M. Sam Mannan

Fuzzy risk matrix.

A risk matrix is a mechanism to characterize and rank process risks that are typically identified through one or more multifunctional reviews (e.g., process hazard analysis, audits, or incident investigation). This paper describes a procedure for developing a fuzzy risk matrix that may be used for emerging fuzzy logic applications in different safety analyses (e.g., LOPA). The fuzzification of frequency and severity of the consequences of the incident scenario are described which are basic inputs for fuzzy risk matrix. Subsequently using different design of risk matrix, fuzzy rules are established enabling the development of fuzzy risk matrices. Three types of fuzzy risk matrix have been developed (low-cost, standard, and high-cost), and using a distillation column case study, the effect of the design on final defuzzified risk index is demonstrated.

Numerical simulations of LNG vapor dispersion in Brayton Fire Training Field tests with ANSYS CFX

Federal safety regulations require the use of validated consequence models to determine the vapor cloud dispersion exclusion zones for accidental liquefied natural gas (LNG) releases. One tool that is being developed in industry for exclusion zone determination and LNG vapor dispersion modeling is computational fluid dynamics (CFD). This paper uses the ANSYS CFX CFD code to model LNG vapor dispersion in the atmosphere. Discussed are important parameters that are essential inputs to the ANSYS CFX simulations, including the atmospheric conditions, LNG evaporation rate and pool area, turbulence in the source term, ground surface temperature and roughness height, and effects of obstacles. A sensitivity analysis was conducted to illustrate uncertainties in the simulation results arising from the mesh size and source term turbulence intensity. In addition, a set of medium-scale LNG spill tests were performed at the Brayton Fire Training Field to collect data for validating the ANSYS CFX prediction results. A comparison of test data with simulation results demonstrated that CFX was able to describe the dense gas behavior of LNG vapor cloud, and its prediction results of downwind gas concentrations close to ground level were in approximate agreement with the test data.

Stabilization of Pickering foams by high-aspect-ratio nano-sheets

We developed Pickering foams highly stabilized by high-aspect-ratio (ξ = diameter/thickness) nano-sheets. The effects of particle aspect ratio, concentration, and hydrophobicity were also investigated. To our knowledge, our study provides the first experimental evidence of the effect of particle aspect ratio on particle-stabilized foams. The adsorption properties of these highly anisotropic nano-sheets are strongly affected by their small thickness and large lateral size (i.e., two-dimensional). These high-aspect-ratio nano-sheets were obtained by exfoliation of α-zirconium phosphate (ZrP) crystals with propylamine (C3H7NH2, PA). The hydrophobicity of the nano-sheets was tailored by adjusting the PA : ZrP molar ratio in the suspension. The morphology and stability of the foam depend on the nano-sheet aspect ratio and concentration as well as on the PA : ZrP molar ratio. Here, we found that using low and high aspect ratio nano-sheets having a high and an intermediate degree of hydrophobicity, respectively, is the successful formula to obtain high foam stability. The aqueous foams were characterized by optical and cross-polarized micrographs. Scanning electron microscopy (SEM) micrographs of dried foams revealed the adsorption of the PA–ZrP nano-sheets on the air–water interface. The foam stability was studied by measuring the foam and the water volume as a function of time to obtain the foam decay and water drainage rate, respectively. We also observed that the foams were stabilized by jammed layers of nano-sheets located in the bulk and at the air–water interface. These layers of particles prevent air diffusion between the bubbles, hence arresting Ostwald ripening and coalescence.

Optimal facility layout under toxic release in process facilities: A stochastic approach

Abstract This paper presents a new approach for the optimal facility siting considering the uncertainty of toxic release in one of the installed facilities. The proposed formulation incorporates the effect of wind speed, wind direction and atmospheric stability to calculate the risk of death via probit functions and Monte Carlo simulation. The overall problem is initially modeled as a disjunctive program where the Cartesian coordinates of each new facility for siting and cost-related variables are the main unknowns to determine. Then, the convex hull approach is used to reformulate the problem as a mixed integer nonlinear program (MINLP). The numerical difficulties are shown in a case study where multiple optimal layouts have been found. In general, the numerical results demonstrate the potential of this approach to improve the process layout design activity.

Adiabatic calorimetric decomposition studies of 50 wt.% hydroxylamine/water.

Calorimetric data can provide a basis for determining potential hazards in reactions, storage, and transportation of process chemicals. This work provides calorimetric data for the thermal decomposition behavior in air of 50wt.% hydroxylamine/water (HA), both with and without added stabilizers, which was measured in closed cells with an automatic pressure tracking adiabatic calorimeter (APTAC). Among the data provided are onset temperatures, reaction order, activation energies, pressures of noncondensable products, thermal stability at 100 degrees C, and the effect of HA storage time. Discussed also are the catalytic effects of carbon steel, stainless steel, stainless steel with silica coating, inconel, titanium, and titanium with silica coating on the reaction self-heat rates and onset temperatures. In borosilicate glass cells, HA was relatively stable at temperatures up to 133 degrees C, where the HA decomposition self-heat rate reached 0.05 degrees C/min. The added stabilizers appeared to reduce HA decomposition rates in glass cells and at ambient temperatures. The tested metals and metal surfaces coated with silica acted as catalysts to lower the onset temperatures and increase the self-heat rates.

Layer of protection analysis for reactive chemical risk assessment

Reactive chemical hazards have been a significant concern for the chemical process industries (CPI). Without sufficient control and mitigation of chemical reaction hazards, reactive incidents have led to severe consequences, such as release of flammable and toxic materials, fires and explosions, and threats to human lives, properties, and the environment. Consequence of reactive hazards can be well understood through calorimetric testing and computational techniques. However, risks of incidents caused by reactive chemicals have not been well addressed due partly to sparse failure frequency data. In this paper, the semi-quantitative layer of protection analysis (LOPA) approach is used to estimate reactive chemical risk, and the probabilities or frequencies of failure scenarios are addressed. Using LOPA, reactive risks can be evaluated with respect to predefined criteria, and the effectiveness of risk reduction measures can be assessed. The hydroxylamine (HA) production system is employed as a case study to demonstrate the application of LOPA to reactive chemical risk assessment.

Utilization of accident databases and fuzzy sets to estimate frequency of HazMat transport accidents.

Risk assessment and management of transportation of hazardous materials (HazMat) require the estimation of accident frequency. This paper presents a methodology to estimate hazardous materials transportation accident frequency by utilizing publicly available databases and expert knowledge. The estimation process addresses route-dependent and route-independent variables. Negative binomial regression is applied to an analysis of the Department of Public Safety (DPS) accident database to derive basic accident frequency as a function of route-dependent variables, while the effects of route-independent variables are modeled by fuzzy logic. The integrated methodology provides the basis for an overall transportation risk analysis, which can be used later to develop a decision support system.

Thermal Decomposition Pathways of Hydroxylamine: Theoretical Investigation on the Initial Steps

Hydroxylamine (NH(2)OH) is an unstable compound at room temperature, and it has been involved in two tragic industrial incidents. Although experimental studies have been carried out to study the thermal stability of hydroxylamine, the detailed decomposition mechanism is still in debate. In this work, several density functional and ab initio methods were used in conjunction with several basis sets to investigate the initial thermal decomposition steps of hydroxylamine, including both unimolecular and bimolecular reaction pathways. The theoretical investigation shows that simple bond dissociations and unimolecular reactions are unlikely to occur. The energetically favorable initial step of decomposition pathways was determined as a bimolecular isomerization of hydroxylamine into ammonia oxide with an activation barrier of approximately 25 kcal/mol at the MPW1K level of theory. Because hydroxylamine is available only in aqueous solutions, solvent effects on the initial decomposition pathways were also studied using water cluster methods and the polarizable continuum model (PCM). In water, the activation barrier of the bimolecular isomerization reaction decreases to approximately 16 kcal/mol. The results indicate that the bimolecular isomerization pathway of hydroxylamine is more favorable in aqueous solutions. However, the bimolecular nature of this reaction means that more dilute aqueous solution will be more stable.

Fault detection and classification in chemical processes based on neural networks with feature extraction.

Feed forward neural networks are investigated here for fault diagnosis in chemical processes, especially batch processes. The use of the neural model prediction error as the residual for fault diagnosis of sensor and component is analyzed. To reduce the training time required for the neural process model, an input feature extraction process for the neural model is implemented. An additional radial basis function neural classifier is developed to isolate faults from the residual generated, and results are presented to demonstrate the satisfactory detection and isolation of faults using this approach.

The development and application of dynamic operational risk assessment in oil/gas and chemical process industry

Abstract A methodology of dynamic operational risk assessment (DORA) is proposed for operational risk analysis in oil/gas and chemical industries. The methodology is introduced comprehensively starting from the conceptual framework design to mathematical modeling and to decision making based on cost–benefit analysis. The probabilistic modeling part of DORA integrates stochastic modeling and process dynamics modeling to evaluate operational risk. The stochastic system-state trajectory is modeled according to the abnormal behavior or failure of each component. For each of the possible system-state trajectories, a process dynamics evaluation is carried out to check whether process variables, e.g., level, flow rate, temperature, pressure, or chemical concentration, remain in their desirable regions. Component testing/inspection intervals and repair times are critical parameters to define the system-state configuration, and play an important role for evaluating the probability of operational failure. This methodology not only provides a framework to evaluate the dynamic operational risk in oil/gas and chemical industries, but also guides the process design and further optimization. To illustrate the probabilistic study, we present a case-study of a level control in an oil/gas separator at an offshore plant.