Band 17, Heft 4 - Special Issue: CPPM special issue in honour of Professor W.Y. “Bill” Svrcek; Guest Editors: Brent R. Young and Ramin Bozorgmehry

Models are required to predict the onset and precipitation of asphaltenes from mixtures of heavy oil and solvents for a variety of heavy oil applications. The regular solution approach is well suited for this objective but has not yet been tested on solvent mixtures. To do so, the onset and amount of asphaltene precipitation were measured and modeled for mixtures of heavy oil with solvent blends made up from n -alkanes, cyclohexane, and toluene at temperatures of 21 and 180 °C and pressures of 0.1 and 10 MPa. Temperature dependent binary interaction parameters (BIP) between the cyclohexane/asphaltene and toluene/asphaltene pseudo-component pairs were proposed to match the data. All other BIP were set to zero. The model with BIP determined from asphaltene precipitation in heavy oil and binary solvents predicted asphaltene precipitation from heavy oil and ternary solvent blends, generally to within the experimental error.

Costs associated with production of favorable biologically produced surfactants continue to be a significant obstacle to large scale application. Using industrial wastes and by-products as substrate and optimization of cultural conditions are two strategies of producing biosurfactants with a reasonable price. Also, modeling the biosurfactant production bioprocess improves the commercial design and monitoring of biomass growth, biosurfactant production, and substrate utilization. In this study, the indigenous Bacillus subtilis N3-1P strain and a local brewery waste as the carbon source were used to produce a biosurfactant. The batch cultivation was performed under the optimum conditions. Models describing the biomass growth, biosurfactant production, and substrate utilization were developed by fitting the experimental data to the logistic, Contois and Luedeking-Piret models using MATLAB software and regression analysis. The kinetic parameters including the maximum specific growth rates ( µ max ), the Contois constant ( K ), parameters of the Luedeking-Piret modelswere calculated. Yields including Y X / S , and Y P / X were found to be 0.143 g X/ g S , and 0.188 g P/ g X , respectively. The experimental and predicted model showed good agreement. The developed models are a key step in designing reactors for scale up of biosurfactant production.

This paper aims to develop an innovative and practical method to accelerate Pinch Analysis problems using a combination of process integration, process simulation tools, and life cycle assessment techniques to design effective industrial energy efficiency measures rapidly. Data extraction and analysis is accelerated using the workflow, reducing pinch analysis duration, and therefore cost. This lowers financial barriers and is expected to help facilitate an increase in the number of conducted pinch analyses each year. An innovative ecological targeting step has been included in the workflow, which represents the option for end-users to further their CO 2 savings, once energy targeting has been completed on a purely cost basis. It is expected that optimizing for costs first will enable the implementation of heat recovery solutions in industry. The methodology is applied to an example case study, and the learnings from industrial application is reported.

Phase equilibrium K values are either estimated with empirical correlations or rigorously calculated based on fugacity values determined from an equation of state. There have been several empirical analytical equations such as Raoult’s Law, the Hoffman Equations (Hoffman A, Crump J, Hocott C. Equilibrium constants for a gas condensate system. J Petrol Technol 1953;5:1–10) and their modifications and the well-known Wilson Equation (Wilson G. A modified Redlich–Kwong equation of state applicable to general physical data calculations. In: AIChE National Meeting Paper15C, May 4–7, Cleveland, OH; 1969). along with several modifications. This work presents a new modification of the Wilson Equation for estimating phase equilibrium K values, predominantly for light hydrocarbon mixtures. The modification is based on correlating a subset of a database of K values, established from convergence pressure data. Results show the method to accurately correlate and predict the K value data, within 10% on average. Moreover, the predicted K factors provide remarkable results for such a simple model when used in a variety of phase equilibrium calculations. The results also show that the new model compares favorably with existing empirical analytical methods. Such a model would provide excellent initial estimates for rigorous thermodynamic calculations.

Athabasca bitumen is an abundant resource that has successfully been upgraded using delayed coking that typically operates at 499 °C (∼930 °F), 207 kPa (∼37 psig), 1–2 min residence time on this type of crude. With society’s desire to reduce industry environmental impact while still providing energy to earth’s growing population, lower energy intensive (and thus lower greenhouse gas emissions) bitumen conversion approaches have been researched and are moving towards commercialization. The paper reviews a correlative model developed on a novel thermal cracking process, operated at lower temperatures (395–405 °C (743–761 °F)), lower pressures (<69 kPa (∼<10 psig) and up to 1 h residence time versus delayed coking, that takes various lab and pilot data, both batch and continuous, as inputs into developing the model. The purpose of the model is for use in industrial operations to provide guidance to operations for representative thermal cracker performance. The model is based on the Arrhenius equation using first order reaction kinetics for easy comprehension and use in an operational environment. Data for developing the model has been taken from various literature sources in the area of study, notably by researchers, Dr. W. Svrcek, Dr. Wiehe, Dr. Mehrotra, and Dr. Yarranton. The public data is used to create a viable range of performance that includes proprietary developments with the novel thermal cracking process. The model is configured on a mass basis so that mass balance closure can be readily calculated. A range of kinetic coefficients are provided that can be used to fit commercial plant performance based on the expected range of product outputs noted in the paper.

The term ‘digital twin’ refers to a virtual simulation/model (virtual twin) of a physical plant or object (physical twin), where data flows between the virtual and physical twins. A digital twin can be used for different purposes, such as process optimisation/control, design, training, and maintenance/service. This manuscript found an increasing number of simulation and modelling publications in literature year on year, which illustrates the current trend towards implementing digital twins in a broad range of process engineering applications. A targeted literature review into the area found several commercial off-the-shelf software solutions (COTS) for different industrial applications providing the necessary flexibility to analyse a broad range of industries. However, most of the process modelling software is designed for petroleum and fine chemicals processes. There is still a need for software solutions that can model a broader range of applications. While most of the software found was licensed, open source process modelling software was also available. There is a lack of independent research into the accuracy of these software solutions. The literature review also found that 37% of the research based on process simulations is carried out to improve energy efficiencies. In comparison, 27% of the research found Decarbonization to be a secondary "added" benefit. It can be concluded that digital twins are ideally suited for driving energy efficiency improvements and decarbonisation goals. However, none of the COTS identified in the literature meets all the requirements for a digital twin. A solution to this problem is to create a layered digital twin, combining and interfacing different tools to accomplish a visually similar, self-optimising, self-learning virtual plant.