Process simulation involves the utilization of computer software resources to develop an accurate, representative model of a chemical process in order to understand its behavior in response to different inputs and control. In the past, process simulation was mainly concerned with the development of sophisticated unit operation blocks to predict mass flows of principal components through a process. In recent years, environmental consciousness has led to demands for tracking trace components (for example, resulting from fugitive emissions) that have an impact on environmental health and compliance, as well as major product and process components. Coupled with this demand for higher resolution models is the need for sophisticated computer-aided process design tools to identify low-cost, environmentally friendly solutions in the presence of considerable uncertainty. This calls for an integrated hierarchy of models, including modules with a high degree of detail for individual unit operations and process engineering activities, to simpler modules for analyzing system interactions at higher scales, with material flows and symbiotic interactions often controlled by exogenous factors, market forces or government regulation.
Many industries, both private and public, are involved in the transformation of raw material to useful products and by-products (some of which may be environmentally unacceptable). Several use process simulation tools to model their core production processes. These include chemical industries involved in the processing of organic and inorganic materials, the electric power industry involved in the transformation of fossil fuel to energy, and municipal treatment plants involved in the transformation of dirty to clean water. Effective facility operation is dependent upon accurate process simulation for assessing the material and energy flows through the process, so that the required thermal, environmental and economic performance can be assessed. These same process simulation tools have the potential to address programs and strategies to improve material and energy flows at higher scales of economic aggregation, providing guidance for industry, governments and citizens wishing to improve efficiency, sustainability and environmental quality through pollution prevention, material re-use, waste recycling, and material and energy conservation.
To understand how process simulation is used to model and design complex systems, the key components of a process simulation software package are identified and reviewed. The essential building blocks of a process simulator or 'flowsheeting' package include the following:
• Thermodynamic models: these are models developed to predict the different physical properties of the components under process conditions.
• Unit module models: these are routines that simulate the different unit operations (distillation, mixing, splitting, heat exchange and so on) and processes (reactions, mass and energy transfer, head loss).
• Data bank: the data on component physical properties, reaction rates and cost coefficients.
In addition to these, there are mathematical routines for numerical computations and cost routines for performing an economic analysis of the process.
Process simulation software can be classified as 'sequential modular', 'equation-oriented' or 'simultaneous modular'. Traditionally, most simulators have adopted a sequential modular approach. With this approach, individual modules are developed for each unit operation and process. Output stream values are computed for each, given the input stream values and the equipment parameters. Each unit module in a flowsheet is solved sequentially. The overall flowsheet calculations in a sequential modular simulator follow a hierarchy. Thermodynamic models and routines are at the bottom of this hierarchy, followed by the unit operation modules that perform the necessary material and energy balances, based on the thermodynamic property routines. At the next level design specifications dictate iterative calculations around the units, superseded by the recycle iterations for stream convergence. Program utilities, such as parameter estimation and optimization, occupy the highest level in the calculation hierarchy in the sequential modular framework.
Equation-oriented simulators define and solve a set of simultaneous non-linear equations that represent the process modules, mass and energy balances in the process. Although these simulators are more flexible in terms of information flow, they are more difficult to construct, and it is often difficult to diagnose errors when they occur. The simultaneous modular approach utilizes individual modules for each unit operation and process, as in the sequential modular approach, but attempts to establish a more immediate link among the inputs, outputs and operations of these individual modules. This is accomplished by defining a set of linear equations that approximately relate the outputs for each module to a linear combination of its input values. These equations are solved simultaneously in the simultaneous modular approach.
While efforts are under way to develop and advance equation-oriented and simultaneous modular software systems for education and research applications, most of the currently available, widely applied commercial simulators are sequential modular in nature. However, as indicated in Table 11.1, a significant effort has been made in recent years to develop and disseminate equation-oriented packages. There are no commercial simulators that use the simultaneous modular approach as yet.
Process simulators are also classified on the basis of their level of temporal aggregation; that is, whether the processes being considered are steady-state or dynamic in nature. Accordingly, steady-state and dynamic simulators are both available for modeling continuous processes. The sequential modular simulators shown in Table 11.1 are steady-state simulators. The equation-oriented simulators in the table can be used for both dynamic and steady-state analysis, but are mostly used for dynamic simulations.
The following example illustrates the use of ASPEN, a sequential modular simulator, to model the steady-state behavior of a benzene production process.
Table 11.1 Process simulation tools
Simulation package Type
FLOWTRAN Sequential modular
FLOWPACK II Sequential modular
PRO II (previously PROII) Sequential modular
ASPEN Plus Sequential modular
MODELLA Equation-oriented gPROMS Equation-oriented
The hydrodealkylation (HDA) of toluene to produce benzene is often used as a benchmark for demonstrating chemical process synthesis methods. The HDA process has been extensively studied by Douglas (1988) using a hierarchical design/synthesis approach. The problem presented and solved here is based on the flowsheet structure analyzed by Diwekar et al. (1992), which involved the selection of the flowsheet configuration and some of the operating conditions that maximize profit. The flowsheet for this case study is described below.
The primary reaction of the HDA process is
In addition to this desired reaction, an undesired reaction
2C6H6 o C6H5 + H2
also occurs. These homogeneous gas phase reactions occur in the range of 894°K and 974°K. A molar ratio of at least 5:1 hydrogen to aromatics is maintained to prevent coking. The reactor effluents must be quenched to 894°K to prevent coking in the heat exchanger following the reactor.
The HDA flowsheet is shown in Figure 11.2. In this process, benzene is formed by the reaction of toluene with hydrogen. The hydrogen feed stream has a purity of 95 per cent (the rest is methane) and is mixed with a fresh inlet stream of toluene, a recycled toluene stream and a recycled hydrogen stream. The feed mixture is heated in a furnace before being fed to an adiabatic reactor. The reactor effluent contains unre-acted hydrogen and toluene, benzene (the desired product), diphenyl and methane; it is quenched and subsequently cooled in a flash separator to condense the aromatics from the non-condensable hydrogen and methane. The vapor stream from the flash unit contains hydrogen that is recycled. The liquid stream contains traces of hydrogen and methane that are separated from the aromatics in a secondary flash unit. The liquid stream from the secondary flash unit consists of benzene, diphenyl and toluene. It is separated in two distillation columns. The first column separates the product,
benzene, from diphenyl and toluene, while the second separates the diphenyl from toluene. The toluene is recycled back into the reactor.
Figure 11.3 presents the ASPEN representation of this flowsheet where unit operation blocks, including splitters, separators and reactors, are used as building blocks to track the material and energy streams through the complete process. Material and energy balances are computed around each unit and the system state variables are calculated, including component flows and system thermodynamic properties like enthalpy, entropy and so on, as shown in Table 11.2.
Table 11.2 Sample results for the HDA flowsheet simulation
Unit Operation Block Results
FLASH:2-OUTL (FLASH2): FLASH INPUT STREAM(S): S01 OUTPUT STREAM(S): S02 S03 PROPERTY OPTION SET SYSOPO
Was this article helpful?