Srinivas Rangarajan

Automated identification of energetically feasible mechanisms of complex reaction networks in heterogeneous catalysis: application to glycerol conversion on transition metals

An automated method is presented to identify energetically feasible mechanisms of heterogeneous catalytic systems comprising of several thousand reactions and species. Specifically, it combines automated rule-based network generation of complex networks with semi-empirical estimation of thermochemical properties to (a) generate large reaction networks, (b) calculate thermochemistry and activation barriers of each step on-the-fly, and (c) identify energetically feasible pathways and mechanisms to experimentally observed products. This method is applied to analyze the mechanism of glycerol conversion on transition metal catalysts for which a network of 3300 reactions and 500 species was generated. Using (i) group additivity methods for calculating reaction and species energies, (ii) linear free energy relationships for activation barriers, and (iii) linear scaling relationships for comparative assessment of energetics of species on different transition metals, the energetically feasible pathways for forming C–C scission and C–O scission products (syngas and propane diol, respectively) are identified on platinum, palladium, rhodium, and ruthenium catalysts. Our results indicate that (a) C–C scission products are preferred on platinum and palladium, (b) rhodium and ruthenium will have a comparatively higher selectivity for C–O scission products, and (c) glycerol tends to undergo several dehydrogenation steps prior to undergoing C–C scission, which are all consistent with experimental observations and DFT calculations. We propose that this method can be used to screen a large number of pathways in complex catalytic networks and to thereby identify the dominant modes in the system, because it is fast and scalable in handling larger networks, flexible in accommodating different types of semi-empirical correlations, and generic in handling any catalytic chemistry.

Reduction of complex energy-integrated process networks using graph theory

Abstract This paper focuses on the analysis of complex (multi-loop) energy-integrated process networks. Simple (single-loop) energy-integrated networks (comprising of large energy recycle or throughput) with two-time scale dynamics are the building blocks for such complex networks. The modular structure of these complex networks lends them to a graph theoretic analysis, whereby weak and strong connections between process units arising from time scale separation are identified from structural information. Subsequently, a graph-theoretic framework for network analysis and control is developed, and connecting links are built to an equivalent analysis using singular perturbations. The proposed analysis framework is illustrated via application to a representative complex process network.

Graph reduction for hierarchical control of energy integrated process networks

In this paper, we propose a graph-theoretic algorithm that can be used to analyze complex chemical processes comprising of multiple energy integration loops. Such networks are known to exhibit dynamics in multiple time scales. The algorithm uses information on the order of magnitude of the different energy flows and determines automatically the time scales where the units evolve, the manipulated inputs acting in the different time scales and the form of the reduced order models in each time scale. The application of the algorithm is illustrated through a case study of a benchmark chemical process.

Multi-Time Scale Dynamics in Energy-Integrated Networks: A Graph Theoretic Analysis

Abstract Energy integrated networks, designed to reduce energy consumption, offer cost savings at the expense of challenging operation. Simple networks with energy integration, involving either a large recycle of energy or a large throughput of energy, have been shown to exhibit dynamic behavior evolving over two time scales. In this paper, the time scale properties in the dynamics of networks involving multiple, interconnected throughputs and recycle loops are investigated. A graph theoretic analysis framework is developed which allows identifying the time scales where each of the process units in the network evolves, based on knowledge of the order of magnitude of the energy flows in the network. An example network is considered to illustrate the application of the proposed framework.

Optimization and Analysis of Chemical Synthesis Routes for the Production of Biofuels

Abstract In this work we extend a recently proposed method to concurrently select biofuel blends which satisfy ASTM standards and their optimal synthesis routes by using thermochemistry based post-processing analysis to compare these routes. Gas phase thermochemistry, estimated from group additivity, was used to calculate the equilibrium extent of reaction for each synthesis step. Situations of reaction, phase, and site coupling were subsequently identified to overcome equilibrium limited steps and reduce the number of reaction systems required. This method can aid process designers in screening and ranking large numbers of potential biofuel candidates and synthesis routes simultaneously from an energetic, thermochemical, economic, or kinetic standpoint.

Graph-theoretic Analysis of Complex Energy Integrated Networks*

Abstract Complex process networks featuring multiple energy integration agents (process-to-process heat exchangers) offer significant cost benefits while adding additional operational constraints. These networks show potential for multi-time scale dynamics owing to the presence of energy flows spanning multiple orders of magnitude. In previous work, we have developed a graph-theoretic framework to systematically uncover this time scale multiplicity. In this paper, we present an application of this framework to a reactor-heat exchanger system used for naphtha reforming. This system involves energy flows spanning three orders of magnitude and the underlying energy balance variables evolve over two time scales. The framework allows for the derivation of control-relevant models in each time scale and classifies the control objectives leading to a hierarchical control strategy. We demonstrate that the analysis uses minimum process information, is efficient, and scalable to large networks.

Network generation and analysis of complex biomass conversion systems

A modular computational tool for automated generation and rule-based post-processing of reaction systems in biomass conversion is presented. Cheminformatics and graph theory algorithms are used to generate chemical transformations pertaining to heterogeneous and homogeneous chemistries in the automated rule-based network generator. A domain-specific language provides a user-friendly English-like chemistry specification interface to the network generator. A rule-based pathway analysis module enables the user to extract and query pathways from the reaction network. A demonstration of the features of this tool is presented using Fructose to 5-Hydroxymethylfurfural as a case study.