Process Operability Algorithms: Past, Present, and Future Developments

Abstract

In this Article, the recently developed multimodel and NLP-based operability approaches are reviewed and further developed in terms of theory, application, and software infrastructure. Classical process operability concepts are also revisited and contrasted with new extensions present in both approaches. A focus is given on the comparison between distinct measures of the operability index and the handling of infeasible portions of the desired operating region. For the software infrastructure outcome, a framework for the development of process operability algorithms is provided. In particular, generated codes from the studied approaches are included in an open-source platform that will grant access of the algorithms to researchers from academia and industry. This platform has the purpose of dissemination and future improvement of process operability algorithms and methods. For the development of a versatile process operability tool, the considered approaches are adapted to support process models of generic dimensionalities. New contributions are also incorporated such as the conventional operability index calculation and other methods to find optimal design regions. Nonlinear energy systems of different input × output dimensionalities (6 × 3, 7 × 3, and 3 × 3) are addressed to demonstrate the applicability of the features available in the developed software tool to process achievability analysis, intensification, and modularization.

Operability and control in process intensification and modular design: Challenges and opportunities

“…Operability describes in general the inherent ability of the plant to perform satisfactorily under conditions different from the nominal design conditions, which can include the requirement to adjust to changes in operating conditions, disturbances, product specification, and so on 50,51 . Extensive efforts have been made in the PSE community over the past several decades to quantify process operability performance from different aspects as summarized below (see detailed reviews 11,40,50,82 ): Steady‐state and dynamic flexibility analysis — to determine the maximum uncertainty/disturbance that can be handled at the nominal operating point in a fixed design configuration 83‐85 Steady‐state and dynamic controllability analysis — to determine the “best” dynamic performance achievable for a system under closed‐loop control 48,51,86 Operability‐based analysis — to determine whether a controller can perform its mission under the constraints posed by the existing input variables and desirable output variables using linear/nonlinear input/output mapping methods 31,87 …”

“…Stepping to the equipment‐based steady state and dynamic design, significant progresses have been made in the PSE community to explore design under uncertainty under multiple time scales. Detailed reviews can be found in Yuan et al 97 and Burnak et al, 82 while an indicative list is given as follows: Steady‐state flexibility with multiperiod design — for example, heat exchanger network, 83 mass and heat exchange network 98 Dynamic flexibility with multiperiod design — for example, distillation column, 99 air separation system 100 Design with controllability considerations — for example, reaction‐separation system, 101 divided wall column 102 Operability‐based design — for example, membrane reactor 31,103 Integrated design and control — for example, combined heat and power system, 104 reactive distillation 92,105 …”

“…Current efforts toward process intensification and modularization through PSE approaches can be generally classified as the following: Design and optimization Process design, synthesis, and intensification — especially highlighting the phenomena‐based approaches to generate innovative process solutions 35‐37 Optimization of modular production — to compare the efficiency, economics, and/or reliability of numbering up vs. scaling up 12,38,39 Optimal operation and control Operability analysis — including steady state and dynamic feasibility, flexibility, controllability analysis, and so on 31,40 Advanced model‐based control — especially highlighting the need for distributed and decentralized model‐based control in intensified and modular systems, as well as integrated considerations of design, control (and scheduling) 22,41 Dynamic/periodic operation — to purposefully perturb operation along an optimal trajectory by exploiting process nonlinearity and output multiplicity 42,43 Supply chain optimization and operational planning of modular production — especially highlighting the flexibility and mobility of modular processing 44,45 …”

“…Optimal operation and control Operability analysis — including steady state and dynamic feasibility, flexibility, controllability analysis, and so on 31,40 Advanced model‐based control — especially highlighting the need for distributed and decentralized model‐based control in intensified and modular systems, as well as integrated considerations of design, control (and scheduling) 22,41 Dynamic/periodic operation — to purposefully perturb operation along an optimal trajectory by exploiting process nonlinearity and output multiplicity 42,43 Supply chain optimization and operational planning of modular production — especially highlighting the flexibility and mobility of modular processing 44,45 …”

Dynamic and Statistical Operability of an Experimental Batch Process

“…Past operability studies have also considered the application of operability to control strategies [4,5], process intensification [6,7], and design optimization [8,9]. The operability framework has been extensively studied from a steady-state point of view [10], in which the graphical regions used consisted of mainly stable points of operation [11,12].…”