106,99 €
Mehr erfahren.
- Herausgeber: Wiley-VCH
- Kategorie: Fachliteratur
- Sprache: Englisch
A pioneering and comprehensive introduction to the complex subject of integrated refinery process simulation, using many of the tools and techniques currently employed in modern refineries. Adopting a systematic and practical approach, the authors include the theory, case studies and hands-on workshops, explaining how to work with real data. As a result, senior-level undergraduate and graduate students, as well as industrial engineers learn how to develop and use the latest computer models for the predictive modeling and optimization of integrated refinery processes. Additional material is available online providing relevant spreadsheets and simulation files for all the models and examples presented in the book.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 523
Veröffentlichungsjahr: 2013
Ähnliche
Contents
Foreword by Steven R. Cope
Foreword by Lawrence B. Evans
Preface
Acknowledgements
About the Authors
1 Characterization, Physical and Thermodynamic Properties of Oil Fractions
1.1 Crude Assay
1.2 Pseudocomponent Generation Based on Boiling-Point Ranges
1.3 Workshop 1.1 – Interconvert Distillation Curves
1.4 Workshop 1.2 – Extrapolate an Incomplete Distillation Curve
1.5 Workshop 1.3 – Calculate MeABP of a Given Assay
1.6 Workshop 1.4 – Duplicate the Oil Fraction in Aspen HYSYS Petroleum Refining
1.7 Property Requirements for Refinery Process Models
1.8 Physical Properties
1.9 Process Thermodynamics
1.10 Miscellaneous Physical Properties for Refinery Modeling
1.11 Conclusions
1.12 Nomenclature
1.13 References
2 Atmospheric Distillation Unit
2.1 Introduction
2.2 Scope of the Chapter
2.3 Process Overview
2.4 Model Development
2.5 Feed Characterization
2.6 Data Requirements and Validation
2.7 Representative Atmospheric Distillation Unit
2.8 Building the Model in Aspen HYSYS
2.9 Results
2.10 Model Applications to Process Optimization
2.11 Workshop 2.1 –Rebuild Model Using “Back-blending” Procedure
2.12 Workshop 2.2 – Investigate Changes in Product Profiles with New Product Demands
2.13 Conclusions
2.14 Nomenclature
3 Vacuum Distillation Unit
3.1 Process Description
3.2 Data Reconciliation
3.3 Model Implementation
3.4 Model Applications to Process Optimization – VDU Deep-Cut Operation
3.5 Workshop – Using Aspen HYSYS Petroleum Refining to Implement the Deep-Cut Operation
3.6 References
4 Predictive Modeling of the Fluid Catalytic Cracking (FCC) Process
4.1 Introduction
4.2 Process Description
4.3 Process Chemistry
4.4 Literature Review
4.5 Aspen HYSYS Petroleum Refining FCC Model
4.6 Calibrating the Aspen HYSYS Petroleum Refining FCC Model
4.7 Fractionation
4.8 Mapping Feed Information to Kinetic Lumps
4.9 Overall Modeling Strategy
4.10 Results
4.11 Model Applications to Process Optimization
4.12 Model Application to Refinery Production Planning
4.13 Workshop 4.1: Guide for Modeling FCC Units in Aspen HYSYS Petroleum Refining
4.14 Workshop 4.2: Calibrating Basic FCC Model
4.15 Workshop 4.3: Build Main Fractionator and Gas Plant System
4.16 Workshop 4.4: Model Applications to Process Optimization – Perform Case Study to Identify Different Gasoline Production Scenarios
4.17 Workshop 4.5: Model Application to Production Planning – Generate DELTA-BASE Vectors for Linear-Programming (LP)-Based Production Planning
4.18 Conclusions
4.20 Nomenclature
4.21 References
5 Predictive Modeling of the Continuous Catalyst Regeneration (CCR) Reforming Process
5.1 Introduction
5.2 Process Overview
5.3 Process Chemistry
5.4 Literature Review
5.5 Aspen HYSYS Petroleum Refining Catalytic Reformer Model
5.6 Thermophysical Properties
5.7 Fractionation System
5.8 Feed Characterization
5.9 Model Implementation
5.10 Overall Modeling Strategy
5.11 Results
5.12 Model Applications to Process Optimization
5.13 Model Applications to Refinery Production Planning
5.14 Workshop 5.1: Guide for Modeling CCR Units in Aspen HYSYS Petroleum Refining
5.15 Workshop 5.2: Model Calibration
5.16 Workshop 5.3: Build a Downstream Fractionation
5.17 Workshop 5.4: Case Study to Vary RON and Product Distribution Profile
5.18 Conclusions
5.19 Nomenclature
5.20 References
6 Predictive Modeling of the Hydroprocessing Units
6.1 Introduction
6.2 Aspen HYSYS Petroleum Refining HCR Modeling Tool
6.3 Process Description
6.4 Model Development
6.5 Modeling Results of MP HCR Process
6.6 Modeling Results of HP HCR Process
6.7 Model Applications to Process Optimization
6.8 Model Application – Delta-Base Vector Generation
6.9 Conclusions
6.10 Workshop 6.1 – Build Preliminary Reactor Model of HCR Process
6.11 Workshop 6.2 – Calibrate Preliminary Reactor Model to Match Plant Data
6.12 Workshop 6.3 – Model Applications to Process Optimization
6.13 Workshop 6.4 – Connect Reactor Model to Fractionator Simulation
6.14 Nomenclature
6.15 References
Supporting Materials: List of Computer Files
Subject Index
Related Titles
Ancheyta, J.
Modeling and Simulation of Catalytic Reactors for Petroleum Refining
2011
ISBN: 978-0-470-18530-8
Al-Qahtani, K. Y., Elkamel, A.
Planning and Integration of Refinery and Petrochemical Operations
2010
ISBN: 978-3-527-32694-5
Lieberman, N.
Troubleshooting Process Plant Control
2008
ISBN: 978-0-470-42514-5
Georgiadis, M., Kikkinides, E. S.,
Pistikopoulos, E. (Eds.)
Process Systems Engineering
Volume 5: Energy Systems Engineering
2008
ISBN: 978-3-527-31694-6
Elvers, B. (ed.)
Handbook of Fuels
Energy Sources for Transportation
2008
ISBN: 978-3-527-30740-1
Papageorgiou, L., Georgiadis, M. (Eds.)
Process Systems Engineering
Volume 3: Supply Chain Optimization
2008
ISBN: 978-3-527-31693-9
Papageorgiou, L., Georgiadis, M. (Eds.)
Process Systems Engineering
Volume 4: Supply Chain Optimization
2008
ISBN: 978-3-527-31906-0
Wiley
Wiley Critical Content
Petroleum Technology, 2 Volume Set
2007
ISBN: 978-0-470-13402-3
Proud, J. F.
Master Scheduling
A Practical Guide to Competitive Manufacturing
2007
ISBN: 978-0-471-75727-6
Bloch, H. P.
A Practical Guide to Compressor Technology
2006
ISBN: 978-0-471-72793-4
The Authors
Ai-Fu Chang, Kiran Pashikanti, and Y. A. Liu
Department of Chemical Engineering
Virginia Polytechnic Institute and State University
Blacksburg, Virginia 24061
USA
All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.
Library of Congress Card No.:
applied for
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library.
Bibliographic information published by the Deutsche Nationalbibliothek
The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at <http://dnb.d-nb.de>.
© 2012 Wiley-VCH Verlag & Co. KGaA, Boschstr. 12, 69469 Weinheim, Germany
All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form – by photoprinting, microfilm, or any other means – nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law.
Composition Manuela Treindl, Fürth
Cover Design Adam design, Weinheim
Print ISBN: 978-3-527-33357-8
Foreword by Steven R. Cope
ExxonMobil Refinery Manager, Baytown, Texas
Petroleum refining is one of the most important, exciting and challenging industries on the face of the earth. It has been in existence for about 100 years and during that time, it has evolved and advanced to the point where today’s modern refinery is full of complex, cutting-edge technologies. Examples include state-of-the-art catalyst systems, complex reactor designs, sophisticated computer control hardware and software, and advanced safety and environmental controls.
A typical medium-size refinery has hundreds of pumps, heat exchangers and drums; dozens of furnaces, compressors, and high temperature/high pressure reactors; and thousands of control loops and associated advanced computer control technologies. This same typical refinery has dozens of different crudes and other feedstocks to choose from and dozens of products to maximize or minimize based on consumer demands and global market-place economics. In addition to daily decisions about feedstocks and products, there are also hundreds of decisions to be made each day about operating temperatures, pressures, unit feed rates, catalyst addition rates, cycle times, distillation cut points, product specifications, inventory levels, etc.
In this very competitive global industry, it is critical to minimize overall operating costs while achieving the maximum possible “upgrade” for each hydrocarbon molecule (called “molecule management”). This process requires complex computer modeling to help select feedstocks and product slates and troubleshoot and optimize the performance of individual refinery processes (e.g. distillation units, fluidized catalytic cracking units). And eventually, all of these individual parts have to be pulled together to feed a linear program (LP) model capable of optimizing the overall refinery. This complex modeling is the subject of this book by Ai-Fu Chang, Kiran Pashikanti and Y. A. Liu.
Based on my review, I believe this book provides a solid introduction to integrated refinery process modeling and optimization, using the tools and techniques currently employed in modern refineries. This book and associated coursework would be a highly desirable investment by any engineering student considering a career in petroleum refining.
Foreword by Lawrence B. Evans
Professor Emeritus of Chemical Engineering
Massachusetts Institute of Technology
Member, National Academy of Engineering
Past President, American Institute of Chemical Engineers
Petroleum refining is a huge industry. Every day the industry worldwide produces more than $ 8 billion of refined products. Small improvements in the design and operation of a refinery can deliver large economic value. Crude petroleum is a natural material containing thousands of chemical compounds. The refinery converts the crude into a wide range of products from transportation fuels and petrochemical feedstocks to asphalt and coke. All of these products must meet demanding specifications while the refinery stays within tight environmental constraints.
Computer models are used routinely today to model petroleum refining processes. Engineers use them to design new refineries, to improve the operation of existing refineries, to make decisions on purchasing crude, and to optimize the planning of production. The ability to accurately model each step in the refining process is the key to optimizing the performance of the integrated refinery. Modeling a refinery is challenging because crude petroleum consists of thousands of chemical compounds. The refinery takes the large molecules in crude oil and cracks them into the smaller molecules of transportation fuels. It must also carry out chemical reactions to tailor the composition of products to meet specifications. These reactions take place through a complex set of reaction pathways.
For most of my career, I have worked on the development of computer models of chemical processes. Today very good commercial software systems exist that enable engineers to build and use sophisticated models for refinery simulation and optimization. But these tools are mainly used by experts. This book by Professor Liu and his colleagues represents a major advance in enabling engineers who are not experts to develop and use state-of-the-art computer models for the simulation and optimization of integrated refinery reaction and fractionation processes.
The book is very well organized and systematic. It starts in the first chapter by showing how to represent the thermodynamic and physical properties of crude petroleum and the complex materials that comprise the intermediate streams in a refinery. The next two chapters cover the major separation units in a refinery: the atmospheric distillation unit (ADU) and the vacuum distillation unit (VDU). The final three chapters cover the most important chemical conversion units together with their product fractionation systems. These include the fluid catalytic cracking (FCC) process, the continuous catalyst regeneration (CCR) reforming process, and the hydroprocessing units. Each chapter follows the same pattern starting first with a description of the unit, methods to organize and use the pertinent data from the refinery, and then the workflows to construct a rigorous model using existing commercial software. Finally, the chapter concludes with strategies to tune the models to match performance followed by case study examples, and the discussion of other applications of the models such as for refinery production planning. The book uses Aspen HYSYS for modeling, but most of the concepts are also applicable to other systems. The supporting materials available from the publisher’s website provide relevant spreadsheets and simulation files for all the models and examples presented in the book.
One of the strengths of the book is that it doesn’t stop with theory, or even case study examples and hands-on workshops. It covers very practical problems: how to work with real data, how to construct the right level of detail for the problem and the data available, and how to tune the model to actual plant data. Individuals who want to contribute to the development of refinery process modeling or explore new directions will find the extensive review of existing work valuable. This book will also be valuable to industrial practitioners and to academic chemical engineers by exposing them to refinery process modeling and optimization and enabling them to solve realistic problems. The book takes this work from a technology used mostly by experts to a tool that refinery engineers can use in their everyday work.
Preface
Overview
Petroleum refining continues to be a major contributor in the production of transportation fuels and chemicals. Current economic, regulatory and environ-mental concerns place significant pressure on refiners to optimize the refining process. New product demands have encouraged refiners to explore alternative processing units and feedstocks. Consequently, refiners have invested in many new technologies to upgrade and optimize the refining process.
Despite these changes, refiners still face the same issues as before: selecting the crude feedstock on the basis of feasibility and profitability, finding the optimal process conditions for the given feedstock (while meeting refinery constraints), and understanding how changes in a given unit cascade upstream and downstream to other units in the refinery. In the past, refiners have traditionally relied on experienced process engineers and guesswork to tackle these issues. This approach is not only unreliable, but the growing tide of retiring industry professionals and the prohibitive costs of test runs at the refinery make it quite infeasible. Hence, detailed modeling and optimization of refinery processes becomes increasingly critical and beneficial.
Modeling commercial-scale refinery reaction processes can be quite difficult for the novice model developer. Refinery reaction processes, such as fluid catalytic cracking (FCC), catalytic reforming and hydroprocessing (including hydrotreating and hydrocracking), involve the complex interplay of thermodynamic, kinetic and transport phenomena. In the literature, many models are available that simplify the operation of these units into standard reaction units that are familiar to undergraduate students. While these models can be useful for a given experimental trial of plant operation, it is difficult to generalize these simple models for modern large-scale processes. In addition, these simple models do not account for complex process phenomena and often cannot be integrated into the overall workflow (since they may be customized solutions using FORTRAN, etc.). Consequently, when the person responsible for the development of model is somehow inaccessible, the model falls by the wayside and the gained knowledge is lost. Hence, the use of familiar and standard commercial software tools provides the refinery a path to reap the benefits of rigorous modeling and optimization, and to retain experience developed during the same process.
The primary goal of this text is to present a rational methodology for the integrated modeling and optimization of key reaction and fractionation processes in the modern refinery. We consider catalytic reaction processes, such as fluid catalytic cracking (FCC), catalytic reforming and hydroprocessing, together with upstream fractionation units, such as atmospheric distillation unit (ADU) and vacuum distillation unit (VDU), as well as downstream fractionation units following the catalytic reaction processes. A rational methodology for modeling and optimization must balance the demands of detailed kinetic models with the availability of plant data. It is unproductive to develop and use kinetic models that we cannot support by using available plant data for the purposes of refinery modeling and optimization.
A secondary goal of this text is to serve as a guide for developing models for units whose details vary from those presented in this work. Using commercial software tools, in lieu of customized software, is very beneficial to engineers attempting to replicate the same work. Although we have used Aspen HYSYS from Aspen Technology, Inc. extensively in this work, much of the workflow described is readily applicable to other process simulation software or custom software. This guide is very important to ensure that models are used continually throughout the refining lifecycle and can be integrated into the overall workflow of the refinery.
This text accomplishes these two goals through the following systematic approach for key refining reaction and fractionation processes:
Thorough process descriptions that highlight key operating phenomena required in models
Methods to organize the vast amount of data available in refinery for modeling purposes
Schemes to convert collected data into a format useful for models using rigorous kinetic and thermodynamic schemes
Workflows to build rigorous rating and optimization models using commercial software
Strategies to calibrate rigorous models to reflect plant performance (No model is perfect!)
Methodologies to build downstream fractionation units to expand the scope of models towards integrated refinery models
Case studies that encompass real-life optimization scenarios in the refinery
Applications that broaden model scope beyond engineering purposes (i.e. refinery production planning)
Hands-on step-by-step workshops to help novice users build and apply complex models using commercial software for process rating and optimization
Spreadsheet tools to simplify model development
To our knowledge, our text, Refinery Engineering: Integrated Process Modeling and Optimization is the first book to present the systematic approach shown above for integrating modeling and optimization into the general refinery workflow. There have been several recent books published by a number of authors.
Refinery Process Modeling (Kaes Enterprises, 2000) by Gerald L. Kaes develops several key workflows and industrial modeling guides for various fractionation units throughout the refinery. However, Kaes does not include any guides for modeling refinery reactors rigorously and uses only black-box reactors for important refinery processes. Our text addresses this oversight by tackling bothreaction and fractionation units in an integrative framework with step-by-step guides. Another related work is Fundamentals of Petroleum Refining (Elsevier, 2009) by Mohamed Fahim, Taher Al-Sahhaf and Amal Elkilani. Fahim and his co-authors give a broad overview of a wide range of refinery processes; however, they do not address the model development in any significant detail that is readily applicable by the industrial practitioners. Further, their models often rely on simple and inaccurate correlation-based yield models to represent complex kinetic phenomena. They provide some guides to using commercial software for refinery modeling, but these guides are not useful in an industrial context. In contrast, our text presents industrially relevant hands-on, step-by-step guides and case studies.
Most recently, the text Modeling and Simulation of Catalytic Reactors for Petroleum Refining (Wiley, 2011) by Jorge Ancheyta addresses many similar topics as our text. Ancheyta gives a detailed review of the existing modeling literature on refinery reaction processes in conjunction with modeling results and a few case studies. Such a review monograph is useful for researchers working towards building new models and approaches for refinery reaction process modeling in general. In addition, Ancheyta presents complex equations and sophisticated models that require special modeling expertise to deploy successfully in the refinery. This approach is not well-suited for a novice model developer or plant engineer using commercial software tools. Practical models that we can use in the refinery must address thermodynamics and physical properties for building significant reaction and fractionation models. In addition, these models must also predict fuel product properties and are applicable to production planning. Our text addresses these practical concerns of model users by focusing on the commercial software that is easy to use, deploy and integrate into the existing refinery workflows. In addition, we present hands-on workshops that will help justify the use of these models on a regular basis for the rating and optimization of integrated refinery reaction and fractionation systems from plant data.
Scope of Textbook
The purpose of this text to guide senior-level undergraduates, graduate students, and industrial practitioners how to quantitatively model key refinery reaction and fractionation processes. In addition, this text contains advanced modeling topics (such as kinetic network calibration) that will prove useful to researchers and practitioners alike. After following the procedures in this text, the reader will be able to: (1) identify key data required for building reaction and fractionation models with commercial software; (2) filter extensive data available at the refinery and use plant data to begin calibrating available models; (3) extend model to include key fractionation sub-models; (4) provide a sound and informed basis to understand and exploit plant phenomena for process optimization to improve yield, consistency and performance of a given unit; and (5) apply models in an overall refinery context through refinery production planning based on linear programming (LP).
We present the topics in a logical progression from basic refinery thermo-dynamics and physical property predictions to detailed guides for modeling complex reaction and fractionation units. Chapter 1 introduces the reader to the basics of dealing with the thermodynamics and physical property predictions of hydrocarbon components in the context of process modeling. Chapters 2 and 3 use the key concepts of fractionation lumps and physical properties to develop detailed models and workflows for atmospheric (ADU) and vacuum (VDU) distillation units. Chapters 4, 5 and 6 are largely self-contained and the reader can read each of these chapters independently of other chapters. These chapters discuss the modeling and optimization of FCC, catalytic reforming and hydroprocessing units. In general, we discuss each unit in the following order:
Process description
Modeling and literature review
Key modeling details
Model calibration
Model validation with industrial data collected by the authors
Model applications to process optimization through industrially relevant case studies
Model application to refinery production planning
Hands-on workshops and step-by-step guides for building and applying models using commercial software
In addition, we provide significant supporting materials alongside the text. The reader may download the supporting materials from the publisher’s website for textbooks: http://www.wiley-vch.de/textbooks/. These materials include relevant spreadsheets, guides and sample simulation files for all models developed in the workshops throughout this text.
We hope that this text allows both academia and industrial practitioners to understand, model and optimize complex refinery reaction and fractionation systems. The goal of all modeling and optimization exercises presented is to improve yield, consistency, profitability and performance of a given unit and the refinery as a whole.
Software Selection and Copyright Notice
Aspen HYSYS and Aspen HYSYS Petroleum Refining are available from Aspen Technology, Burlington, MA (http://www.aspentech.com/).
Microsoft Excel and Visual Basic for Applications (VBA) for available as part of Microsoft’s Office software package (http://office.microsoft.com/en-us/default.aspx).
Screen images of input information and output results from Aspen HYSYS® and Aspen HYSYS Petroleum Refining are printed with permission by Aspen Technology, Inc. AspenTech®, aspenONE®, Aspen HYSYS®, Aspen HYSYS Petroleum Refining, and the Aspen leaf logo are trademarks of Aspen Technology, Inc. All rights reserved.
Acknowledgements
It is a pleasure to thank a number of very special persons and organizations that contributed to the preparation of this book.
The idea for this book originated from the doctoral work of the junior authors, Ai-Fu Chang and Kiran Pashikanti. The junior authors would like to thank the members of their advisory committee at Virginia Tech, in particular: Professor Y. A. Liu, who developed the original idea of the book and was the major advisor, and Professors Luke Achenie, Richey M. Davis and Preston Durrill.
We would like to express our sincere appreciation to the engineering product management and refinery process modeling experts at Aspen Technology, in particular Stephen Dziuk, Hiren Shethna, Dhaval Dave, Darin Campbell, Maurice Jett, John Adams, Glenn Dissinger and Vikas Dhole for teaching us the principles and practice of refinery process modeling. We thank Chau-Chyun Chen for his continued guidance in our learning of process modeling. We also want to thank Desmond Jacas and Blanca Yanulis, Global University Program, for providing us software tools.
We would like to thank the China Petroleum and Chemical Corporation (SINOPEC) and Formosa Petrochemical Corporation (FPCC) for challenging us to enter the field of refinery process modeling in 2007.
We thank Alliant Techsystems, Aspen Technology, China Petroleum and Chemical Corporation (SINOPEC), Milliken Chemical, Novozymes Biological, and Mid-Atlantic Technology, Research and Innovation Center for supporting our educational programs in computer-aided design and process system engineering at Virginia Tech. We are very grateful to Mr. Cao Xianghong, for his strong support of this work during his tenure as Senior Vice President and Chief Technology Officer of SINOPEC.
We thank the following academic and industrial leaders who kindly took time to write the FOREWORD for our text: Mr. Steven R. Cope, Manager of the Baytown Refinery, ExxonMobil Corporation, and Professor Lawrence B. Evans of Massachusetts Institute of Technology and Founder of Aspen Technology, Inc.
Ai-Fu Chang would like to thank his wife, I-Chun Lin, for her patience in enduring years of suffering as the girlfriend, fiancée, and now wife of a Ph. D. student, and to his parents and big sister for their unconditional love and encouragement in my life and studies. Kiran Pashikanti would like to thank to his parents for their continuing support throughout his graduate studies. The senior author would like to thank his wife, Hing-Har Liu, for her support through the laborious process of this book writing and revision.
About the Authors
Ai-Fu Chang received his Ph. D. in the Department of Chemical Engineering at Virginia Polytechnic Institute and State University (“Virginia Tech”) in September, 2011. He received his B. S. in chemical engineering from National Taiwan University in 2001. He completed his doctoral dissertation on integrated process modeling and product design of biodiesel manufacturing, and refinery reaction and fractionation systems. The latter was the basis of this textbook. He has worked on several industrial modeling projects, including poly (acrylonitrile-vinyl acetate), hydrocracking, and biodiesel. These projects were collaborative efforts between Virginia Tech, Aspen Technology, and industrial manufacturers. He is currently employed by Chevron Phillips Chemical Company.
Kiran Pashikanti was a Ph. D. student in the Department of Chemical Engineering at Virginia Tech. He received his B. S. in chemical engineering from Virginia Commonwealth University in 2005, and his Ph. D. in chemical engineering from Virginia Tech in September, 2011. He has worked on several industrial modeling projects on integrated modeling of refinery reaction and fractionation systems, and of carbon-dioxide capture processes. This textbook grows out of his doctoral dissertation on the predictive modeling of fluid catalytic cracking and catalytic reforming processes. He is currently employed by Chevron Phillips Chemical Company.
Y. A. Liu, the Frank C. Vilbrandt Endowed Professor of Chemical Engineering at Virginia Tech, received his B. S. (1967), M. S. (1970), and Ph. D. (1974) degrees from National Taiwan University, Tufts University and Princeton University, respectively.
Professor Liu taught at Auburn University from 1974 to 1981, where he received the Outstanding Engineering Faculty Award four times, and his last position was Alumni Associate Professor endowed by the Auburn Alumni Association. He joined Virginia Tech as a Professor of Chemical Engineering in 1982. In 1983, he was appointed the Vilbrandt Professor. He has published numerous papers and eight books, including four pioneering chemical engineering textbooks on artificial intelligence in chemical engineering (with Thomas E. Quantrille) and on neural networks in bioprocessing and chemical engineering (with D. Richard Baughman) in 1991 and 1995, respectively, published by Academic Press, San Diego, California, on industrial water reuse and wastewater minimization (with James G. Mann) in 1999, published by McGraw-Hill, New York, and on step-growth polymerization process modeling and product design (with Kevin Seavey) in 2008, published by John Wiley and Sons, New York.
Professor Liu’s contributions to chemical engineering teaching and research have been recognized by university, national and international awards. He is a Fellow of the American Institute of Chemical Engineers, a member of Virginia Tech’s Academy of Teaching Excellence, and a recipient of the 1996 AspenTech International Award for University Teaching Excellence in computer-aided design. He has received three awards from the American Society of Engineering Education (ASEE): the Fred Merryfield Design Award (1993) for creativity and excellence in teaching and research of engineering design; the George Westinghouse Award (1990), ASEE’s highest honor for an engineering educator under age 45 for outstanding achievements in both teaching and scholarship; and the Western Electric Award (1984) for excellence in instruction of engineering students. In 1986, he received the National Catalyst Award for excellence in chemical education from the Chemical Manufacturers Association. He received the Distinguished Chemical Engineering Alumni Award in 1990, and the Outstanding Career Achievement Award in 2010, both from Tufts University.
Over the past 25 years, Professor Liu devoted his school breaks helping petrochemical industries in developing countries and chemical industries in Virginia with technology development and engineering training. He has taught intensive training courses on computer-aided design, process system engineering, energy and water savings, and refinery and polymerization process modeling to over 6,000 practicing engineers in China, Taiwan and United States. For his contributions to teaching, research and industrial outreach, he received the Virginia Outstanding Faculty Award from Governor Jim Gilmore in 2000. He also received the National Friendship Award from China’s Premier Zhu Ronjie in 2000.
1
Characterization, Physical and Thermodynamic Properties of Oil Fractions
This chapter introduces the common methods for characterizing crude oils and petroleum fractions (i.e., oil fractions), and for estimating their thermophysical properties. We begin by defining the essential bulk and fractional properties for oil fractions, and by explaining the various types of distillation curves and their interconversion (Section 1.1). Next, we discuss the generation of pseudocomponents of oil fractions based on boiling-point ranges, and the estimation of density and molecular weight distributions of the resulting pseudocomponents (Section 1.2). Sections 1.3 to 1.6 present four hands-on workshops using Excel spreadsheets and Aspen HYSYS Petroleum Refining for the interconversion of distillation curve data, the extrapolation of an incomplete distillation-curve data, the calculation of the mean-average boiling point of a given oil fraction, and specifying the oil fraction in Aspen HYSYS Petroleum Refining. Section 1.7 intro duces the essential thermophysical properties for developing refinery reaction and fractionation process models. Section 1.8 presents the useful methods for estimating the thermophysical properties (e.g., molecular weight, liquid density, critical properties, ideal gas heat capacity, and heat of vaporization) of pseudocomponents of oil fractions. Section 1.9 describes the important thermodynamic models for refinery reaction and fractionation processes. Section 1.10 discusses the estimation methods for other physical properties such as flash point, freeze point and PNA (paraffin, naphthalene and aromatic) content of a refinery feed. Finally, Section 1.11 summarizes the conclusions of this chapter.
1.1 Crude Assay
Crude oils and petroleum fractions are the most important feedstocks for refining processes. To properly simulate the refining processes, we must have good understanding of the compositional information and thermophysical properties of crude oils and petroleum fractions. However, the complexity of molecular composition of crude oils and petroleum fractions makes it hardly possible to identity individual molecules. Instead, modern refiners use assay to characterize crude oils and petroleum fractions.
Table 1.1 A typical crude assay.
A typical crude assay includes two types of information for an oil sample: (1) bulk properties; and (2) fractional properties. Table 1.1 gives examples of both properties of a crude assay. For design and modeling purposes, it is always the best practice to have process data obtained in the same period as assay data, since the properties and composition of crude change over time as it is produced from a given well. Kaes [1] suggests that assay data should not be two years older than the process data used to build process simulation. We explain both bulk and fractional properties in the following subsections.
1.1.1 Bulk Properties
The sulfur content is expressed as a percentage of sulfur by weight, and varies from less than 0.1% to greater than 5%. Crude oils with less than 1 wt.% sulfur are called low-sulfur or sweet crude, and those with more than 1 wt.% sulfur are called high-sulfur or sour crude. Sulfur-containing constituents of the crude oil include simple mercaptans (also known as thiols), sulfides, and polycyclic sulfides. Mercaptan sulfur is simply an alkyl chain (R–) with –SH group attached to it at the end. The simplest form of R–SH is methyl mercaptan, CH3SH.
The pour point is a measure of how easy or difficult to pump the crude oil, especially in cold weather. Specifically, the pour point is the lowest temperature at which a crude oil will flow or pour when it is chilled without disturbance at a controlled rate. The pour point of the whole crude or oil fractions boiling above 232 °C (450 °F) is determined by the standard test ASTM D97.
The flash point of a liquid hydrocarbon or an oil fraction indicates its fire and explosion potential, and it is the lowest temperature at which sufficient vapor is produced above the liquid to form a mixture with air that a spontaneous ignition can occur if a spark is present. One of the standard ASTM test methods for the flash point is D3278.
The freeze point is the temperature at which the hydrocarbon liquid solidifies at atmospheric pressure. It’s an important property for kerosene and jet fuels, because of the very low temperatures encountered at high altitudes in jet planes. One of the standard test methods for the freeze point is ASTM D4790.
The smoke point refers to the height of a smokeless flame of fuel in millimeters beyond which smoking takes places. It reflects the burning quality of kerosene and jet fuels, and is determined by the standard testing method ASTM D1322.
The aniline point represents the minimum temperature for complete miscibility of equal volumes of aniline and petroleum oil. It’s an important property of diesel fuels, and is measured by ASTM D611.
The cloud point refers to the temperature at which solidifiable components (waxes) present in the oil sample begin to crystallize or separate from solution under a method of prescribed chilling. It’s an important specification of middle distillate fuels, as determined by ASTM D2500.
The Conradson carbon residue (CCR) results from ASTM test D189. It measures the coke-forming tendencies of oil. It is determined by destructive distillation of a sample to elemental carbon (coke residue), in the absence of air, expressed as the weight percentage of the original sample. A related measure of the carbon residue is called Ramsbottom carbon residue. A crude oil with a high CCR has a low value as a refinery feedstock.
The acid number results from ASTM test method D3339-11 that determines the organic acidity of a refinery stream.
The refractive index represents the ratio of the velocity of light in a vacuum to that in the oil. It is determined by ASTM D1218.
The gross heat of combustion or high heating value (HHV) is the amount of heat produced by the complete combustion of a unit quantity of fuel. We obtain the gross heat of combustion by cooling down all products of the combustion to the temperature before the combustion, and by condensing all the water vapor formed during combustion.
The net heat of combustion or lower heating value (LHV) is obtained by subtracting the latent heat of vaporization of the water vapor formed by the combustion from the gross heat of combustion or higher heating value.
The true boiling point (TBP) distillation [1] of a crude oil or petroleum fractions results from using the U. S. Bureau of Mines Hempel method and the ASTM D-285 test procedure. Neither of these methods specifies the number of theoretical stages or the molar reflux ratio used in the distillation. Consequently, there is a trend toward applying a 15: 5 distillation according to ASTM D2892, instead of the TBP. The 15: 5 distillation uses 15 theoretical stages and a molar reflux ratio of 5.
A key result from a distillation test is the boiling point curve, that is, the boiling point of the oil fraction versus the fraction of oil vaporized. The initial boiling point (IBP) is defined as the temperature at which the first drop of liquid leaves the condenser tube of the distillation apparatus. The final boiling point or the end point (EP) is the highest temperature recorded in the test.
Additionally, oil fractions tend to decompose or crack at a temperature of approximately 650 °F (344 °C) at one atmosphere. Thus, the pressure of TBP distillation is gradually reduced to as low as 40 mmHg, as this temperature is approached to avoid cracking the sample and distorting measurements of true components in the oil.
The TBP distillation typically takes much time and labor. In practice, we carry out the distillation test of oil fractions using other less costly ASTM methods and convert the resulting boiling point curve to TBP curve using correlations, as given in the API Technical Data Book-Petroleum Refining [2]. We have implemented these correlations in an Excel spreadsheet of the Interconversion of boiling point curves from typical ASTM distillation methods in a hands-on workshop in Section 1.3.
The ASTM D86 distillation of an oil fraction takes place at laboratory room temperature and pressure. Note that the D86 distillation will end below an approximate temperature of 650 °F (344 °C), at which petroleum oils begin to crack at one atmospheric pressure.
The ASTM D1160 distillation of an oil fraction is applicable to high-boiling oil samples (e.g. heavy heating oil, cracker gas oil feed, residual oil, etc.) for which there is significant cracking at atmospheric pressures. The sample is distilled at a reduced pressure, typically at 10 mmHg, to inhibit cracking. In fact, at 10 mmHg, we can distill an oil fraction up to temperatures of 950 to 1000 °F (510 to 538 °C), as reported on a 760-mmHg basis. The reduced pressure used for D1160 distillation produces a separation of components that is more ideal than that for D86 distillation.
The ASTM D2887 distillation of an oil fraction is a popular chromatographic procedure to “simulate” or predict the boiling point curve of an oil fraction. We determine the boiling point distribution by injecting the oil sample into a gas chromatograph that separates the hydrocarbons in a boiling-point order. We then relate the retention time inside the chromatograph to the boiling point through a calibration curve.
1.1.2 Fractional Properties
Bulk properties provide a quick understanding of the type of the oil sample such as sweet and sour, light and heavy, etc. However, refineries require fractional properties of the oil sample that reflects the property and composition for specific boiling-point range to properly refine it into different end products such as gasoline, diesel and raw materials for chemical process. Fractional properties usually contains paraffins, naphthenes and aromatics (PNA) contents, sulfur content, nitrogen content for each boiling-point range, octane number for gasoline, freezing point, cetane index and smoke point for kerosene and diesel fuels.
The octane number is a measure of the knocking characteristics of a fuel in a laboratory gasoline engine according to ASTM D2700 [1]. We determine the octane number of a fuel by measuring its knocking value compared to the knocking of a mixture of n-heptane and isooctane or 2-2-4-trimethylpentane (224TMP). By definition, we assign an octane number of 0 to pure heptane and of 100 to 224TMP. Therefore, a mixture of 30% heptanes and 70% isooctane has an octane number of 70.
There are two specific octane numbers in use. The motor octane number (MON) reflects the engine performance at highway conditions with high speeds (900 rpm), while the research octane number (RON) corresponds to the low-speed city driving (600 rpm). RON is typically higher than MON because of engine test efficiencies. The posted octane number is an average of MON and RON.
The cetane number measures the ease for self-ignition of a diesel fuel sample and is essentially an opposite of the octane number. It represents the percentage of pure cetane (n-hexadecane) in a blend of cetane and alpha methyl-naphthalene that matches the ignition quality of a diesel fuel sample. This quality is important for middle distillate fuels.
The cetane index is a number calculated from the average boiling point and gravity of a petroleum fraction in the diesel fuel boiling range, which estimates the cetane number of the fraction according to ASTM D976 (see, for example, http://www.epa.gov/nvfel/testproc/121.pdf).
1.1.3 Interconversion of Distillation Curves
While building a refining process simulation, distillation curve of the oil sample is the most confusing information among assay data since there are different methods used to obtain volatility characteristics of an oil sample. The most widely used tests of distillation curve are ASTM D86, ASTM D1160 (atmospheric distillation), ASTM D1160 (vaccum distillation), ASTM D2887 (chromatographic simulation) and true boiling point (TBP). API Technical Databook [35] presents the characteristics of each test and gives the correlations to perform interconversion among these ASTM distillation types. Most commercial process simulators include the capability to convert one type of distillation curve to the other. We develop a MS Excel spreadsheet which automates the API conversion between any two of the ASTM distillation types (see Figure 1.1). Section 1.3 presents a hands-on workshop for this interconversion of distillation-curve data.
Figure 1.1 Conversion spreadsheet for distillation curves.
1.2 Pseudocomponent Generation Based on Boiling-Point Ranges
To simulate refining processes, the first task is to construct a pseudocomponent scheme to characterize the feedstock. Data requirement and definition of the pseudocomponents depend on the type of the refining process to be modeled. There are different concerns to be addressed when building pseudocomponents for fractionation and reaction units. The pseudocomponents for fractionation units have to accurately characterize volatilities of the hydrocarbons in the feedstock in order to calculate vapor-liquid-equilibrium in distillation columns. Therefore, refiners use pseudocomponents based on boiling-point ranges to represent the feedstock and model fractionation units. For modeling of reaction units, refiners partition the hydrocarbons into multiple lumps (or model compounds) based on molecular structure or/and boiling-point ranges, and assume the hydrocarbons of each lump to have an identical reactivity in order to develop the reaction kinetics for reaction units. This section deals with pseudocomponent generation based on boiling-point ranges for fractionation units. Chapters 4 to 6 will present the pseudocomponent schemes for the three major reaction units in modern refinery – fluid catalytic cracking (FCC), catalytic reformer and catalytic hydrocracker.
Most commercial process simulators include the capability to generate pseudo components based on boiling-point ranges representing the oil fractions. Workshop 1.4 will demonstrate how to use Aspen HYSYS to generate pseudocomponents based on boiling-point ranges for a given oil fraction with required analysis data. Conventionally, there are four steps to develop pseudocomponents based on boiling-point ranges to represent petroleum fraction:
Figure 1.2 Relationship between pseudocomponent properties and the TBP curve (redraw from [1]).
Table 1.2 Typical boiling-point ranges for pseudocomponents in commercial process simulators.
Boiling-point Range
Suggested Number of Pseudocomponents
IBP to 800 °F (425 °C)
30
800 °F to 1200 °F (650 °C)
10
1200 °F to 1650 °F (900 °C)
8
(1.1)
(1.2)
(1.3)
(1.4)
(1.5)
(1.6)
(1.7)
Figure 1.3 Iteration spreadsheet for MeABP calculation.
Section 1.4 presents a hands-on workshop for applying our spreadsheet to extrapolate an incomplete distillation curve. We note that the density distribution along with boiling point should be used (in step 3) whenever it is available because the assumption of constant Watson K factor always fails in low and high boiling-point ranges. Figure 1.5 compares the pseudocomponents generated from constant Watson K factor and from density distribution. The pseudocomponents generated from constant Watson K factor shows significant deviations from assay data on estimating the densities of pseudocomponents, particularly in both light and heavy ends of the distillation curve. On the other hand, using density distribution is able to provide a good estimation of the densities of pseudocomponents. Estimating the densities of pseudocomponents is the most important part when developing pseudocomponents because density is required for most of the physical property estimations.
Figure 1.4 Spreadsheet for extrapolating distillation curve.
Figure 1.5 Comparison of the pseudocomponents generated from constant Watson K factor and density distribution (data obtained from [1]).
1.3 Workshop 1.1 – Interconvert Distillation Curves
There are two situations we may encounter when the distillation curve available is not a TBP curve and needs to be converted – (1) It is another ASTM type; and (2) It is ASTM D1160 at vacuum pressure. The spreadsheet we have developed is able to solve these two cases. The following steps demonstrate how to convert an ASTM D1160 (at 10 mmHg) curve into a TBP curve.
Step 1: Open ASTMConvert.xls.
Figure 1.6 ASTMConvert.xls.
Step 2: Copy and paste the ASTM D1160 curve into the sheet for interconversion among different testing pressures of ASTM D1160.
Figure 1.7 Input cells of ASTM D1160 interconversion in ASTMConvert.xls.
Step 3: Input the testing pressure which is 10 mmHg in this case.
Figure 1.8 Input pressure for ASTM D1160 interconversion.
Step 4: The blue cells will show the converted results which correspond to ASTM D1160 at 1 atmosphere.
Figure 1.9 The results of ASTM D1160 interconversion.
Step 5: Copy the values of ASTM D1160 (at 1 atm) to the sheets for converting ASTM D1160 at 1 atm into TBP.
Figure 1.10 Input cells for other ASTM interconversion in ASTMConvert.xls.
Step 6: The blue cells reveals the converted TBP curve.
Figure 1.11 Result cells for other ASTM interconversion in ASTMConvert.xls.
1.4 Workshop 1.2 – Extrapolate an Incomplete Distillation Curve
Step 1: Open Beta.xls. Cells B2 to B5 show the adjustable parameters in beta distribution function, the cells A8 to B16 require the input of the distillation curve, cells H8 to K16 and the graph indicate the fitted results.
Figure 1.12 Beta.xls.
Step 2: Input the incomplete distillation curve into cells A8 to B16. The user is allowed to add/remove according to the number of points in the distillation curve.
Figure 1.13 Input cells in Beta.xls.
Step 3: Click “solve” to run the fitting program.
Figure 1.14 Activation button in Beta.xls.
Step 4: The cells B2 to B5 show the fitted parameters. Cells H8 to K16 and the graph represent the extrapolated distillation curve.
Figure 1.15 Fitted results in Beta.xls.
1.5 Workshop 1.3 – Calculate MeABP of a Given Assay
Step 1: Open MeABP Iteration.xls.
Figure 1.16 MeABP.xls.
Step 2: Select the type of the oil fraction. We choose naphtha in this case.
Figure 1.17 Select oil type.
Step 3: Input TBP curve and specific gravity in highlighted cells.
Figure 1.18 Input distillation curve and specific gravity.
Step 4: Go to Tool/Goal Seek.
Figure 1.19 Activate “goal seek” in Beta. slx.
Step 5: Assign cell 19 to “By changing cell” and cell B30 to “Set cell” and input “0” in “To value”. And then, click “OK”.
Figure 1.20 Assign tuning and objective cells.
Step 6: Row 28 reveals the calculated MeABP for the given oil fraction.
Figure 1.21 Iterative MeABP in MeABP.xls.
1.6 Workshop 1.4 – Duplicate the Oil Fraction in Aspen HYSYS Petroleum Refining
Step 1: Start a new case in Aspen HYSYS Petroleum Refining.
Figure 1.22 Start a new case in Aspen HYSYS Petroleum Refining.
Step 2: Click “add” to add a new component list.
Figure 1.23 Add a new component list.
Step 3: Click “view” to edit the component list. Add light components which are shown in assay data.
Figure 1.24 Add light components.
Step 4: Click “add” in “fluid pkgs” tab to add the thermodynamic model.
Figure 1.25 Enter the list of thermodynamics models.
Step 6: Select the Peng-Robinson method.
Figure 1.26 Select a thermodynamics model.
Step 7: Click “enter oil environment” in “oil manager” tab.
Figure 1.27 Enter the “oil environment”.
Step 8: Click “add” to add a new assay and click “view” to edit the assay data.
Figure 1.28 Add and edit assay data.
Step 9: In this case, we have TBP curve, bulk density and light end composition. Therefore, we use these three properties to build the assay in Aspen HYSYS Petroleum Refining. Users are allowed to input molecular weight curve, density curve and viscosity curve if available.
Figure 1.29 Select the data to be used to define an assay.
Step 10: Check “distillation” and click “edit assay” to input the distillation curve.
Figure 1.30 Enter the distillation curve.
Step 11: Check “bulk props” to input the bulk density and other bulk properties if available.
Figure 1.31 Enter the bulk density.
Step 12: Check “light ends” to input the light-end composition.
Figure 1.32 Enter the composition of light components.
Step 13: Click “calculate” to enable the Aspen HYSYS Petroleum Refining’s calculation for working curves which are used to generate pseudocomponents.
Figure 1.33 Enable the pseudocomponent generation.
Step 14: Go to “cut/blend” tab and click “add” to add a new cut. Then, click “view” to edit the cut.
Figure 1.34 Add cut/blend.
Step 15: Select “assay-1” and click “add” to use the assay we input to generate the corresponding pseudocomponents.
Figure 1.35 Select the assay used to be cut or blended.
Step 16: Go to “table” tab to check the generated pseudocomponents.
Figure 1.36 The pseudocomponents used to represent the cut or blend.
Step 17: Close the window in previous step. And then, go to “install oil” tab, check “install” box and enter stream name (it is oil in this case).
Figure 1.37 Install the cut/blend into simulation.
Step 18: Click “return to basis environment”.
Figure 1.38 Return to the basis environment.
Step 19: Click “return to simulation environment”.
Figure 1.39 Return to the simulation environment.
Step 20: The oil fraction is duplicated in Aspen HYSYS Petroleum Refining.
Figure 1.40 The stream in the simulation environment used to represent the oil fraction.
1.7 Property Requirements for Refinery Process Models
We classify the processes in modern refinery into two categories: separation units and reaction units. To develop a process model for any unit, we need to check mass and energy balances of the flowsheet and perform calculations to describe the performance of the target unit. Therefore, the essential properties (physical and chemical) used to simulate these processes depend on the target unit, the chosen pseudocomponent scheme and the selected kinetic model for reaction unit. Chapters 4 through 6 will represent the relevant issues for the three major reaction units in a modern refinery – FCC, catalytic reformer and hydrocracker. While this chapter focuses primarily on the thermophysical properties required for modeling fractionation processes, the general framework for developing these properties for different kinds of pseudocomponents (i.e. those generated by kinetic lumping networks) is the same.
The previous sections in this chapter address the creation of pseudocomponents by cutting an assay curve into a set of discrete components based on boiling-point ranges. We also briefly alluded to physical properties and process thermodynamics selection in the earlier workshops of this chapter. In this section, we consider, in detail, the problem of how to represent these components in the process modeling software. There are two major concerns in this area: physical properties of pseudocomponents and selection of a thermodynamic system that can deal with these hydrocarbon pseudocomponents in the context of refinery modeling. A correct selection of physical properties and process thermodynamics results in a process model that can accurately account for material and energy flows in both vapor and liquid process streams.
1.8 Physical Properties
For any process simulation that involves only vapor-liquid phases, certain key physical and thermodynamic properties must be available for each phase. Table 1.3 lists these properties for all phases. We can typically obtain these properties for pure components (i.e. n-hexane, n-heptane, etc.) from widely available databases such as DIPPR [2]. Commercial process simulation software (including Aspen HYSYS) also provides a large set of physical and thermodynamic properties for a large number of pure components. However, using these databases requires us to identify a component by name and molecular structure first, and use experimentally measured or estimated values from the same databases. Given the complexity of crude feed, it is not possible to completely analyze the crude feed in terms of pure components. Therefore, we must be able to estimate these properties for each pseudocomponent based on certain measured descriptors.
It is important to note the properties given in Table 1.3 are the minimal physical properties required for rigorous accounting of the material and energy flows in the process. As we will discuss in the subsequent sections, process models may require additional properties (especially vapor pressure) depending on the type of thermodynamic models being considered.
Table 1.3 Required properties for each phase.
Phase
Required Properties
Vapor
Ideal Gas Heat Capacity (CP
IG
)
Liquid
Liquid Heat Capacity (CP
L
), Liquid Density (
ρ
L
), Latent Heat of Vaporization (Δ
H
VAP
), Vapor Pressure (
P
VAP
)
Both
Molecular weight (MW)
1.8.1 Estimating Minimal Physical Properties for Pseudocomponents
We show in previous sections that the minimal amount of information to create pseudocomponents is a distillation curve and a specific gravity or density distribution. If only the bulk density is available, we can use the constant Watson K-Factor assumption to estimate the density distribution. If only a partial density distribution is available, we can use the beta function to extrapolate an incomplete distillation curve. Note that it is almost always better to incorporate as much experimentally measured information about the density curve as possible when building the process model. Once the distillation and density curve are available, we can cut the curve into a set of discrete pseudocomponents, each with its own boiling point and density. We will use these two measured properties to estimate a variety of different types of physical properties (i.e. molecular weight, critical temperature, critical pressure, acentric factor, etc.). Using these estimated physical properties, we can derive additional estimates for minimal physical properties required for process simulation. We have also provided a Microsoft Excel spreadsheet in the material that accompanies this text which includes many of the correlations given in this section.
1.8.2 Molecular Weight
The molecular weight is the most basic information for a given pseudocomponent. Molecular weight is a required property to ensure an accurate material balance throughout the process flowsheet. Researchers have studied extensively the trends of molecular weight for a variety of pure hydrocarbons and oil fractions. There are several correlations available to estimate the molecular weight as a function of boiling point, density and viscosity. In general, correlations that only require the boiling point are the least accurate and correlations that require values of boiling point, density and viscosity tend to be the most accurate. Viscosity is used as a parameter in these correlations because it correlates well with molecular type – which can further refine the molecular weight estimate. In most cases, we use correlations that require the boiling point and density of a given component. Two popular correlations are the Lee-Kesler [9, 10] correlation, Equation (1.8), and the Twu [11] correlation, Equations (1.9) to (1.12), respectively.
(1.8)
(1.9)
(1.10)
(1.11)
(1.12)
(1.13)
(1.14)
(1.15)
(1.16)
Riazi [4] lists several other correlations such as Cavett and Goosens for molecular weight, but they generally do not have significant advantage over the Lee-Kesler or Twu correlations. The Lee-Kesler correlation was developed by correlating light oil fractions (< 850 °F or 454 °C) from a variety of sources. As a result, the Lee-Kesler correlation tends to be less accurate for pseudocomponents with high boiling point temperatures. The Twu correlation includes a significant number of data points to account for heavier components. We recommend using the Twu correlation, especially for heavier feed types processed in the crude vacuum towers. The correlation is quite easy to change in most process modeling software. Figure 1.41 shows how to select the molecular weight correlation for a particular blend (shown in earlier workshops) in Aspen HYSYS.
Figure 1.41 Modify molecular weight correlation in Aspen HYSYS.
1.8.3 Critical Properties
Many properties that are required for rigorous accounting of material and energy flows (Table 1.3) in process models are not well defined for pseudocomponents. Fortunately, researchers have found that these required properties correlate well with critical temperature (Tc), critical pressure (Pc) and acentric factor (ω) for different types of hydrocarbons from many sources. Therefore, when we use pseudocomponents of any kind, we must also estimate these critical properties. Just as with molecular weight, there are many critical property estimation methods available in the literature. These correlations differ on the basis of the parameters required and underlying data used to create the correlation. We note that as the components get heavier and boil at higher temperatures, the associated change in critical pressure tends to diminish. Hence, correlations for critical pressure tend to be logarithmic formulas. A modeling consequence is that particularly accurate measures of these critical pressures are not required for good modeling results. In addition, most refinery processes conditions do not approach the critical properties of these pseudocomponents.
Lee-Kesler [9, 10] and Twu [11] have also produced correlations for critical properties. In our work, we have used the Lee-Kesler correlations extensively. Equations (1.17) and (1.18) give the correlations for critical temperature (Tc) and critical pressure (Pc) using the Lee-Kesler correlations. We recommend using these correlations for all boiling-point ranges since the differences that arise from using other correlations are often minor. Figure 1.42 and Figure 1.43 show how we can change the correlation for each blend in Aspen HYSYS.
(1.17)
(1.18)
A related property is the acentric factor. The acentric factor accounts for the size and shape of various kinds of molecules. Simple molecules have an acentric factor close to 0, whereas large or complex hydrocarbon molecules may have values approaching 0.5 to 0.66. The acentric factor is not measured, but defined as an explicit function of the ratio of vapor pressure at the normal boiling point to the measured or estimated critical pressure. We show the definition of the acentric factor in Equation (1.19).
Figure 1.42 Modify Tc correlation in Aspen HYSYS.
Figure 1.43 Modify Pc correlation in Aspen HYSYS.
(1.19)
where represents the reduced vapor pressure, that is, the pseudocomponent vapor pressure divided by its critical pressure, when the reduced temperature, Tr that is, the temperature divided by the critical temperature, is equal to 0.7.
Given the small range of values for the acentric factor, most correlations can provide useful results. The accuracy of the acentric correlation depends largely on the accuracy of the critical temperature and pressure correlations. However, even large relative errors do not result in significant deviation of derived properties such as ideal gas heat capacity, etc. We again choose the Lee-Kesler [9, 10] correlation for the acentric factor. This correlation, given by Equation (1.16), relies on extensive vapor pressure data collected by Lee and Kesler for the critical temperature and pressure correlations. The correlation is technically limited to the reduced boiling point temperature (Tbr) of less than 0.8, but has been successfully used at high Tbr values. Figure 1.44
