Informatics Rich Integrated Chemical Safety Risk Assessment (ICSRA) Platform

Informatics Rich Integrated Chemical Safety Risk Assessment (ICSRA) Platform

M Karthikeyan1*, Renu Vyas2, HV Prasad3

1 Centre of Excellence in Scientific Computing and Digital Information Resource Centre, CSIR-National Chemical Laboratory, Pune – 411 008, India

2 Chemical Engineering and Process Development, CSIR-National Chemical Laboratory, Pune – 411 008, India

3 CSIR-Indian Institute of Chemical Technology, Habsiguda, Hyderabad-500 607, India

Abstract

Inherently Safer Practices (ISP) is a systematic innovative scientific methodology integrated with the process, design and operations for facilitating the improvement of process safety at all stages of process life cycle.   ISP constitutes intensification (minimization), substitution, attenuation, limitation of effects and simplification as procedural options for implementation in process plants.   

For this study we conducted literature data search focused on recent domestic and international incidents where the primary cause was related to chemical reactivity and runaway reactions; Several parameters were taken into consideration including chemical manufacturing (i.e., raw material storage, chemical processing, and product storage) and other industrial activities involving bulk chemicals, such as storage/distribution, waste processing, and petroleum refining. For purposes of the incident search, only reactive incidents that caused serious consequences would be examined. Source of data: Four decades of data collection involve searching over 40 data sources, focusing on incidents where the primary cause was related to chemical reactivity. For the purposes of the investigation and analysis, an "incident" was defined as a sudden event involving an uncontrolled chemical reaction—with significant increases in temperature, pressure, and/or gas evolution etc.,

The proposed experimental and simulation platform will serve as a repository for plant-specific information, and facilitates the classification and ranking of Indian chemical industry - for the first time - through a Risk index.

Objectives

Computational and Simulation of Industrial chemical reactions to predict runaway reactions and conditions as safety measures. To our knowledge, no systematic studies on inherent safety are yet reported in India. The studies comprise important research areas e.g., process evaluation, estimation of reactivity, hazard identification and evaluation, assessment and prevention of thermal runaways, early fault diagnosis etc. The computational risk assessment platform intends to collect new chemical process safety and risk data, and plant-specific incident data from participating industries across the world for model building and simulation.

To integrate and include a dynamic interactive information platform, viz., chemical safety risk ranking and incident reporting system, by the application of High performance Computing (Cloud, Grid and Distributed systems) for simulation, modeling and response for use by Govt., Industry, Emergency Responders, Research and academic communities, and the public. Upgradation of Thermo chemical Laboratories to assess the proposed/existing industry from the context of implementing ISP, i.e., towards safer process chemistry including selection of solvents, safer process routes, safer process operating conditions for preventing Runaways, early detection of runaways, evaluation/selection of process and designs to handle lesser inventories etc.

Methodology

 The methodology devised comprises recognition of the hazards and risks posed by the process, and efforts to reduce or control them to the lowest practical levels, keeping in view the principle objectives of the business.  Here we studied data based on literature to identify and collect chemically significant industrially important data pertaining to physico-chemical, biological and toxicological profiles of chemicals (reactants, reagents, catalysts, solvents, products, side products) for their suitability and reusability in simulation and modelling. This includes development of suitable algorithms, chemical reaction component descriptors and methods to simulate experimental conditions and validation through experiments to improve the predictions of risk assessment and prevention of disaster in industrial conditions (Lab-scale, pilot plant and large scale manufacturing).

This process also includes mapping known literatures (Publications, patents and reports) to the existing knowledge for interactive risk assessment of chemicals and incidents in industrial environment. The computational approaches include classical methods, semi-empirical methods, and quantum methods to determine properties at the molecular level. We are developing a computational problem-solving platform and these tools, data and methods to build ICSRA a cloud-computing infrastructure for integrated chemical safety risk assessment of chemicals manufactured, imported and exported from world chemical sources. Proper documentation and accurate predictive models on chemicals of Indian origin would facilitate compliance with other global compliance on chemicals being exported especially in REACH, EPA, FDA and other global federal regulation context.

Development of QSPR (Quantitative Structure Property Relationships) models to relate macroscopic behaviour to molecular properties with detailed molecular computations to obtain thermo chemical data for reactive materials and to predict calorimetric data based on molecular properties. These open source based tools are integrated to this platform (ChemInfoCloud) includes identification and assessment of hazards due to chemical reactivity and risk assessment. For this systematic study we considered major potential chemical group (CG) which are frequently encountered in industrial reagents like Organic peroxides, Organic Nitrites, Unsaturated Hydrocarbons, Epoxides, Hydrides and Hydrogen, Metal acetalides etc. Efforts to study and model major Chemical Reaction Types (CRT) include Decomposition, Polymerization, Pyrophoric, Peroxide Former, Water sensitive and Oxidation reactions. Molecular computing would help to predict heat of formation, electronic energy, core-core repulsion, dipole, no. of filled levels, ionization potential, molecular weight, topological, electronic and three dimensional descriptors for critical components of reactants, products, solvents.  Theoretical study of transition states of chemical reactions, computation of activation energies and experimental validation to predict that one reaction pathway is favoured over another reaction pathway. In order to predict the type of polymerization and runaway reactions enthalpy of reactions and kinetics was calculated using the semi-empirical quantum methods, molecular modelling and dynamics simulations of reactants under simulated diverse experimental conditions of Industrial Chemical Processes (ICP), including homogeneous and heterogeneous catalysis

<>Computational Approach to Process Safety

Solvents

Computed Data of Solvents

Digital mapping of incidental data concerning process safety and their availability to decision makers would help effective risk assessment and rapid response during unexpected events.

During the presentation the detailed roadmap, current scenario and future direction of the chemical safety risk assessment platform will be discussed and the input from delegates for this meeting would be considered for integration into the tool if required.