role of resins in asphaltenes stability. For Re/As ratios larger than 2.5, results show a decrease in initial fouling rate indicating a possible role of resins in keeping asphaltenes in solution and therefore increasing compatibility. On the other hand, at lower Re/As ratios it seems that Chapter 5: Results and Discussion 63 there are factors overcoming the role of resins in keeping asphaltenes in solution. It is indicated by Leontaritis et al. [1988], that there is a threshold concentration of resins required to keep asphaltenes in solution below which asphaltenes would precipitate. It could be possible that a Re/As ratio 2.5 indicates the threshold- resins concentration ratio required for this system to enable the positive role of resins in stabilizing the crude and below which its role is eliminated. On the other hand, asphaltenes concentration varied from 0.04 - 3.4 wt %, while the resins content was limited to between 3.1 - 4.9 wt %. The wide variation in the resulting Re/As ratio (from 1.4 to 81.3) was dictated primarily by the changing asphaltene content. The narrow range of resin concentration variation in the mixtures made it difficult to study its role as a peptizing agent in more detail. If the fuel oil content was kept low, the viscosity of the mixture became too high for pumping. In addition, the high saturates content of these mixtures (55.3 - 67.3 %) made it even harder to examine the effect of other constituents. Aromatics varied between 28.7 and 37.1 %. Therefore, when the initial fouling rates were plotted versus (Ar + Re)/As in Figure 5.9, the graph was similar to that of Figure 5.8. A detailed list of constituent concentrations of test fluids is available in a later section in Table 5.6. 5.3 Effect of Resins to Asphaltenes Ratio on Hot Filtration and Pentane Insolubles Pentane insolubles basically provide a measure of the total asphaltenes content in the mixture. Since hot filtrate insoluble material is asphaltene, its measurement represents the flocculated asphaltene concentration. The hot filtration insolubles test therefore gives a rough indication of the compatibility of the mixture. There is some inaccuracy in measurements of hot filtration insolubles which is due to the low Chapter 5: Results and Discussion 64 concentrations of insolubles present at bulk conditions and in some cases samples were not filtered directly after sampling which might have affected its concentration of filterable solids. Properties of all test fluids are presented in Table 5.4. Results for concentrations of hot filtration insolubles and pentane insolubles are plotted against Re/As ratio for all test fluids in Figure 5.10. . a o 0.060 0.055 H 0.050 0.045 0.040 -\ 0.035 0.030 0.025 H 0.020 0.015 H 0.010 0.005 0.000 ^  -0.005 10 — • — 0 % DAO (Asomaning 1997) ® 5 %DAO — A — io % DAO —v— 15 % DAO - T 1 1 1—i—i—i—p 100 (Re+Ar)/As - A 1000 Figure 5.9: Relationship of Initial Fouling Rate with (Re + Art/As Content in Mixture for 0. 5.10 and 15 wt % DAO in HO/FO Mixture Figure 5.10 shows the variation of pentane insolubles and hot filtration solids with the Re/As ratio for all the mixtures. Pentane insolubles and hot filtration insolubles concentrations were measured at the beginning and end of each run and the average was used in this study since there was a negligible difference in both measurements. The pentane insolubles (and calculated asphaltene contents) decreased monotonically with the increasing Re/As ratio. The hot filtration solids appeared to first increase and then to Chapter 5: Results and Discussion Table 5.4: Properties of Test Fluids Test Fluid Hot Filtration Insolubles (a/i) Pentane Insolubles (g/1) Initial Fouling Rate (m2K/kWh) Reynolds Numbers* Reynolds Numbers"1" Re/As 0 % DAO 5 %HO 95 %FO 2.7 7.3 0.010 1998 2734 3.7 5 %DAO 0 %HO 95 %FO 0.9 1.3 Neg. fouling 1748 2449 81.3 5 %DAO 5 %HO 90 % FO 4.0 8.4 0.020 1816 2527 4.0 5 %DAO 10 %HO 85 %FO 2.5 10.2 0.039 1376 2016 2.3 5 %DAO 15 %HO 80 %FO 3.6 14.2 0.026 1433 2083 1.7 5 %DAO 20 %HO 75 %FO 3.3 21.0 0.026 1250 1865 1.4 10 %DAO 0 %HO 90 %FO 1.6 1.6 Neg. fouling 1680 2370 45.5 10 % DAO 3 % HO 87 %FO 2.4 5.2 Neg. fouling 1549 2218 6.4 10 % DAO 5 %HO 85 %FO 4.2 9.4 0.025 1512 2175 4.2 10 %DAO 10 %HO 80 %FO 5.2 11.3 0.045 1453 2106 2.4 15 %DAO 2 %HO 83 %FO 0.9 5.0 0.001 1414 2060 8.9 15 %DAO 3.5 %HO 81.5 %FO 1.2 5.8 0.001 1396 2040 5.9 15 %DAO 5 % K O 80 %FO 5.2 10.9 0.036 1317 1945 4.4 15 %DAO 10 %HO 75 %FO 5.4 11.8 0.054 1054 1626 2.6 15 %DAO 15 %HO 70 %FO 7.2 18.6 0.039 1120 1707 1.9 * Calculated based on bulk temperature properties + Calculated based on estimated film temperature properties Chapter 5: Results and Discussion 66 decrease with the ratio Re/As. Comparing pentane insolubles values with the hot filtration results suggests that on average, less than 50 % of the potential asphaltenes are insoluble at the bulk temperature, whereas the rest remain in solution. At low Re/As ratios, only 15 % of the asphaltenes are insoluble, whereas at the highest Re/As levels, about 70 % of the asphaltenes are insoluble. ^ 25 • -59 <D £ 20 • O a is 12 10 S3 O o OH • Hot Filtration Insolubles e Pentane Insolubles - i 1 1 r -10 Re/As 100 Figure 5.10: Relationship of Properties of Mixtures with Re/As ratio for all oil Mixtures Pentane Insoluble contents of test fluids are plotted against their Re/As ratio for various DAO concentrations in Figure 5.11. The graph shows the decrease in pentane insolubles with increasing Re/As ratio. In addition, it is clear that at a constant Re/As ratio, the pentane insoluble concentration increases with increasing DAO content in the mixture for Re/As < 5. This result indicates that DAO addition to the mixture increases the content of pentane insolubles and therefore affects the compatibility of the mixture. However, it Chapter 5: Results and Discussion 67 is necessary to make sure that this increase in pentane insolubles content is not due to an increase in the total calculated asphaltene content of the mixture due to the addition of DAO. Therefore, the pentane insoluble concentrations are plotted versus the calculated asphaltene content for various DAO concentrations as shown in Figure 5.12. Examination of Figure 5.12 shows that the pentane insoluble concentration increases with increasing DAO content at a constant asphaltenes content. These findings lead to the conclusion that the addition of DAO to the mixture may enhance the formation of asphaltenes that were not present in the mixture and therefore affect the oil compatibility. 25 20 jE^ 15 o C O I 10 OH -•—5 %DAO ® 10 % DAO A -- 15 % DAO A._ • - A - T 1 1 — I — I — r I I I — I — r 10 Re/As 100 Figure 5.11: Pentane Insolubles Variation with Re/As ratio for Various DAO Concentrations The effect of resins concentration on hot filtration insolubles at contant asphaltene content is shown in Figure 5.13. As the resin concentration is raised at a given asphaltene level, the hot filtration solids concentration generally increased indicating less compatibility in the mixture. This result supports the fouling rate findings in Figure 5.7. Chapter 5: Results and Discussion 68 Figure 5.12: Measured Pentane Insloluble Concentration Variation with Calculated Asphaltene Contents for Various DAO Concentrations 7.5 2.0 H 1 1 1 1 1 1 1 1 r 3.0 3.5 4.0 4.5 5.0 Calculated Resins Content (wt. %) Figure 5.13: Measured Hot Filtration Insoluble Concentration Variation with Calculated Resins Contents for Various Asphaltene Concentrations Chapter 5: Results and Discussion 69 Hot filtration insoluble concentrations are plotted against the Re/As ratios for various DAO concentrations in Figure 5.14. Although there is considerable scatter in the data, it appears that hot filtration insolubles generally decrease with increasing Re/As for all DAO concentrations. Mixtures with higher DAO concentrations have a higher hot filtration insoluble concentration at a constant Re/As ratio, with the exception of the data at 15 % DAO (Re/As > 5). This figure shows that DAO is generally playing a negative role in the compatibility of the mixture as noted previously with the pentane insolubles concentrations. 7.5-7.0 H 6.5-6.0-5.5-5.0-4.5-4.0-3.5-3.0-2.5-2.0-1.5-1.0-0.5-• 5 %DAO a 10 % DAO A 15 % DAO • i i i i i | l 10 Re/As 100 Figure 5.14: Hot Filtration Insolubles Variation with Re/As ratio for Various D A O Concentrations A rough correlation of the fouling rate with concentration of hot filtration solids is shown in Figure 5.15, where the dashed line, and some of the data are taken from Asomaning and Watkinson [1999]. The Asomaning data shows fouling rates increasing Chapter 5: Results and Discussion 70 by a factor of about one hundred as the insoluble solid concentration goes from about 2 g/L to 6 g/L. The present data shows as two horizontal clusters on the plot. The initial fouling rate is about 0.001 m 2 K / k W h where the concentration of precipitated solids is from about 0.7 to 1.5 g/L, and at insolubles concentrations of 3-7 g/L, the rate did not change substantially from its average value of 0.03 m 2 K / k W h . o.i-d o o.oi-j I? 1 fa 13 1E-3 ' 4 3 • Asomaning [1997] e This Work i l i I i l 0 1 2 3 T — 1 — r 4 5 Hot Filtration Insolubles (g/L) Figure 5.15: Initial Fouling Rate Dependence on Solids ConcentrationstTh 85" C Tg, 230° C . V 0.75 m/s) Hot filtration insolubles results are plotted in terms of percentages of H O and F O in Figure 5.16. The plot shows general trends of increasing hot filtration insolubles with increasing concentrations of both H O and D A O . Although there is some scatter, the contours do show a general trend which suggests that increasing the D A O concentrations makes the system less compatible at any level of H O . As well , where no heavy oil is present, a small concentration of filterable solids exists. Chapter 5: Results and Discussion 71 Elemental analysis was performed on a sample of hot filtration insolubles collected by fdtering an oil mixture of 15 % DAO - 15 % HO - 70 % FO. Results of sample analyses are show in Table 5.5. The H/C ratio is lower in this sample of precipitate than that found in the original test fluids, HO, DAO and FO as reported in Tables 3.1, 3.3 and 3.4; and is very close to values of asphaltenes given by Suzuki et al. [1982] as presented in Table 2.2, and by Strausz [1992]. Table 5.5: Elemental Analysis of Hot Filtration Insolubles Oil Sample C H N S H/C atomic 15%DAO-15%HO-75%FO 77.38 7.09 1.36 8.91 1.1 •A g -•—0 %HO ® -5 %HO A 10% HO - T - 15 % HO 10 % DAO i 12 14 16 Figure 5.16: Relationship of % DAO and % HO to Hot Filtration Insolubles As well, as expected for asphaltenes, the nitrogen content is found to be higher and the sulfur content is much higher than that of the original test oils. Chapter 5: Results and Discussion 72 5.4 Deposit Characterization Fouling occurs on the surface of the probe in the fouling rig. Deposit builds up as the run proceeds appearing as a thin black layer on the surface, for low fouling rates as shown in Figure 5.17, and a thick black layer in cases of high fouling rates as it appears in Figure 5.18. Figure 5.19 shows a close-up of the thick deposit evenly distributed along the heated length and the roughness is apparent at the surface. Deposits are collected at the end of each run after photographing the probe. Elemental analyses were performed for some deposit samples and these results are presented in Table 5.6. Chemical characteristics of collected probe deposit show that the H/C atomic ratio of the deposits is in the range 1.2-1.3, nitrogen content is 0.77- 1.1 %, and the sulphur content is 4.5-5.8 %. These values are characteristic of asphaltenes as reported by Wiehe [1999b]. On average, the H/C ratios and nitrogen contents of the deposits are similar to that of the hot filtration insolubles as shown in Table 5.5 indicating that insolubles are depositing on the heat exchanger surface causing fouling. However, the sulfur content of the hot filtration insolubles is much higher than that of the deposits. Table 5.6 Chemical Characteristics of Probe Deposits of Some Runs C H N S . H/C atomic 0%DAO-5%HO-95%FO 79.83 8.07 1.1 4.46 1.21 5%DAO-5%HO-90%FO 72.22 7.10 0.86 5.83 1.18 10%DAO-5%HO-85%FO 75.19 8.06 0.82 4.78 1.29 15%DAO-5%HO-80%FO 78.24 8.67 0.77 4.68 1.33 Chapter 5: Results and Discussion 73 Figure 5.18: Deposit Formed on Probe Surface for Run with 10 % D A O - 1 0 % H O - 80 % F O with High Fouling Rate. Chapter 5: Results and Discussion 74 Figure 5.19: Close-Up of Deposit Formed on Probe Surface for R u n wi th 10 % D A O - 1 0 % H O - 80 % F O with H i g h Foul ing Rate. The H/C atomic ratio and the nitrogen content are plotted versus the percentage of DAO in the mixture in Figure 5.20. It is shown that the H/C atomic ratio of the deposits increases and nitrogen content decreases with increasing DAO content. These results show that DAO may contribute to precipitation as the properties of the deposits formed are consistently changing with DAO concentrations. Fouling deposits of selected runs were further studied using a scanning electron microscope (SEM) along with an Energy-Dispersion X-ray (EDX). The SEM micrographs of the deposits of these runs are shown in Figures 5.21, 5.23 and 5.25 while the EDX plots are shown in Figures 5.22, 5.24 and 5.26. Chapter 5: Results and Discussion 75 1.35 —•— H/C Atomic Ratio 1.30 125 o O 1.20 > 3 o' 70 0.7 1.15 0 2 4 6 8 10 12 14 16 Wt % DAO Figure 5.20: Deposit Characteristics Var ia t ion with D A O Content in M i x t u r e SEM micrographs of fouling deposit show clusters of agglomeration that could be asphaltenes that have undergone some form of chemical change on the hot probe surfaces. The E D X analyser is attached to the SEM to examine the deposits for the presence of elements using carbon as the standard. The E D X analyses showed the presence of sulphur, silicon, sodium, chlorine and trace quantities of sodium and copper. The large content of sulphur is noticeable in all samples. The presence of most of these elements is believed to be due to heavy oil since DAO and fuel oil samples are processed while heavy oil is not. However, it should be noted that the quantities of these elements are very low and they are present in most fluid mixtures in similar quantities, therefore their effect on the fouling behaviour of these mixture is believed to be negligible. Chapter 5: Results and Discussion 76 Figure 5.21: S E M Mic rog raph of Foul ing Deposit for R u n with 5 % D A O - 5 % H O and 90 % F O Counts 1875 4 1500 4 1125 4 Figure 5.22: E D X Plot of Foul ing Deposit for R u n with 5 % D A O - 5 % H O and 90 % F O Chapter 5: Results and Discussion 77 H H S Figure 5.23: S E M Micrograph of Foul ing Deposit for R u n wi th 15 % D A O - 5 % H O and 80 % F O Counts 1065 852 4 639 - J 426 213 Figure 5.24: E D X Plot of Fouling Deposit for R u n with 15 % D A O - 5 % H O and 80 % F O Chapter 5: Results and Discussion 78 M l Figure 5.25: S E M Mic rograph of Fouling Deposit for R u n with 5 % D A O - 15 % H O and 80 % F O Counts 1025 4 820 615 410 J 205 . Cu 8 9 keV Figure 5.26: E D X Plot of Foul ing Deposit for R u n with 5 % D A O - 15 % H O and 80 % F O Chapter 5: Results and Discussion 79 5.5 Colloidal Instability Index Asomaning and Watkinson [1997] reported that their fouling rate data for HO-FO mixtures, including those with pentane and xylene additions, could be correlated by the Colloidal Instability Index. However the present data which involves HO-FO-DAO mixtures could not. The reason for the failure of the CII to correlate the present data is unclear; however, it is noted that in contrast to the present data, all of Asomaning's data except one point was taken at Re/As less than 2.5, and hence below the condition for maximum fouling rate shown in Figure 5.8. Both sets of results were explored through the fouling regime map of Figure 5.27. The boundary between a "no-fouling" regime (initial fouling rate < 0.001 m2K7kWh), and the "fouling" regime can be approximated by the condition, R f o - » 0 , for CH < (Re/As) 0 3 (5.1) Additional data are required to fix this boundary with greater accuracy, however Figure 5.27 strongly suggests that additional parameters beyond the CII are needed to predict fouling rates in asphaltene-containing systems where a wide variation in Re/As ratio exists. The Asomaning data showed measurable fouling rates once CII reached about 1.2, whereas the present data suggest that CII values as high as 1.6 will not produce fouling, if the ratio of Re/As is large. Calculated oil constituent contents of test fluids along with their CII values are presented in Table 5.7. Chapter 5: Results and Discussion 80 10-Fouling • • • y 0.1 • A No Fouling • Th i s W o r k (Fouling) ® Asoman ing 1997 (Fouling) ^ Th i s W o r k (No Foul ing) • Asoman ing 1997 ( N o Foul ing) 10 100 R e / A s Figure 5.27: Foul ing Regime M a p ( " N o Fou l ing" corresponds to R f < or equal to 0.001 m 2 K / k W h ) Chapter 5: Results and Discussion -81 Table 5.7: Compositions of Test Fluids Test Fluid Calculated Saturates Content (%) Calculated Aromatics Content (%) Calculated Resins Content (%) Calculated Asphaltenes Content (%) Re/As CII 4 % DAO 5 % HO 95 % FO 67.30 28.77 3.10 0.83 3.7 2.14 6 % DAO 7 % HO 95 % FO 67.17 29.70 3.09 0.04 81.3 2.05 8 % DAO 9 % HO 90 % FO 64.85 30.81 3.47 0.87 4.0 1.92 10 % DAO 11 %HO 85 % FO 62.53 31.92 3.86 1.70 2.3 1.80 12 %DAO 13 % HO 80 % FO 60.20 33.03 4.24 2.53 1.7 1.68 14 % DAO 20 %HO 75 % FO 57.88 34.14 4.62 3.36 1.4 1.58 10 %DAO 0 % HO 90 %FO 64.73 31.74 3.46 0.08 45.5 1.84 10 %DAO 3 % HO 87 % FO 63.34 32.45 3.69 0.57 6.4 1.77 10 %DAO 5 % HO 85 % FO 62.40 32.85 3.84 0.91 4.2 1.73 10 %DAO 10 %HO 80 %FO 60.08 33.96 4.22 1.74 2.4 1.62 15 % DAO 2 % HO 83 % FO 61.35 34.20 3.98 0.45 8.9 1.62 15 % DAO 3.5 %HO 81.5 %FO 60.66 34.60 4.09 0.70 5.9 1.59 15 %DAO 5 %HO 80 %FO 59.96 34.89 4.21 0.95 4.4 1.56 15 %DAO 10 %HO 75 % FO 57.64 36.0 4.59 1.78 2.6 1.46 15 %DAO 15 %HO 70 %FO 55.31 37.11 4.97 2.61 1.9 1.38 Chapter 5: Results and Discussion 82 5.6 Oil Compatibility Model Relation to Asphaltenes Fouling The Oil Compatibility Model as proposed by Wiehe [1999] is used to predict the order of blending and the ratio of certain oils to ensure compatibility of the oil mixture. It is assumed that if the fluid mixture is incompatible, fouling will occur. In this study, Oil Compatibility tests were performed on test fluids HO, DAO and FO to obtain model parameters required to enable correlation of model results with available fouling data. Two sets of tests were performed. In the first set, heavy oil was used as the reference oil and both DAO and FO were tested against it. A blend of 50-50 weight percent of HO-D AO was used as the reference oil and FO was tested against it in the second test set. Results obtained in both test sets are discussed in the following sections. 5.6.1 Heavy Oil as Reference Oil Heavy oil contains 16 % of asphaltenes and therefore is defined as heptane insoluble. However, DAO and FO have low asphaltene contents and therefore are defined as heptane soluble. Heptane Dilution tests and Toluene Equivalence tests were performed on HO. Since FO is not a good solvent for asphaltenes, the Nonsolvent Dilution test was performed. The solvency of DAO for asphaltenes was not clearly known, therefore both the Nonsolvent Oil Dilution and the Solvent Oil Dilution tests were performed. 5.6.1.1 Test Results Heavy oil and DAO are very viscous at room temperature. Therefore, the sample was usually heated to 70°C for half an hour at least with intermediate shaking using an ultrasonic bath to lower the viscosity of the mixture and assure adequate mixing. Chapter 5: Results and Discussion 83 The Heptane Dilution test was performed for heavy oil. It was found that 9.2 ml. of n-heptane was needed to precipitate asphaltenes. The Toluene Equivalence test was performed on heavy oil and it was found that a 23.5% toluene in a toluene-n-heptane test sample is required to keep asphaltenes in solution at a concentration of 10 ml. of test sample to 2 grams of oil. Heptane Dilution test was also carried out for DAO to ensure that it is heptane soluble. There were no asphaltenes detected after the addition of 25 ml. of n-heptane to 5 ml. of DAO, so it was confirmed that it is heptane soluble. The Nonsolvent Oil Dilution test was performed on fuel oil using heavy oil as the reference oil. It was found that 5.75 ml. fuel oil can be added to 5 ml. of heavy oil before precipitating asphaltenes. The Nonsolvent Oil Dilution test was performed on De-asphalted oil and it was found to be asphaltene soluble at a concentration of 25 ml. DAO to 5 ml. heavy oil. Therefore, the Solvent Equivalence test was performed and it was found that 37.50 % of DAO is required in a DAO-n-heptane mixture to keep asphaltenes in solution at a concentration of 10 ml. of DAO-n-heptane mixture to 2 grams of heavy oil. Results of these tests are listed in Table 5.8. It should be noted that major difficulties were encountered in trying to detect the point of asphaltene precipitation. This was mainly due to the high viscosity of the mixtures and the dark color of heavy oil that made it very hard to identify the end point clearly. It is suspected that these results have low accuracy; however, the trend is believed to be of importance here. This work was done to check whether the extent of Chapter 5: Results and Discussion 84 incompatibility from this model would correlate with the initial fouling rates obtained for these mixtures. Table 5.8: Oil Compatibility Test Results using H O as the Reference Oil Heptane Dilution test ml. of n-heptane added to 5 ml. HO 9.2 Toluene Equivalence test % toluene in 10 ml. of toluene-n-heptane mixture added to 2 grams of HO 23.5 Nonsolvent Oil Dilution test (FO) ml. of FO added to 5 ml. HO 5.75 Solvent Oil Equivalence test % DAO in 10 ml. of DAO-n-heptane mixture added to 2 grams of HO 47.5 5.6.1.2 Model Prediction Model parameters were calculated using equations 2.4 to 2.7. Results obtained are shown in Table 5.9. Wiehe [1999c], had previously performed these tests on Cold Lake HO and found that HO gave an IN of 30 and SBN of 81 compared to an IN of 37.1 and SBN of 105.5 obtained in this study. Results obtained by Wiehe give a V H of 8.5 ml and T E of 30 % compared to a V H of 9.2 ml and T E of 23.5 %. Test results obtained in this study are close to those obtained by Wiehe considering that both samples were not identical. Therefore test results can be considered accurate within experimental error. The solubility blending number of all the mixtures were calculated using equation 2.8 and the volume fraction along with the solubility blending numbers of individual oils. Solubility blending numbers of all mixtures are listed in Table 5.10. The maximum Chapter 5: Results and Discussion 85 Table 5.9: Calculated O i l Compatibi l i ty M o d e l Parameters Using H O as Reference IN SBN HO 37.1 105.5 FO 0 -22.3 (calculated using eqn 2.10) DAO 0 49.5 (calculated using eqn 2.11) insolubility index in all mixtures is that of heavy oil being 37.1. Therefore, to achieve compatibility for these mixtures, the solubility blending number should be greater than 37.1. As seen in table 5.10, all test fluids were found to be incompatible according to the compatibility criterion, having solubility blending numbers between -18.3 and 7.6. The results obtained from this model did not correlate with the initial fouling rate results obtained for these mixtures. For example, the two mixtures with the most negative blending numbers contained no heavy oil, and showed negligible fouling. The three mixtures with the highest positive blending numbers, showed significant but not the highest fouling rates. However, it was found that the compatibility results of this model correlate well with the colloidal instability index CII as shown in Figure 5.28. This could suggest that both tests give similar results therefore, it is preferred to use the CII since it is based on more accurate test methods. 5.6.2 Heavy O i l - D A O Blend as Reference O i l Oil compatibility tests were repeated using a 50-50 mixture of HO-D AO as the reference oil instead of HO. The objective of these tests was to examine the role of DAO on the stability of asphaltenes. It is reported in literature that resins and aromatics are peptizing agents that keep asphaltenes in solution. DAO is low in asphaltenes and Chapter 5: Results and Discussion 86 Table 5.10: Oil Compatibility Model Prediction for Test Fluids Test Fluid Solubility Blending Numbers 0 % DAO - 5 % HO - 95 % FO -15.6 5 % DAO - 0 % HO - 95 % FO -18.3 5 % DAO - 5 % HO - 90 % FO -12.0 5 % DAO - 10 % HO - 85 % FO -5.77 5 % DAO -15 % HO - 80 % FO 0.5 5 % DAO - 20 % HO - 75 % FO 6.88 10 % DAO - 0 % HO - 90 % FO -14.7 10 % DAO - 3 % HO - 87 % FO -11.0 10 % DAO - 5 % HO - 85 % FO -8.5 10 % DAO -10 % HO - 80 % FO -2.2 15 % DAO - 2 % HO - 83 % FO -8.7 15 % DAO - 3.5 % HO - 81.5 % FO -6.8 15 % DAO - 5 % HO - 80 % FO -4.9 15 % DAO -10 % HO - 75 % FO 1.4 15 % DAO -15 % HO - 70 % FO 7.6 contains high percentages of aromatics and resins. Therefore, it was blended with heavy oil to observe its effect on the stability of asphaltenes. 5.6.2.1 Test Results Heptane dilution test was performed on a mixture of 2.5 ml. of HO and 2.5 ml of DAO. Insoluble asphaltenes were detected after adding 3.75 ml. of n-heptane to the oil blend. This n-heptane volume is much less than 9.2 ml., which was added to 5 ml. of HO before asphaltenes were detected. This shows that DAO is not helping in keeping asphaltenes in solution, instead, it caused asphaltenes to precipitate more readily. Chapter 5: Results and Discussion 87 The Toluene Equivalence test was also performed using 1 gram of HO and another gram of DAO. The oil mixture was heated to 70°C for half an hour with intermediate shaking in the ultrasonic bath to ensure mixing. It was found that 37.5 % of toluene in a toluene-n-heptane mixture was required to keep asphaltenes in solution compared to 23.5 % which was required for the 2 grams of HO. This also indicates that DAO is playing a negative role in stabilizing asphaltenes in solution. Figure 5.28: Relationship of Oil Compatibility Model Index to Colloidal Instability Index The Nonsolvent Oil Dilution test was performed. Fuel oil was added to a 5 ml. of a 50-50 mixture of HO-DAO mixture. In this case, asphaltenes were detected after adding 1.75 ml. of FO to 5 ml. oil mixture compared to 5.75 ml. when HO was used as the reference oil. All the results obtained in this part of the study show that DAO is playing a negative role in asphaltene stability and therefore causing more precipitation of asphaltenes. Results are shown in Table 5.11 Chapter 5: Results and Discussion 88 Table 5.11: Oil Compatibility Test Results Using H O - D A O as The Reference Oil Heptane Dilution Test ml. of n-heptane added to 5 ml. HO-DAO 3.75 Toluene Equivalence Test % toluene in 10 ml. of toluene-n-heptane mixture added to 2 grams of HO-DAO 37.50 Nonsolvent Oil Dilution Test (Fuel Oil) ml. of FO added to 5 ml. HO-DAO 1.75 5.6.2.2 Model Prediction Model Parameters were calculated using equations 2.6, 2.7 and 2.10. The results obtained are shown in Table 5.12. Results show a discrepancy between the SBN index calculated for FO using heavy oil as the reference oil and that calculated using HO-DAO as the reference oil. However, when calculations were carried out, it was found that the difference in the indices is due to a volume difference of 0.75 ml. This indicates that the model parameters are very sensitive to a small error in test results, Test results obtained in this study are of greater importance at this stage of the research, than the calculated model parameters, in explaining the fouling behavior observed for these oil mixtures. Table 5.12: Calculated Oil Compatibility Model Parameters Using H O - D A O Blend as Reference IN SBN HO-DAO Blend 43.8 76.7 FO 0 -50.1 (calculated using eqn 2.10) 6.0 CONCLUSIONS AND R E C O M M E N D A T I O N S 6.1 Conclusions This study of heat exchanger fouling of mixtures of heavy oil , de-asphalted oil and fuel oi l at initial surface temperature of 230 °C and bulk temperature o f 85 °C, and a fixed velocity, led to the following conclusions: • A t a fixed D A O concentration, the fouling rate first increased, and then decreased as the H O concentration was raised from 0 to 20 % and as Re/As ratio decreased. The initial fouling rate passed through a maximum as the Re/As ratio was raised over the range 1.3 to 81.3. The maximum fouling rate occurred at Re/As = 2.5 (which corresponded to about 1.75 % asphaltenes), and decreased to essentially zero for Re/As > 5.8. • Fouling rate, pentane insolubles concentrations and hot filtration insolubles concentrations all increased as D A O concentration was raised at a fixed Re/As ratio. This suggests that resins from the D A O are involved in enhancing precipitation, and fouling. • High fouling rates were generally encountered at high pentane insolubles concentrations, which increased as D A O concentration was raised at a fixed Asphaltenes concentration. 89 Chapter 6: Conclusions and Recommendations 90 • Hot filtration insolubles and probe deposits have chemical properties similar to asphaltenes. There was some indication of higher H/C ratios in deposits where D A O concentrations were higher. • Fouling rates for mixtures containing DAO, did not correlate with the colloidal instability index alone. A fouling regime map indicated that low fouling rates were dependent on both the colloidal instability index and the resins/asphaltene ratios. High colloidal instability indexes could be tolerated, provided that Re/As ratio was sufficiently large. • Oil Compatibility Model predictions correlated well with the colloidal instability index, and therefore also did not predict which mixture would foul most heavily. • Oil compatibility model titrations showed that all mixtures used were incompatible to some extent. This was also reflected by the presence of hot filtration insolubles in all fluid mixtures. Addition of DAO to heavy oil caused asphaltenes to precipitate at a lower concentration of n-heptane. Hence blending of DAO with heavy oil required a higher percentage of toluene in a toluene-n-heptane mixture to keep asphaltenes in solution. This result was consistent with the findings on the role of D A O in raising hot filtration insolubles concentrations and fouling rates. Model parameters are very sensitive to small errors in test method results. 6.2 Recommendations The role of resins in asphaltene stability is not very clear in this study due to the narrow range of resin concentrations covered. This was due to the presence of other oil Chapter 6: Conclusions and Recommendations 91 constituents in large quantities in the test samples which might have overcome the effect of the resin fraction. Therefore, it is recommended to isolate a resin fraction from a heavy oil sample through precipitation or adsorption/desorption methods. This would have the advantage of both the asphaltenes and the added resins originating from the same source. The concentrations of hot fdtration insolubles in the samples were too low to detect with high accuracy levels, therefore it is recommended to use a larger oil sample in the fdtration tests and use insulation around the filter funnel to reduce heat loss and keep the sample at the desired bulk temperature. In this study, experiments were performed at a constant bulk temperature. It is recommended to examine this behaviour at higher bulk temperatures to examine the effect of temperature on the fouling behaviour of these mixtures. It was noted in this study that the pentane insolubles concentration increased as DAO concentration was raised in the mixture. This issue can be investigated further by preparing different oil mixtures and measuring their pentane insolubles. The asphaltene precipitation detection method used in the oil compatibility testing involves high levels of inaccuracy. A better method could provide an accurate measure of oil compatibility and improve the chance of correlating oil properties to fouling behaviour. Abbreviations and Nomenclature 92 Abbreviations API American Petroleum Institute Ar Aromatics As Asphaltenes A S T M American Standard Test Methods CII Colloidal Instability Index C M C Critical Micelle Concentration DAO De-Asphalted Oil E D X Energy-Dispersion X-Ray FO Fuel Oil HO Heavy Oil HPLC High Pressure Liquid Chromatography O C M Oil Compatibility Model Re Resins Sa Saturates SARA Saturates, Aromatics, Resins and Asphaltene Fractionation SEM Scanning Electron Microscope VPO Vapour Pressure Osmometry Nomenclature Aor Cross sectional area of the orifice m 2 b Constant in Kern-Seaton equation s"1 C Concentration mol/L, g/L Abbreviations and Nomenclature c d Orifice discharge coefficient -d Diameter m D Shear rate 1/s d e q Equivalent diameter of annular test section m H v Heat of vaporization J/mol, kJ/n IN Insolubility number -k f Thermal conductivity of deposit kW/m2K m Mass deposit per unit area kg/m2 P Pressure Pa q Heat flux kW/m2 Q Power input W Pv Universal gas Constant J/mol.K Rf Thermal fouling resistance m2K/kW Rr* Asymptotic fouling resistance m2K/kW Initial fouling rate m2K7kWh SBN Solubility blending number -SNSO Solubility blending number for non-solvent oils -SOE Solvent oil equivalence % Sso Solubility blending number for solvent oils -t Time s T Temperature ° C , K T b Bulk temperature °C, K T E Toluene equivalence % 94 T m Thermocouple temperature reading ° C , K T s Surface temperature ° C , K U Overall heat transfer coefficient kW/m 2K V Molar volume m3/mol V Volumetric flowrate m3/s V H Heptane dilution volume ml VNSD Volume for non-solvent dilution test ml X Deposit thickness m X S Thermocouple depth from surface m Subscripts a Component a f Foulant 0 Initial s Surface Greek P Ratio of orifice diameter to pipe diameter -r Gibbs surface excess mol/m2 Y Suface tension of the solution N/m 5 Solubility parameter MPa 0' 5 ^met Thermal Conductivity of metallic wall kW/m. P Density kg/m3 Dynamic viscosity Pa.s X Shear stress N / m 2 .3\0.5 References 95 References Andersen, S. I. and J. G. Speight, "Observation of The Critical Micelle Concentration of Asphaltenes", Fuel 72, 1343-1345, (1993). Andersen, S. I., and K. S. Birdi, "Aggregation of Asphaltenes as Determined by Calorimetry" Journal of Colloid and Interface Science 142, 497, (1991). Asomaning, S. and A.P. Watkinson, "Deposit Formation by Asphaltene-Rich Heavy Oil Mixtures on Heat Transfer Surfaces Preprints" Preprints, Engineering Foundation Conference, Italy (1997). Asomaning, S., "Heat Exchanger Fouling by Petroleum Asphaltenes", PhD Thesis, University of British Columbia, (1997). Asomaning, S. and A.P. Watkinson, "Stability and Heteroatom Effects in Heat Exchanger Fouling by Petroleum Asphaltenes" paper submitted to Can. J. Chem. Eng., July (1999). A S T M D2007- 93: "Standard Test Method for Characteristic Groups in Rubber Extender and Processing Oils by the Clay-Gel Adsorption Chromatographic Method", A S T M (1993). Barre, L. , D. Espinat, E. Rosenberg and M . Scarsella, "Colloidal Structure of Heavy Crudes and Asphaltene Solution", Revue De LTnstittut Francais Du Petrole 52, (1997). Blazek J., and G. Sebor, "Formation of Asphaltenes and Carbenes During Thermal Conversion of Petroleum Maltenes in Presence of Hydrogen", Fuel 72, 199-204 (1993). Bott, T. R , "The Cost of Fouling", in Fouling of Heat Exchangers, Elsevier Scinece B.V., (1994). Clarke, P. F. and B. B. Pruden, "Asphaltene Precipiatation: Detection Using Heat Transfer Analysis, and Inhibition Using Chemical Additives", Fuel 76, 607-614, (1997). Crittenden, B. D., S. T. Kolaczkowski and S. A. Hout, "Modeling Hydrocarbon Fouling,: Chem. Eng. Res. & Des. 65, 171-179 (1987b). Dickakian, G. and S. Seay, "Asphaltene Precipitation Primary Crude Exchanger Fouling Mechanism", Oil and Gas Journal, 47-50, (1988). Dickakian, G. B. and S. Seay, "Asphaltene Precipitation Primary Crude Exchanger Fouling Mechanism" Oil and Gas Journal 86, 47-50 (1988). References 96 Dikie, J. P. and T. F. Yen, "Macro-structures of the Asphaltic Fractions by Various Instrumental Methods" Anal. Chem. 39, 1847-1851, (1967). Eaton, P. and R. Lux, "Laboratory Fouling Test Apparatus for Hydrocarbon Feedstocks" ASME HTD 35, 33-42 (1984). Epstein, N., "A Model of the Initial Chemical Reaction Fouling Rate for Flow within a Heated Tube and Its Verification", Proc. 10th International Heat Transfer Conf, Brighton, U K (1994). Epstein, N., "Thinking about Heat Transfer Fouling: A 5 X 5 Matrix", Heat Transfer Engineering 4, 43-56 (1983). Espinat, D., "Colloidal Structure of Asphaltenes Solutions and Heavy-oil Fractions Studied by Small-angle Neutron and X-ray Scattering". SPE # 25187. New Orleans, L A , pages 365-373, (1993). Ferworn, K. A., A. K. Mehrota and W. Y. Svrcek, "Measurement of Asphaltene Agglomeration from Cold Lake Bitumen Diluted with n-Alkanes", Can. J. Chem. Eng. 71, 699-703 (1993). Galtsev, V. E. , I. M . Ametov and O. Y. Grinberg, "Asphaltene Association in Crude Oil as Studied by ENDOR" Fuel 74, 670, (1995). Gastel, C , R. Smadja and K. A. Lamminan, "Contribution a La Connaissance Des Proprietes Des Bitumen Routiers", Revue Gingirale De Routes Et Des Airodromes 466, 85, (1971). Hasson, D., "Precipitation Fouling", in Fouling of Heat Transfer Equipment Sommerscales, E.F.C. and J. G. Knudsen (eds.) Hemisphere Publishers, Washington DC (1981). Hilderbrand, J. H , J. M . Prausnitz and R. L. Scott, Regular Solutions, Van Nostrand-Reinhold, New York (1970). Hirschberg, A., L. N. J. de Jong, B. A. Schipper, and J. G. Meijers, "Influence of Temperature and Pressure on Asphaltene Flocculation", Soc. Pet. Eng. J., 283 (1984). Hunt, A., "Uncertainties Remain in Predicting Paraffin Deposition", Oil and Gas J. 94, 96-103, (1996). Jamaluddin, A. K. M . , T. W. Nazarko, and S. Sills, "Deasphalted Oil: A Natural Asphaltene Solvent", SPE Production and Facilities, 161-165, (1996). Kawanaka, S., S. J. Park and G. A. Mansoori, "The role of asphaltene deposition in EOR gas flooding: a predictive technique" SPE 17376. Oklahoma, (1988). References 97 Kawanaka, S., S. J. Park and G. A. Mansoori, "Organic Deposition from Reservoir Fluids: A Thermodynamic Predictive Technique", Soc. Petrol. Eng. Res. Eng. 5, 185-194 (1991). Kern, D. Q. and R. E. Seaton, "A Theoretical Analysis of Thermal Surface Fouling", Br. Chem. Eng. 4, 258-262 (1959). Koots, J. A. and J. G. Speight, "Relation of Petroleum Resins to Asphaltenes", Fuel 54, 179-184, (1975). Lambourn, G. A. and M . Durrieu, "Fouling in Crude Oil Preheat Trains" in Heat Exchangers Theory and Practice, Taborek Hewitt and Afghan (eds.) 219-261, GRETh Seminar, Grenoble (1992). Leontariitis, K. J., and G. A. Mansoori, "Asphaltene Flocculation During Oil Produciton and Processing: A Thermodynamic Colloidal Model", SPE 16258, SPEInt. Symp. Oilfield Chemistry, San Antonio, T X (1987). Leontaritis, K. J. and G. A. Mansoori, "Asphaltene Deposition: A Survey of Field Experiences and Research Approaches" J. of Pet. Sc. and Eng. 1, 229, (1988). Leontaritis, K. J., S. Kawanaka and G. A. Mansoori, "Descriptive Accuonts of Thermodynamic and Colloidal Models of Asphaltene Flocculation", In Oil Field Chemistry, American Chemical Society, Washington DC, 196-204, (1988). Li, S., C. Liu, G. Que, W. Liang and Y. Zhu, "A Study of the Interactions Responsible for Colloidal Structures in Petroleum Residua", Fuel 76, 1459-1463, (1997). Lian, H , J. R. Lin and T. R Yen, "Peptization Studies of Asphaltene and Solubility Parameter Spectra" Fuel 73, 423-428, (1994). Lichaa, P. M. , "Asphaltene Deposition Problem in Venezuela Crudes - Usage of Asphaltenes in Emulsion Stability" Can. Pet. Tech. J., oil sands, 609-624, (1977). Long, R. B., "The Concept of Asphaltene", Adv. Chem. Ser. 195, 17-27 (1981). Mansoori, G. A., "Asphaltene, Resin, and Wax Deposition from Petroleum Fluids" The Arabian Journal for Science and Engineering 21, 48, 707-723, (1996). Mitchel, D. L. and J. G. Speight, "The Solubility of Asphaltenes in Hydrocarbon Solvents", Fuel 52, 149-152 (1973). Murphy, G. and J. Campbell, "Fouling in Refinery Heat Exchangers: Causes, Effects, Measurements and Control", in Fouling Mechanisms. M . Bohnet et al. (eds.) 249-261, GRETh Seminar, Grenoble (1992). References 98 Park, S. J. and G. A. Mansoori, "Aggregation and Deposition of Heavy Organics in Petroleum Crude" Energy Sources 10, 109, (1988). Patterson, W. R. and P. J. Fryer, "A Reaction Engineering Approach to the Analysis of Fouling", Chem. Eng. Sci. 43, 1714-1717 (1988). Pfeiffer, J. P. and R. N. Saal, Journal of Physical Chemistry 44, 139, (1940). Pinhero, J. D. R. S., "Fouling of Heat Transfer Surfaces" in Heat Exchangers Thermal Hydraulic Fundamentals and Design. McGraw Hill, New York (1981). Pnachal, C. B. and A. P. Watkinson, "Chemical Reaction Fouling Model for Single-Phase Heat Transfer", A.I.Ch.E Symp. Ser. 295, 89, 323-334 (1993). Ravey, J. C , G. Ducouret, and D. Espinat, "Asphaltene Macrostructure by Small Angle Neutron Scattering" Fuel 67, 1560, (1988). Rogel, E. , "Theroretical Estimation of the Solubility Parameter Distributions of Asphaltenes, Resins, and Oils from Crude Oils and Related Materials" Energy and Fuels 11, 920-925, (1997). Sheu, E. Y., "Physics of Asphaltene Micelles and Microemulsions-theory and Experiment" J. Phys.: Condens. Matter 8, A125-A141, (1996). Speight, J. G. in The Chemistry and Technology of Petroleum. Marcel and Dekker Inc. New York (1991). Storm, D. A., R. J. Barresi, E. Y. Sheu, A. K. Bhattacharya and T. F. DeRosa, "Microphase Behavior of Asphaltic Micelles During Catalytic and Thermal Upgrading", Energy and Fuels 12, 120-128, (1998). Strausz, O. P. and I. Rubinstein in Atomic and Nuclear Methods in Fossil Fuel Research. Royston, Filby and Carpenter, (eds.), Plenum Publishers, New York (1980). Strausz, O. P., T. W. Mojelsky and E. M. Lown, "The Molecular Structure of Asphaltene: An Unfolding Story", Fuel 71, 1355-1363 (1992). Thawer, R , D. C. A. Nicoll, and G. Dick, "Asphaltene Deposition in Production Facilities", SPE Production Engineering, 475-480, (1990). Watkinson, A. P. and N. Epstein, "Gas Oil Fouling in a Sensible Heat Exchanger", CEP Symp. Ser. 65 No. 92, 84-90 (1969). Watkinson, A. P., "Chemical Reaction Fouling of Organic Fluids", Chem. Eng. Technol. 15, 82-90, (1992). References 99 Wiehe, I. A. and K. S. Liang, "Asphaltenes, Resins, and Other Petroleum Macromolecules", Fluid Phase Equilibria 117, 201-210, (1996). Wiehe, I. A., "Solvent-Resid Phase Diagram for Tracking Resid Conversion", Ind. Eng. Chem. Res. 31, 530-536, (1992). Wiehe, I. A., "The Oil Compatibility Model and Crude Oil Incompatibility", Preprints of International Conference on Petroleum Phase Behavior and Fouling, AIChE, Spring Meeting, Houston, (1999a). Wiehe, I., A., "Tutorial on the Phase Behavior of Asphaltenes and Heavy Oils", Preprints of International Conference on Petroleum Phase Behavior and Fouling, A.I.Ch.E., Spring Meeting, Houston, (1999b). Wiehe, I. A. - Private Communication (1999c). Wiehe, I. A., Ind. Eng. Chem. Res. 32, 2447-2457, (1993). Yen T. F., "Structure of Petroleum Asphaltene and Its Significance" Energy Sources 1, 447, (1974). Zhang, G. and A. P. Watkinson, "Modeling of Autoxidation Fouling", presented at C.S.Ch.E Conf, Vancouver (1991). Appendices 100 Table A l . l : STIMMARY OF FOULING RUNS* Test Fluid 5 95 % DAO %HO %FO % DAO %HO 95 %FO 0 90 % DAO %HO %FO Tb,avg (°C) 86.3 83.8 86.1 Ts,i 226 231 233 Heat Flux (kW/m2) 350 250 366 Initial Fouling Rate (m2K/kWh) 0.013 Neg. Fouling 0.020 1/U. (m2K/kW) 0.40 0.57 0.385 Range of Initial Rate Calculation (h) 4-32 0-26 3 - 22.5 Final R f (m2K/kW) 0.141 0.008 0.164 5 % DAO 10 %HO 85 %FO 5 % DAO 15 %HO 80 %FO 85.7 228 365 0.039 0.395 86.1 233 411 0.026 0.354 1.1-6.5 3 - 19 0.272 0.305 5 % DAO 20 %HO 75 %FO 86.1 230 451 0.026 0.301 0-16.5 0.269 10 %DAO 0 % HO 90 %FO 10 %DAO 3 % HO 87 %FO 10 %DAO 5 % HO 85 %FO 10 10 80 % DAO %HO %FO 84.7 232 243 Neg. Fouling 0.587 85.1 230 238 Neg. Fouling 0.596 87 232 353 0.025 0.392 86.6 231 452 0.045 0.324 0-21 2-25 0-18 1 - 11 0.0049 0.007 0.183 0.279 15 2 83 % DAO % HO %FO 15 %DAO 3.5 %HO 81.5% FO 15 %DAO 5 % HO 80 %FO 15 %DAO 10 %HO 75 %FO 15 15 70 % DAO %HO %FO 86.6 230 237 Neg. Fouling 0.587 85.4 227 237 Neg. Fouling 86.3 231 85.5 222 86.3 214 389 0.036 431 0.054 516 0.039 0.600 0.353 0.317 0.231 0-27 0-27 0-19 1-6.3 0-11 * Detailed data points for all runs are available on diskette from A. P. Watkinson 0.017 0.020 0.257 0.305 0.271 Appendices 101 ^2 Sample Calculations 1 Bulk Velocity and Reynolds Number The flow rates are calculated using Equation 3.1 V = C d A o r \ 2(AP) C d = 0.6102 d 2 = 0.0158 m (pipe diameter) di = 0.0008 m (orifice diameter) fj = 0.5024 Aor= 4.7 x 10"5 m 2 p =851 kg/m3 (15% DAO - 2% HO - 83% FO) For a velocity of 0.75 m/s, the volumetric flow rate is at follows, V = u • ACT where, A = - ( d 2 0 -d 2 i )=- (o .025 2 -0.0103 2) 4 4 - 0.000401 m 2 where d 0 = annulus outer diameter di = annulus inner diameter therefore, the volumetric flow rate is, V = (0.000401) (0.75) V = 0.0003 m3/s using the calculated flow rate, then the required pressure drop across the orifice to obtain the desired velocity is given by solving the following for P, 0.0003 = (0.6102)(4.7xl0-5) P = 43592.62 Pa 2p 85l(l-0.50244) Appendices 102 Since the mercury manometer is used, then the required manometer reading required is given by, AP = A z ( p H g - p f ) g A z = 0.350m A z = 13.78 inches given P H G = 13543 kg/m3. The bulk reynolds number is given by, u d e q Re = 2_ where d e Q is the equivalent diameter of the annulus - (d0 - d;) q = 0.025-0.0103 =0.0143 m and the viscosity of this fluid is 7.6 x 10"6 m2/s (0.75X0-0147) R e - , 7.6xl0 - 6 Re =1414 2. Test Fluid Concentration Heavy oil and DAO are measured by mass while FO is measured by volume, using its density. Normally for a 10 % heavy oil mixture, 860 g of heavy oil is utilized. Calculating the required volume of FO for preparing 15% DAO - 2% HO - 83% FO mixture is as follows, 0.17 = W H 0 + W D A 0 W H 0 + W D A 0 -1- HFO V FO + P F 0 V F Appendices 103 0.172 kg+ 1.290 kg 0.17 = - ~ — 851*| V m y V V F O 0.172 kg+ 1.290 kg + V F O = 8.39 Liters of FO 3. Thermal Fouling Resistances Sample calculation will be performed for run with 5 % DAO - 15 % HO - 80 % FO. The data logger records the voltage and the current at each data point. The power supplied to the probe is given by, Q = V x l Q = (213X6.83) Q = 1455 W the heat flux is given calculated as follows, Q _ 1455^ W q A / , 2 / l 0 0 0 W (3.41xl0- 3m 2l v \ l k W q = 426.7 kW/m 2 where A is the surface area of the heated section of the probe obtained as follows, A = T C / D A = 7t( 0.1016 m) (0.0107 m) A = 3.41 x 10"3 m 2 The reciprocal of the clean overall heat transfer coefficient for T s o = 232.7 °C and Tb 85.9 °C is given by, Appendices 104 _L - ZkzZk = = Q 3 5 4 m 2 K / k W U o q 0 426.7 The thermal resistance or the reciprocal of overall heat transfer coefficient at time t (under fouled condition, in this case at 8 hours), is evaluated as follows, J _ = T * - T b = 272.5-86 = Q _ 4 7 2 m 2 K y k W U f q 409.4 The thermal fouling resistance is calculated as follows, R f =— — = 0.472- 0.354 = 0.118m2K/kW U f U 0 4. Initial Fouling Rate Table A2 lists the calculated initial fouling rates, method of determination and the regression of the slope or the standard deviation of the model fitting. In cases where the fouling resistance did not reach the asymptotic fouling value, the initial fouling rate is obtained by determining the slope of the fouling resistance versus time curve as, d R f _ r1 f 1 ^ dt dt vU( t )y However, in cases where asymptotic behavior is observed, the Kern-Seaton asymptotic fouling resistance model is used to fit the data and obtain the initial fouling rate as follows, R f = R f * ( l - e x p ( - b t ) ) Whence R f =bxR* f Appendices 105 Table A2.1: Modeling Values of Initial Fouling Rates of A l l Mixtures Test Fluid Initial Fouling Rate (m2K/kWh) Modeling/ Slope R 2 for rates obtained using slope Fitted to Asymptotic Fouling Model Standard Deviation R / b 0 % DAO 5 % HO 95 %FO 0.013 Modeling - 4.40E-3 0.151 0.089 5 % DAO 0 % HO 95 %FO Neg. Fouling - - . - - -5 % DAO 5 % HO 90 %FO 0.020 Modeling - 3.81E-3 0.188 1.070 5 % DAO 10 %HO 85 %FO 0.039 Slope 0.988 - - -5 % DAO 15 %HO 80 %FO 0.026 Modeling - 5.29E-3 0.578 0.047 5 % DAO 20 %HO 75 %FO 0.026 Modeling - 5.29E-3 0.578 0.047 10 %DAO 0 % HO 90 %FO Neg. Fouling - - - - -10 %DAO 3 % HO 87 %FO Neg. Fouling - - - - -10 %DAO 5 % HO 85 %FO 0.025 Slope 0.990 - - -10 %DAO 10 %HO 80 %FO 0.045 Modeling - 5.90E-3 0.451 0.097 15 %DAO 2 % HO 83 %FO Neg. Fouling - - - - -15 %DAO 3.5 %HO 81.5% FO Neg. Fouling - - - - -15 %DAO 5 % HO 80 %FO 0.036 Modeling - 7.68E-3 0.281 0.127 15 %DAO 10 %HO 75 %FO 0.054 Slope 0.987 - - -15 %DAO 15 %HO 70 %FO 0.039 Modeling - 4.00E-3 0.447 0.088 Appendices 106 5. Colloidal Instability Index The colloidal instability index is calculated using the definition, Q J _ (Saturates + Asphaltenes) (Aromatics + Resins) Table A2.2 Average SARA Analysis of Working Fluids. Working Fluid Saturates Weig Aromatics ht % Resins Asphaltenes Heavy Oil 23.1 49.8 10.4 16.6 DAO 20.7 68.5 10.1 0.8 Fuel Oil 69.6 27.7 2.7 Trace The CII for mixtures are calculated as follows, (0.166WHO +0.008WDAO)+(0.231WHO +0.207WDAO +0.696WFO) (0.104WHO +0.101WDAO +0.027WFO)+(0.498WHO +0.685WDAO +0.277WFO) 80 % FO oil mixture, the CII works out to be, (0.0253 + 0.6018) 6 g (0.0423 + 0.329) ~ 6. Oil Compatibility Model Parameters Insolubility index for heavy oil is obtained as follows, H 25p For 5% DAO - 1 5 % H O -CII = Appendices 107 T E = 23.5 % of toluene required in a toluene-heptane mixture to keep asphaltenes in solution in the toluene equivalence test V H = 9.2 ml of heptane resulted in asphaltene precipitation in the heptane dilution test, p = (1.038)(0.997 g/L) which is the density of heavy oil at 25°C. 23.5 9.2 25 1.038x0.997 g = 37.1 The solubility blending number of heavy oil is given by, 5 SB N=37.1x 1 + -9.2 The non-solvent oil dilution test is performed for fuel oil and its solubility blending number is calculated as follows, S-ro IVNSD V H ] JNSO V NSD 1 + ^ 5 Sxo = Solubility blending number for heavy oil. Y N S D = 5 75 m i of fuel oil that results in asphaltene precipitation in the non-solvent oil dilution test. Appendices 108 105.5[5.75-9.2] 'NSO 5.75 1 + 9.2 -22.3 The solvent oil equivalence test is used for DAO and its solubility parameter is calculated as follows, S s o = 100 T E SOE T E = Tolene equivalence test result for heavy oil SOE = 47.5 % of DAO required to keep asphaltene in solution in a heptane-DAO solution. s s o = i o o | 23.5 47.5 = 49.5 The solubility blending number for test fluid mixtures are obtained as follows, ' BNmix V , + v 2 + v 3 + . . . For a mixture of 5 % DAO - 15 % HO - 80 % FO, the solubility blending number is calculated as ' BNmix 5(49.5) +15(105.5) +80(-22.3) 100 0.50 Appendices 109 A3 Reproducibility of Thermal Fouling Experiments It is critical to examine the reproducibility of certain sets of data. Therefore, the experiments with 5% DAO -15% HO - 80% FO and 51% DAO -10% HO - 75% FO were repeated. Table A3.1 lists the properties of these runs. Table A3.1 Test of Reproducibility of Data Trial No. Test Fluid Tb,avg (°C) Ts,i Heat Flux (kW/m2) Initial Fouling Rate (m2K/kWh) 1/Uo (m2K/kW) Range of Initial Rate Calculation (h) Final R r (m2K/kW) 1 5 % DAO 15 %HO 80 %FO 86.0 227.5 406 0.026 0.35 3 . 7 - 2 0 . 3 0.322 2 5 % DAO 15 %HO 80 %FO 86.1 232.7 411 0.026 0.35 3.1 - 19.3 0.305 1 15 %DAO 10 %HO 75 %FO 85.5 221.6 431 0.054 0.317 0 - 6 . 3 0.305 2 15 %DAO 10 %HO 75 %FO 85.3 228.0 411 0.050 0.35 0 - 1 0 . 5 0.312 The thermal fouling resistance results of both runs are shown in Figures A3.1 and A3.2. Results in Figure A3.1 are very similar, however results in Figure A3.2 show some discrepancy in both runs with time which may be due to small differences in variables. Results indicate that the results of the thermal fouling experiments are reliable and reproducible within experimental error. Sources of error in thermal fouling experiment can be due to several factors. These include consistency in test sample properties, cleaning of the fouling rig, maintaining constant variables such as bulk temperature, pressure and heat flux over the Appendices 110 course of the experiment. These sources of errors can be avoided by through mixing of the test samples upon receiving to insure consistency of oil samples. In addition, the fouling rig is cleaned rigorously prior to each experiment to avoid contamination of test fluid. The probe is cleaned thoroughly prior to each experiment to ensure consistency in the clean overall heat transfer coefficient. Great caution is taken through the course of experiment to ensure constant variables through out the experiment. 0.35-, , -0.05 -| , 1 1 1 1 1 • r 0 5 10 15 20 Time (h) Figure A3.1: Fouling Resistance over Time for A Repeat Run with 5% DAO -15% HO - 80% FO Oil Mixture. Appendices 111 o° CP o.oo H • First Trial o Second Trial T • r Time (h) 10 12 Figure A3.2: Fouling Resistance over Time for A Repeat Run with 15% D A O -10% H O - 75% F O Oil Mixture. Appendices 112 A4 Viscosity Data Table A4.1 Kinematic Viscosiv of a mixture of 10% DAO, 10% HO and 80% FO at Various Temperatures Temperature (°C) Kinematic Viscoisty (m /s) xlO 6 (m2/s) 25 76.894 40 31.609 60 14.429 70 10.387 85 6.905 92 5.705 Figure A4.1 Fitting Kinematic Viscosity Varation with Temperature of 10% DAO, 10% HO and 80% FO Oil Mixture into a First Order Exponential Decay Function 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0058530/manifest

Comment

Related Items