This dissertation presents an analysis of two aspects of the chromatographic separation process known as Simulated Moving Bed (SMB) chromatography. The first aspect is system design, and the second is improving computer simulations to generate heuristics for choosing operational modes.

For the past 15-20 years, there has been a surge of interest in the use of Simulated Moving Bed systems for the chromatographic separation of chemicals¹. A wide variety of methods, nomenclatures, and conventions have been adopted over the years²⁻⁴, as teams from different backgrounds adopt and improve on the SMB technology. This work presents a unifying discussion of the two major design methods, Triangle Theory and Standing Wave Design, used in the SMB field. We provide the complete computer code required to execute both design methods. A sample problem is worked, which demonstrates the novelty and ease of use that such tools provide. Mathematica was chosen for the implementation of these design methods, because of its strong symbolic analysis capabilities, and simplicity of creating interfaces for new users. We present derivations of the classic Langmuir results in Mathematica, and proceed to extend those implementations. When analytic solutions are impossible, we use Mathematica's numerical methods.

This work also develops a distributed computing tool known as ChromRunner which allows large numbers of detailed numerical simulations to be run simultaneously. The motivations and benefits of this approach are discussed alongside implementation details. We apply the distributed computing system to two separate SMB separations in order to optimize them, as well as determine heuristics governing their operational modes. We wrote ChromRunner in C#, and took advantage of Visual Studio's Entity Framework to create the database backend. The user interface for this software was created using Microsoft's "Windows Presentation Foundation" (WPF) technologies.