Evaluation of historical simulations | Construpedia
Navegación
Evaluation of historical simulations
Introduction
Process simulation is used for the design, development, analysis and optimization of technical processes such as: chemical plants, chemical processes, environmental systems, power plants, complex manufacturing operations, biological processes and similar technical functions.
Main principle
Process simulation is a model-based representation of chemical, physical, biological and other technical processes and unit operations in software. The basic prerequisites are a deep knowledge of the physical and chemical properties[1] of pure components and mixtures, of reactions and of mathematical models that, in combination, allow the calculation of a process on computers.
Process simulation software describes processes in flowcharts where unit operations are positioned and connected by products or information flows. The software must solve the mass and energy balance to find a stable operating point. The goal of a process simulation is to find the optimal conditions for an examined process. This is essentially an optimization problem that must be solved in an iterative process.
Process simulation always uses models that introduce approximations and assumptions, but allow the description of a property over a wide range of temperatures and pressures that may not be covered by real data. The models also allow interpolation and extrapolation ("Extrapolation (mathematical)"), within certain limits, and allow the search for conditions outside the range of known properties.
Modeling
The development of models[2] for a better representation of real processes is the core of further development of simulation software. The development of the model is carried out on the chemical engineering side, but also in control engineering and for the improvement of mathematical simulation techniques. Process simulation is therefore one of the few fields where scientists from chemistry, physics, computer science, mathematics and various engineering fields work together.
Many efforts are made to develop new and improved models for property calculations. This includes, for example, the description of.
• - Thermophysical properties such as vapor pressures, viscosities, caloric data, etc. of pure components and mixtures.
Evaluation of historical simulations
Introduction
Process simulation is used for the design, development, analysis and optimization of technical processes such as: chemical plants, chemical processes, environmental systems, power plants, complex manufacturing operations, biological processes and similar technical functions.
Main principle
Process simulation is a model-based representation of chemical, physical, biological and other technical processes and unit operations in software. The basic prerequisites are a deep knowledge of the physical and chemical properties[1] of pure components and mixtures, of reactions and of mathematical models that, in combination, allow the calculation of a process on computers.
Process simulation software describes processes in flowcharts where unit operations are positioned and connected by products or information flows. The software must solve the mass and energy balance to find a stable operating point. The goal of a process simulation is to find the optimal conditions for an examined process. This is essentially an optimization problem that must be solved in an iterative process.
Process simulation always uses models that introduce approximations and assumptions, but allow the description of a property over a wide range of temperatures and pressures that may not be covered by real data. The models also allow interpolation and extrapolation ("Extrapolation (mathematical)"), within certain limits, and allow the search for conditions outside the range of known properties.
Modeling
The development of models[2] for a better representation of real processes is the core of further development of simulation software. The development of the model is carried out on the chemical engineering side, but also in control engineering and for the improvement of mathematical simulation techniques. Process simulation is therefore one of the few fields where scientists from chemistry, physics, computer science, mathematics and various engineering fields work together.
• - Properties of different devices such as reactors, distillation columns, pumps, etc.
• - Chemical and kinetic reactions.
• - Environmental and safety data.
Two main types of models can be distinguished:
Rather simple equations and correlations where the parameters fit the experimental data.
Predictive methods where properties are estimated.
Equations and correlations are usually preferred because they describe the property (almost) exactly. To obtain reliable parameters, it is necessary to have experimental data that are generally obtained from factual data banks[3][4] or, if there is no publicly available data, from measurements.
The use of predictive methods is much cheaper than experimental work and also than data from databases. Despite this great advantage, predicted properties are typically only used in the early steps of process development to find the first approximate solutions and exclude erroneous paths because these estimation methods typically introduce higher errors than correlations obtained from real data.
Process simulation also encouraged the further development of mathematical models in the fields of numeration and complex problem solving.[5][6].
History
The history of process simulation is strongly related to the development of computing and hardware and programming languages. Simple and early implementations of aspects of chemical processes were introduced in the 1970s when suitable hardware and software were available (here mainly the FORTRAN and C programming languages \u200b\u200bC (programming language)). Modeling of chemical properties began much earlier, particularly the cubic equation of states and Antoine's equation were precursor developments of the 19th century.
Steady state and simulation of dynamic processes
Initially, process simulation was used to simulate steady-state processes. Steady-state models perform a mass and energy balance of a stationary process (a process in an equilibrium state) that does not depend on time.
Dynamic simulation is an extension of steady-state process simulation, in which time dependence is built into the models through derived terms, i.e., mass and energy accumulation. The advent of dynamic simulation means that time-dependent description, prediction and control of real processes in real time have become possible. This includes describing plant startup and shutdown, changes in conditions during a reaction, delays, thermal changes, and more.
Dynamic simulations require longer computation time and are mathematically more complex than a steady-state simulation. It can be viewed as a multiplied repeated steady state simulation (based on a fixed time step) with constantly changing parameters.
Dynamic simulation can be used both online and offline. The online case is the predictive control model, where real-time simulation results are used to predict the changes that would occur for a change in control input, and control parameters are optimized based on the results. Offline process simulation can be used in process plant design, troubleshooting, and optimization, as well as in conducting case studies to evaluate the impacts of process modifications. Dynamic simulation is also used for operator training.
• - Advanced simulation library")[7].
• - Computer simulation.
• - List of chemical process simulators").
• - Process simulation software.
References
[1] ↑ Rhodes CL, “La revolución de la simulación de procesos: necesidades y preocupaciones de propiedades termofísicas”, J. Chem.
[2] ↑ Gani R., Pistikopoulos EN, “Modelado y simulación de propiedades para diseño de productos y procesos”, Fluid Phase Equilib., 194-197, 43-59, 2002.
[3] ↑ Marsh K., Satyro MA, “Integración de bases de datos y su impacto en la simulación y diseño de procesos”, Conferencia, Lake Tahoe, EE. UU., 1994, 1-14, 1994.
[4] ↑ Wadsley MW, “Bases de datos de propiedades termoquímicas y termofísicas para simulación de procesos químicos computacionales”, Conferencia, Corea, Seúl, 30 de agosto al 2 de septiembre de 1998, 253-256, 1998.
[5] ↑ Saeger RB, Bishnoi PR, "Un algoritmo modificado 'de adentro hacia afuera' para la simulación de procesos de separación de múltiples componentes de múltiples etapas utilizando el método de contribución de grupo de UNIFAC", Can.
[6] ↑ Mallya JU, Zitney SE, Choudhary S., Stadtherr MA, "Solucionador frontal paralelo para simulación y optimización de procesos a gran escala", AIChE J., 43 (4), 1032-1040, 1997.
Many efforts are made to develop new and improved models for property calculations. This includes, for example, the description of.
• - Thermophysical properties such as vapor pressures, viscosities, caloric data, etc. of pure components and mixtures.
• - Properties of different devices such as reactors, distillation columns, pumps, etc.
• - Chemical and kinetic reactions.
• - Environmental and safety data.
Two main types of models can be distinguished:
Rather simple equations and correlations where the parameters fit the experimental data.
Predictive methods where properties are estimated.
Equations and correlations are usually preferred because they describe the property (almost) exactly. To obtain reliable parameters, it is necessary to have experimental data that are generally obtained from factual data banks[3][4] or, if there is no publicly available data, from measurements.
The use of predictive methods is much cheaper than experimental work and also than data from databases. Despite this great advantage, predicted properties are typically only used in the early steps of process development to find the first approximate solutions and exclude erroneous paths because these estimation methods typically introduce higher errors than correlations obtained from real data.
Process simulation also encouraged the further development of mathematical models in the fields of numeration and complex problem solving.[5][6].
History
The history of process simulation is strongly related to the development of computing and hardware and programming languages. Simple and early implementations of aspects of chemical processes were introduced in the 1970s when suitable hardware and software were available (here mainly the FORTRAN and C programming languages \u200b\u200bC (programming language)). Modeling of chemical properties began much earlier, particularly the cubic equation of states and Antoine's equation were precursor developments of the 19th century.
Steady state and simulation of dynamic processes
Initially, process simulation was used to simulate steady-state processes. Steady-state models perform a mass and energy balance of a stationary process (a process in an equilibrium state) that does not depend on time.
Dynamic simulation is an extension of steady-state process simulation, in which time dependence is built into the models through derived terms, i.e., mass and energy accumulation. The advent of dynamic simulation means that time-dependent description, prediction and control of real processes in real time have become possible. This includes describing plant startup and shutdown, changes in conditions during a reaction, delays, thermal changes, and more.
Dynamic simulations require longer computation time and are mathematically more complex than a steady-state simulation. It can be viewed as a multiplied repeated steady state simulation (based on a fixed time step) with constantly changing parameters.
Dynamic simulation can be used both online and offline. The online case is the predictive control model, where real-time simulation results are used to predict the changes that would occur for a change in control input, and control parameters are optimized based on the results. Offline process simulation can be used in process plant design, troubleshooting, and optimization, as well as in conducting case studies to evaluate the impacts of process modifications. Dynamic simulation is also used for operator training.
• - Advanced simulation library")[7].
• - Computer simulation.
• - List of chemical process simulators").
• - Process simulation software.
References
[1] ↑ Rhodes CL, “La revolución de la simulación de procesos: necesidades y preocupaciones de propiedades termofísicas”, J. Chem.
[2] ↑ Gani R., Pistikopoulos EN, “Modelado y simulación de propiedades para diseño de productos y procesos”, Fluid Phase Equilib., 194-197, 43-59, 2002.
[3] ↑ Marsh K., Satyro MA, “Integración de bases de datos y su impacto en la simulación y diseño de procesos”, Conferencia, Lake Tahoe, EE. UU., 1994, 1-14, 1994.
[4] ↑ Wadsley MW, “Bases de datos de propiedades termoquímicas y termofísicas para simulación de procesos químicos computacionales”, Conferencia, Corea, Seúl, 30 de agosto al 2 de septiembre de 1998, 253-256, 1998.
[5] ↑ Saeger RB, Bishnoi PR, "Un algoritmo modificado 'de adentro hacia afuera' para la simulación de procesos de separación de múltiples componentes de múltiples etapas utilizando el método de contribución de grupo de UNIFAC", Can.
[6] ↑ Mallya JU, Zitney SE, Choudhary S., Stadtherr MA, "Solucionador frontal paralelo para simulación y optimización de procesos a gran escala", AIChE J., 43 (4), 1032-1040, 1997.