Search

International audience; Ammonia has a potential as carbon-free and high-energy density compound for chemical storage of renewable energies and its synthesis from green H2 requires to be as energetically efficient as possible. In this work, a superstructure optimization methodology for process synthesis is proposed and applied to an ammonia production process. The approach covers three different scales: process, equipment, and molecules. Process scale refers to finding the optimal process structure, while the equipment scale is related to the best set of operating conditions, and the molecular scale studies two catalysts (Fe and Ru) with their respective kinetic rates. The optimization intends to minimize the Levelized Cost of Ammonia (LCOA) and maximize the energy efficiency. The best trade-off is found with the Ru catalyst, with an LCOA of 766 €/tNH3 and 57.2 % of energy efficiency. In comparison with reference cases, the LCOA decreases 0.6 % and the energy efficiency increases 1.5%. The main improvement is found in the pressure, 75 bar, reduced in 25 %. The production of 11.6 tNH3/day equals to 2.5 MW of stored power from 3 MW supplied as H2 to the process, with a total energy consumption of 10.67 kW h/kgNH3, including prior H2 and N2 production processes.

This work presents a fully open-source superstructure optimization model for bi-criteria process design optimization. This model is implemented in the Open sUperstrucTure moDeling and OptimizatiOn fRamework (OUTDOOR). It combines mass, energy and cost balances with a more sophisticated heat integration concept with low computational effort and acceptable accuracy. The model is applied to a Power-to-Methanol (PtM) process design case study. A cost optimal methanol plant is identified at net production costs (NPC) of 892 €/tMeOH and net production emissions (NPE) of -1.937 tCO2-eq./tMeOH. It utilizes CO2 captured from refinery flue gas and hydrogen supply via. ambient pressure alkaline electrolysis. A plant configuration, designed for minimum CO2 emissions yields costs of 979 €/tMeOH with emissions of -2.191 tCO2-eq./tMeOH, using CO2 from ambient air and refinery and cement factory flue gases. A sensitivity analysis on electricity prices forecasts cost competitive methanol at large production capacities and low electricity prices of 2 ct/kWh.

The main objective of this study is to develop a decision-making tool for the design of the optimal municipal solid waste (MSW) facilities based on superstructure optimization. Currently, the disposal of MSW is a major problem due to the lack of awareness of the negative impacts resulting from dumping MSW into the environment. This poses a challenge for the authorities. MSW valorization such as anaerobic digestion, pyrolysis, gasification etc has been increasingly focused on as an approach when handling MSW to enhance both economic and environmental sustainability. However, with an increasing array of processing technologies, the design of MSW facilities involving the integration of these technologies is becoming tedious and unmanageable. To deal with this problem, superstructure optimization is proposed. It is an effective tool for the design of several chemical processes because it is able to consider all potential process alternatives including the optimal solution using mathematical models based on mass and energy balances. Uncertainty is incorporated into the optimization framework to enhance the robustness of the solution. The proposed methodology was applied in the design process of the MSW facility in Ubon Rathathani province, Thailand, with the objective function of maximizing the profit. The optimization problem was developed as Mixed Integer Linear Programming and it was solved using an optimization platform, General Algebraic Modeling System, with CPLEX as the solver related to obtaining the optimal solution. The results show there to be as positive profit that is economically viable compared to the use of landfill technology.

International audience; In this paper, we survey the main superstructure-based approaches in process system engineering, with a particular emphasis on the existing literature for automated superstructure generation. We examine both classical and more recent representations in terms of generality, ease of use, and tractability. We also discuss the implications that different representations may have on strategies for algebraic modeling and optimization. We then review the state-of-the-art in software implementations to support synthesis. Finally, we examine the use of evolutionary—recently referred to as superstructure-free—approaches, in which algorithmic procedures dynamically generate and evaluate candidate process structures.

As breakthrough cellular therapy discoveries are translated into reliable, commercializable applications, effective stem cell biomanufacturing requires systematically developing and optimizing bioprocess design and operation. This article proposes a rigorous computational framework for stem cell biomanufacturing under uncertainty. Our mathematical tool kit incorporates: high‐fidelity modeling, single variate and multivariate sensitivity analysis, global topological superstructure optimization, and robust optimization. The advantages of the proposed bioprocess optimization framework using, as a case study, a dual hollow fiber bioreactor producing red blood cells from progenitor cells were quantitatively demonstrated. The optimization phase reduces the cost by a factor of 4, and the price of insuring process performance against uncertainty is approximately 15% over the nominal optimal solution. Mathematical modeling and optimization can guide decision making; the possible commercial impact of this cellular therapy using the disruptive technology paradigm was quantitatively evaluated. © 2017 American Institute of Chemical Engineers AIChE J, 64: 3011–3022, 2018

This paper presents a novel methodology for the optimisation of the preliminary design of heat-integrated multicomponent distillation sequences. This is achieved through use of a reduced process superstructure where the role of splitting each adjacent key component pair is assigned to an individual separation column, greatly reducing the size and complexity of the problem. This methodology uses information about other process streams on site with which heat can be exchanged within the initial design in order to find a plant-wide optimal separation configuration with the aim of reducing the cost of heating provided by utilities including those used for heating and cooling process streams. In order that this model can be formulated as a mixed-integer linear program, this model utilises a discretised temperature grid where stream temperatures are allowed to vary. This methodology was tested on an example of a mixed alkane feed stream, with the sequence changing dependent on the degree of process integration. The method was found to have the potential for significant cost reductions compared to a heuristic design, with the example exhibiting a cost saving of over 50% and a reduction in CO2 associated with process heating of almost 60%, though the magnitude of these savings is highly dependent on the specific example to which it is applied.

Advances in simulation and optimization technologies coupled with the continued growth in computing power now increasingly pave the way for the development of advanced model-based engineering design frameworks. In this work, we propose an extensive computational framework, which brings together state-of-the-art engineering practices, such as high fidelity process simulation, superstructure-based conceptual design, global sensitivity analysis, Monte Carlo procedures for uncertainty quantification, and a stochastic simulation-based design space optimizer in order to foster decision making under uncertainty. The capabilities of the framework are highlighted in a case study, which addresses the challenges of how to synthesize and design wastewater treatment plant configurations under influent uncertainties. In order to handle multiple stochastic constraints, a black-box solver using a new infill criterion for surrogate-based optimization is also proposed. The results demonstrate the promising potential of the simulation and sampling-based framework for effectively addressing stochastic design problems arising in broader engineering domains.

The past decades have seen a diversification of the sugarcane industry with the emergence of new technology to produce bioenergy from by-product and waste process streams. Given Brazil’s ambitious goal of reducing green-house gas emissions by over 40% below 2005 levels by 2030, it is of paramount importance to develop reliable decision-making systems in order to stimulate investment in these low-carbon technologies. This paper seeks to develop a more accurate optimization model to inform risk-conscious investment decisions for bioenergy generation capacity in sugarcane mills. The main objective is for the model to enable a better understanding of how Brazilian government policies, such as the electricity price in the regulated market, may impact these investments, by taking into account the uncertainty in sugar, ethanol and spot electricity markets and the interdependency between production and investment decisions in terms of saleable product mix. The proposed methodology combines portfolio optimization theory with superstructure process modeling and it relies on simple surrogates derived from a detailed sugarcane plant simulator to retain computational tractability and enable scenario analysis. The case study of an existing sugarcane plant is used to demonstrate the methodology and illustrate how the model can assist decision-makers. In all of the scenarios assessed, the model recommends investment in extra bioelectricity capacity via the anaerobic digestion of vinasse but advises against investment in second-generation ethanol production via the hydrolysis of surplus bagasse. Furthermore, the decision to upgrade the cogeneration system with a condensation turbine is highly sensitive to the electricity price practiced in the regulated market, capacity constraints on the sugar-ethanol mix, and the accepted level of risk. Another key insight drawn from the case study is that recent market conditions have favored a production focused on the sugar business, making it challenging for policy-makers to create attractive scenarios for biofuels. Long-term electricity contracting appears to be the main hedging strategy for de-risking other products and investments in the sugarcane business, provided it is priced adequately.

Industrial operations consume energy and water in large quantities without accounting for potential economic and environmental burdens on future generations. Consumption of energy (mainly in the form of high pressure steam) and water (in the form of process water and cooling water) are essential to all processes and are strongly correlated, which requires development of systematic methodologies to address their interconnectivity. To this end, the subject of heat-integrated water allocation network design has received considerable attention within the research community in recent decades while further growth is expected due to imposed national and global regulations within the context of sustainable development. The overall mathematical model of these networks has a mixed-integer nonlinear programming formulation. As discussed in this work, proposed models in the literature have two main difficulties dealing with heat–water specificities, which result in complex formulations. These difficulties are addressed in this work through proposition of a novel nonlinear hyperstructure and a sequential solution strategy. The solution strategy is to solve three sub-problems sequentially and iteratively generate a set of potential solutions through the implementation of integer cut constraints. The novel mathematical approach also lends itself to an additional innovation for proposing multiple solutions balancing various performance indicators. This is exemplified with both a literature test case and an industrial-scale problem. The proposed solutions address a variety of performance indicators which guides decision-makers toward selecting the most appropriate configuration(s) among a large number of potential possibilities. Results exhibit that despite having a sequential solution strategy, better performance can be reached compared to previous approaches with the additional benefit of providing many potential solutions for further consideration by decision-makers to select the best case-specific solution.

Genome-scale models have become indispensable tools for the study of cellular growth. These models have been progressively improving over the past two decades, enabling accurate predictions of metabolic fluxes and key phenotypes under a variety of growth conditions. In this work, an efficient computational method is proposed to incorporate genome-scale models into superstructure optimization settings, introducing them as viable growth models to simulate the cultivation section of biorefinaries. We perform techno-economic and life-cycle analyses of an algal biorefinery with five processing sections to determine optimal processing pathways and technologies. Formulation of this problem results in a mixed-integer nonlinear program, in which the net present value is maximized with respect to mass flowrates and design parameters. We use a genome-scale metabolic model of Chlamydomonas reinhardtii to predict growth rates in the cultivation section. We study algae cultivation in open ponds, in which exchange fluxes of biomass and carbon dioxide are directly determined by the metabolic model. This formulation enables the coupling of flowrates and design parameters, leading to more accurate cultivation productivity estimates with respect to substrate concentration and light intensity.

Artificial Neural Networks (ANNs) have been used in a wide range of applications for complex datasets with their flexible mathematical architecture. The flexibility is favored by the introduction of a higher number of connections and variables, in general. However, over-parameterization of the ANN equations and the existence of redundant input variables usually result in poor test performance. This paper proposes a superstructure-based mixed-integer nonlinear programming method for optimal structural design including neuron number selection, pruning, and input selection for multilayer perceptron (MLP) ANNs. In addition, this method uses statistical measures such as the parameter covariance matrix in order to increase the test performance while permitting reduced training performance. The suggested approach was implemented on two public hyperspectral datasets (with 10% and 50% sampling ratios), namely Indian Pines and Pavia University, for the classification problem. The test results revealed promising performances compared to the standard fully connected neural networks in terms of the estimated overall and individual class accuracies. With the application of the proposed superstructural optimization, fully connected networks were pruned by over 60% in terms of the total number of connections, resulting in an increase of 4% for the 10% sampling ratio and a 1% decrease for the 50% sampling ratio. Moreover, over 20% of the spectral bands in the Indian Pines data and 30% in the Pavia University data were found statistically insignificant, and they were thus removed from the MLP networks. As a result, the proposed method was found effective in optimizing the architectural design with high generalization capabilities, particularly for fewer numbers of samples. The analysis of the eliminated spectral bands revealed that the proposed algorithm mostly removed the bands adjacent to the pre-eliminated noisy bands and highly correlated bands carrying similar information.

The growing desire to improve resource efficiency and environmental impact of industrial processes is directly linked to optimal management of heat, mass and power flows. The concept of industrial symbiosis tackles this issue by proposing interplant heat recovery and resource transfer which can bring economical and environmental benefits to each party. A comprehensive methodology is required which can easily be incorporated in the planning of industrial clusters. Therefore, a generic hybrid mixed integer linear programming superstructure has been developed to address simultaneous heat, water, and power optimization in interplant operations. Additional concepts are included in the previously-proposed water network superstructure (Kermani et al., 2017) to account for the issues related to interplant heat and mass exchange. A cold utility superstructure is included in the water network while a steam network superstructure is modified to better represent the feedwater heaters and heat recovery opportunities. The proposed methodology is applied to an industrial case study. Results exhibit a large potential for synergies among industrial sites, even in disparate sectors, and emphasize the importance of a generic approach.

A multistage multilayer systematic procedure has been developed for the selection of the optimal product portfolio from waste biomass as feedstock for systems involving water–energy–food nexus. It consists of a hybrid heuristic, metric-based, and optimization methodology that evaluates the economic and environmental performance of added-value products from a particular raw material. The first stage preselects the promising products. Next, a superstructure optimization problem is formulated to valorize or transform waste into the optimal set of products. The methodology has been applied within the waste to power and chemicals initiative to evaluate the best use of the biomass residue from the olive oil industry toward food, chemicals, and energy. The heuristic stage is based on the literature review to analyze the feasible products and techniques. Next, simple metrics have been developed and used to preselect products that are promising. Finally, a superstructure optimization approach is used to design the facility that processes leaves, wood chips, and olives into final products. The best technique to recover phenols from “alperujo”, a wet solid waste/byproduct of the process, consists of the use of membranes, while the adsorption technique is used for the recovery of phenols from olive leaves and branches. The investment required to process waste adds up to €110.2 million for a 100 kt/yr for the olive production facility, while the profit depends on the level of integration. If the facility is attached to an olive oil production, the generated profit ranges between 14.5 MM €/yr (when the waste is purchased at prices of €249 per ton of alperujo and €6 per ton of olive leaves and branches) and 34.3 MM €/yr when the waste material is obtained for free., Methodology and case study for the selection of added-value products out of waste and process design.

This work deals with the design of integrated facilities for the production of xylitol and sorbitol from lignocellulosic biomass. Xylitol can be obtained from xylose via fermentation or catalytic hydrogenation. Sorbitol is obtained from glucose, but preferably from fructose, and also via fermentation or catalytic hydrogenation. Fructose can be obtained from glucose via isomerization. Thus, a superstructure of alternatives is formulated to process switchgrass, corn stover, miscanthus, and other agricultural and forestry residues. Different pretreatments, such as dilute acid or ammonia fiber explosion (AFEX), for the fractionation of the biomass are evaluated. Next, after hydrolysis, the C5 and C6 sugars are processed separately for which a catalytic or a fermentation stage are considered. Glucose has to be isomerized before it can be processed. Finally, crystallization in a multistage evaporator system is used for purification. The optimization of the system is done by the use of dilute acid and the catalytic system. A system of 3 crystallizers is selected. For a facility that produces 145 kt/yr of xylitol and 157.6 kt/yr of sorbitol, the investment adds up to 120.74 M€ for a production cost of 0.28 €/kg products. The inverse engineering of biomass was also performed resulting in a composition of 15% water, 20% cellulose, 40% hemicellulose, 15% lignin, and 5% ash. The closest biomass corresponds to Sargassum (brown algae), which is capable of producing 230.5 kt/yr of xylitol and 116 kt/yr of sorbitol with investment and production costs of 120.5 M€ and 0.25 €/kg products, respectively.

Industries consume large quantities of energy and water in their processes which are often considered to be peripheral to the process operation. Energy is used to heat or cool water for process use; additionally, water is frequently used in production support or utility networks as steam or cooling water. This enunciates the interconnectedness of water and energy and illustrates the necessity of their simultaneous treatment to improve energy and resource efficiency in industrial processes. Since the seminal work of Savulescu and Smith in 1998 introducing a graphical approach, many authors have contributed to this field by proposing graphically- or optimization-based methodologies. The latter encourages development of mathematical superstructures encompassing all possible interconnections. While a large body of research has focused on improving the superstructure development, solution strategies to tackle such optimization problems have also received significant attention. The goal of the current article is to study the proposed methodologies with special focus on mathematical approaches, their key features and solution strategies. Following the convention of Jezowski, solution strategies are categorized into: decomposition, sequential, simultaneous, meta-heuristics and a more novel strategy of relaxation/transformation. A detailed, feature-based review of all the main contributions has also been provided in two tables. Several gaps have been highlighted as future research directions.