Process Modeling
This section explains the model structure on which IPSE GO process models are based. It gives an overview of the different model types that are available and explains how they are related to each other. In this section you will also find information about how component models are combined to model a complete process.
What is a Process?
Engineers frequently have to set up and solve mathematical models for systems that can be structured as a network of components.
An example of such a system is a power plant. A power plant is composed of many components, such as turbines, pumps, heat exchangers, etc. These components are connected by streams which carry the working fluids and by shafts.
In general, what appears to be a single component is again structured in the same way as a process model. A turbine, as mentioned in the previous example, can itself be decomposed into components like valves, guide vanes, stages, etc. It depends on the level of detail you use when looking at a process to define the components which make up the process.
In IPSE GO such a structured system is called process. Hence, a process is a system that has the following characteristics:
-
A process is composed of one or more objects or process components.
-
The objects are connected in a defined way.
-
The behavior of each object can be formulated mathematically.
-
The overall behavior of the process is determined by the behavior of the objects that compose the process and by the connections which link these objects.
The components which make up a process are more generally called objects.
| Notice that the objects are not necessarily pieces of equipment. They can also represent connections among components or chemical compositions, for instance. |
In summary, in IPSE GO a process is any system that can be structured into a network of discrete components, whose operation causes some data changes.
This general interpretation of the term process goes far beyond the definition that is used in chemical engineering, for instance: A process is any operation or series of operations that causes a physical or chemical change in a substance or a mixture of a substance [1].
A process in IPSE GO is not necessarily related to changes in a substance. The changes are not limited to physical or chemical changes. They may affect any kind of information that can be represented numerically. Such information can be, for example, geometric or financial data. Hence, the term process in IPSE GO is not based on underlying physical or chemical phenomena, but on the structural characteristics of a system. Mathematically, a process is described by a system of nonlinear equations. This system can consist of several thousand variables.
Model Structure
In IPSE GO, a process model is created using component models from a model library. Component models are mathematical descriptions of the behavior in terms of equations and different other items like variables and parameters. Objects which are based on specific component models are created and arranged appropriately to form a process model.
| This section describes the model structure on a general basis. The units, connections and global models available in your project depend on the specific model library that you are using. Check your library documentation for more details. |
From Separate Objects to the Process Model
Mathematically, an object is a set of equations and variables. Setting up the mathematical model of the process means to combine the equations and variables of all objects into one single system of equations which represents the process model. This is done automatically by IPSE GO while you connect component models on the flowsheet.
The following example illustrates how the component models are structured so that it is possible to combine them into a system of equations:
Consider two pipes A and B as shown in Figure 1. The pipes are connected and a mass flow \(m\) exists through the pipes. To satisfy the continuity principle, the mass flow entering a pipe must be the same as the one leaving the pipe:
\(m_{A, in} = m_{A, out}\) |
(2.1) |
|||
\(m_{B, in} = m_{B, out}\) |
(2.2) |
Furthermore, the mass flow leaving A must be the same as the one entering B:
\(m_{A, out} = m_{B, in}\) |
(2.3) |
These four variables are displayed as case 1. If equation (2.3) is used, it is not possible to represent the object as a set of equations that is independent from other objects in the system. However, this independence from other objects is essential if process models are created from predefined component models in a library.
It is possible to avoid equation (2.3) by guaranteeing that \(m_{A, out}\) and \(m_{B, in}\) are the same variable, which automatically satisfies (2.3). This is achieved by introducing streams, additional objects that are used to connect the pipes. The mass flows are variables of the streams (\(m_{1}\), \(m_{2}\), \(m_{3}\)). This is shown as case 2. Using the variables from the streams, the equations of the pipes are now:
\(pipe A: m_{1} = m_{2}\) |
(2.4) |
|||
\(pipe B: m_{2} = m_{3}\) |
(2.5) |
The pipes are now referencing the streams in order to use their data. Although the equations are still using variables from other objects, the referenced object is now well defined without limiting the flexibility.
To use the concept explained in the previous example, IPSE defines three types of models:
IPSE restricts how objects of different component models can reference each other. In consequence, IPSE establishes a hierarchical structure where objects can only reference other objects of a lower hierarchical level, as shown in Figure 2.
Graphical Representation
IPSE GO uses the concept of a network structure also to represent a process graphically. This simplifies the interaction, since this method of representing a process is commonly applied, especially in non-computational approaches.
The unit models from a model library are represented by graphical icons. These icons can be placed on the drawing area and be connected to build up a process scheme. Such a scheme which describes how matter is transferred between the different units of a process is named the process flowsheet. To have more flexibility in drawing the flowsheet, for some units more than one icon is available. Yet what matters when it comes to system solution are the mathematical equations that stand behind each unit icon. For each unit at least one mathematical model is defined. If a unit contains different models, for example a design and an off-design model, IPSE GO allows switching between these models without having to change the unit icons. See Advanced Data Management for more details.
Model Libraries
IPSE GO stores all information about models for units, connections and globals in a model library. Model libraries are created and maintained using the Model Development Kit of IPSEpro (MDK). MDK compiles the models into a binary format. IPSE GO includes a processor that evaluates the instructions that are required to solve the model equations.
To use IPSE GO for setting up a process model, you must have the appropriate model library available. IPSE GO itself is the environment that is required in order to use an existing model library.