Simultaneous product and processes.design using reverse design algorithm
Chemical Product Design and Engineering (CPD&E)
Chemical Product Design & Engineering - II (CPD&E - 2)
Keywords: Reverse design approach, Simultaneous process product design, Membrane based processes
In order to meet the demands of specifically architectured fine chemicals it is advantageous to design the process and product simultaneously. A systematic model based approach, is developed which consists of a framework of multiscale process and product models. The idea is to design simultaneously process and product and generating design alternatives for a given set of performance criteria with proper justification and validation. This gives us more flexibility in terms of choosing the design variables that could either be the process parameters (like temperature, pressure etc.), product parameters (like molecular structure, selectivity, solubility etc.) or both in order to achieve the process demands. A design algorithm based on the reverse design methodology and a generic model that is able to handle wide variety of products and processes will be presented.
The solution strategy based on reverse design approach, for designing the process and product, splits the solution steps in two stages. In the first stage, the process model that includes the balance and constraint equations are solved keeping of property parameters of the system as unknown variables. In the second stage various property models (constitutive equations of the original process model) are solved in order to get the design variables, that matches the target properties calculated in stage I. These are the key properties of the system which affects the performance of the process. For example, reaction rate constant or dissociation constant for reactive systems, driving force for distillation or liquid-liquid extraction etc., thermodynamic or kinetic properties for solution diffusion kind of separation, selectivity of solvents for solvent based separation etc. These key properties in turn depend on parameters like process conditions (temperature, pressure, flowrates etc.), parameters related to the equipment, chemical structure of solvent or entrainers (for azeotropic distillation), microscopic structure of polymers and/or support layer for membrane based separation processes etc. Instead of choosing these parameters for any process and evaluating property parameters using constitutive equations during the solution of process model and then checking if the desired performance criteria is achieved or not, it is computationally convenient to split the solution steps in two stages as mentioned before. In this way, the hierarchal approach converges from the inlet and outlet specifications of a process to the product and process properties which leads to the design of the product to match the performance criteria of process, hence designing both of them simultaneously in one framework.
This methodology as compared to the conventional forward approach is computationally inexpensive as it does not require the property models to be embedded in the process model and solved simultaneously. It also gives the opportunity of combining models that are multidisciplinary and multiscaled. In the second stage of the algorithm, as many property models can be used in order to generate the design alternatives for both product and process. It is not an iterative process like the forward approach where different designs are tested in a trial and error method in order to match the performance criteria.
The methodology along with its application to design membrane based separation processes will be shown for processes like Vacuum membrane distillation (VMD), pervaporation and membrane based gas separation. The physics of the system depends very much on the separation process and the kind of polymer used as membrane in the process. For example, for both pervaporation (liquid separation) and gas separation the polymer generally used is dense polymer and hence the transport mechanism is solution-diffusion mechanism, which makes permeability the key property in the separation process. While on the other hand for VMD, usually porous membranes are used making porosity, tortousity etc. as properties that enhance or retard the separation process. Property models to calculate some of these properties will also be shown. Results for permeability calculations done using molecular modeling for different structures of polymers (polyethylene and polyisobutylene in particular) will be highlighted.
Presented Wednesday 19, 15:40 to 16:00, in session Chemical Product Design & Engineering - II (CPD&E - 2).
