Preface

Motivation and Goals

Modern adsorptive and chromatographic technologies have been leaning heavily on cyclic processes, as opposed to traditional steady-state unit operations. A number of such technologies, including temperature swing adsorption (TSA), pressure swing adsorption (PSA), and simulated moving bed (SMB) chromatography, are becoming more common in laboratory-, pilot-, and industrial-scale processes. Engineers are grappling with a wide variety of methods for characterizing the cyclic operations of these technologies, and researchers are still inventing new heuristics and approximations for design and optimization.

Simulated moving bed technologies in particular are beginning to fulfill their earlier promise of creating efficient continuous chromatographic alternatives to older batch separations. They are finding widespread applications in separating and purifying enantiomers, petrochemicals, pharmaceuticals, and biochemicals with higher yield and lower solvent consumption. Because of this, there are a large number of engineers who need actionable methods for the selection, design, simulation, and optimization of these processes.

Simulated moving beds are new enough that there is a dearth of heuristics and shortcut design methods for practicing engineers to use. Indeed, even the ones that do exist are developed for narrow classes of separations and frequently do not map well to the separations of interest. This has led to a significant amount of duplicated work and a reliance on large over-design of SMB systems.

Engineers working on gas-adsorptive separations have similar needs to those who do SMB chromatography but are slightly better off since there are numerous sources of information. There are several well-established textbooks on the theory and design of TSA and PSA units, and resources are available for both practicing engineers and academic researchers. However, for systems as complex as TSA and PSA, being informed of the theory is necessary, but not sufficient, to perform the kind of simulation and optimization required in this competitive industry.

Often, experienced engineers and operators can make good estimates of unit performance. However, the growing tide of retiring professionals and loss of experience throughout the industry makes this task difficult or impossible. An engineer who wants to actually implement the modern theories to build computer models of his adsorptive and chromatographic processes will not find much help in current resources. There are software packages, which can put the finite-element analyses within reach of most engineers, but those without previous computer modeling experience will be hard pressed to use them effectively. This is because there is very little educational material that is devoted to the mapping between the theories of adsorptive and chromatographic separations and their implementation in such tools. Also, these software packages often omit intermediate methods (such as standing wave design) that allow the engineer to incorporate a wider set of physical effects (such as mass transfer resistance) than the idealized design equations without performing a rigorous, but time-consuming, finite-element simulation.

We prepare this text in order to teach undergraduate and graduate students, engineers, and scientists what they need to know to effectively design, simulate, and optimize industrial adsorptive and chromatographic separations. We do this by presenting a unified approach to the development of the ideal and intermediate design equations, while simultaneously teaching the reader to use the rigorous simulation packages, Aspen Adsorption and Aspen Chromatography (Aspen Technology, Inc., Bedford, Massachusetts). We explain the physical phenomena involved in adsorptive and chromatographic processes, and then perform the derivations of the various design equations using the computer algebra system, Mathematica (Wolfram Research, Inc., Campaign, Illinois). In this way, the reader can see the path from problem specification to solution, without getting bogged down in the details required to solve the complex systems of differential equations.

This book has a number of secondary goals, which all serve the main goals. These include the following:

1. 1. Describe new software techniques, which can dramatically speed up simulations, and simplify model development. These include techniques to convert dynamic simulations into a simulation that attempts to directly solve for a cyclic steady state. We also describe techniques for storing results to avoid unnecessarily duplicating calculations.
2. 2. Introduce the modeling of adsorptive separations, such as temperature and pressure swing absorbers (TSA/PSA). Because liquid and gas adsorption systems have extensive mathematical similarities, a comparison between these two types of separations can help those engineers who are already familiar with gas-phase adsorption understand the SMB processes.
3. 3. Introduce some premade tools for performing the shortcut design of SMB systems, as well as instructions for modifying them to apply to systems outside the coverage of this book. This has largely been lacking in the field so far, and while the simplest methods can be easily implemented in a tool such as Excel, the majority of the methods require more advanced coding to solve.
4. 4. Teach practicing engineers how to do parameter regression in adsorptive and chromatographic separations. Regression requires a solid understanding of the underlying physical processes and is frequently required during scale-up. While regression against steady-state models is well understood, the cyclic nature of the SMB and gas adsorption systems introduces some additional nuances that must be taken into account. Moreover, regressing against dynamic models can require careful planning to be successful.

Approaches

To accomplish these goals, this text employs the following methodical approach:

• Thorough descriptions of adsorptive and chromatographic processes that highlight the underlying physical phenomena to be captured by our models.
• Description of how these phenomena manifest themselves in the ultimate models, or which assumptions cause them to be omitted.
• Best practices for using the developed models, including workflows that use multiple software packages.
• Strategies for improving the model performance, including parameter regression as well as customizing model code.
• Case studies of real processes to help the reader understand the details of TSA/PSA and SMB operations.
• Step-by-step, hands-on workshops to ensure that the reader is able to construct models with the complexity necessary to handle an immense variety of adsorptive and chromatographic separations.
• Instructions on how to develop new models in commercial software packages.
• Hands-on practice problems with hints and solutions.

Contents

In Chapter 1, we introduce the pressure and temperature swing strategies, and then learn the detailed design considerations by building a model of a PSA hydrogen separator, the classic Skarstrom process for producing an oxygen-enriched stream from air, and a simulation of TSA for air drying. In Chapter 2, we transfer our knowledge of Aspen Adsorption to Aspen Chromatography, and learn the differences between these software packages by building a model of an SMB process. We also discuss the differences in the assumptions that our models are allowed to make and the nuances of liquid- versus gas-phase isotherms. Because of the expensive calculations involved, we also discuss ways to improve the Aspen simulation performance. Chapter 3 covers the shortcut design methods for SMB processes. Some detailed derivation is performed using Mathematica, and the results are generalized to systems of arbitrary isotherms. Specifically, we provide a derivation and implementation of both triangle theory and standing wave design, two of the modern design techniques of SMBs. The Mathematica tools are used in a workshop to find the optimum operating conditions for the system modeled in Chapter 3. In Chapter 4, we extend our consideration to the so-called "Operational Modes" of SMB chromatography, which can yield large improvements in product purity, recovery, and other performance indices. Based on our research experience, we teach the reader how to simulate and optimize SMB operations by varying the feed flow rates ("PowerFeed" and "Partial Feeding"), varying the feed concentrations ("ModiCon"), and shifting asynchronously the inlet/outlet ports ("Varicol"), within the switching interval of the SMB operation. In Chapter 5, we cover methods for performing parameter estimation and regression using the models we built in Chapters 1 and 2, along with insights we gained from the analysis in Chapter 3.

Our review of currently available materials failed to find any text that encompassed the range of modeling techniques laid out here that emphasize our approach of integrating fundamental principles, industrial applications, and hands-on workshops and practice problems of commercial software tools for design, simulation, and optimization.

Since 1993, the senior author and the instructors he trained have taught industrial training courses to over 7,000 practicing engineers in 2017 FORTUNE global top one and two refining/chemical companies (SINOPEC and PetroChina) about hands-on applications of commercial software tools to the design, simulation, and optimization of petrochemical...