How to improve process integration with dynamic simulation Integration of process design, control and operation
Today, the simulation of process dynamics is bringing us as close as possible to the real process. With dynamic simulation, it si possible to integrate process design, process technology, safety, process control, control strategy and process operation into one discipline. The dynamic simulation represents knowledge synergy of all these disciplines and the basis for different kinds of analysis, optimization, training and education. Its development employs a huge amount of knowledge, but it gives even more knowledge back to users.
SimulateLive will be covering all the most interesting topics of process dynamic simulation, bringing to our reader practical stories from the industry.
Dynamics vs. steady state
First of all, as the introduction to dynamic simulation, let’s clear some of the most important terms.
By the state of the process, mathematical models are divided into two groups:
- Static, and
- Dynamic model.
A static model is describing steady state condition and ignores changes of process variables with time while dynamic models are describing unsteady state and defining relationships between the variables as they change with time.
Dynamic simulation of a chemical process is a crown of chemical engineering and represents the most advanced discipline of the simulation in general. When describing more complex processes, even a steady state simulation can be very challenging because of convergence and calculation stability problems and can easily cause headaches for the user developing it. Since dynamic simulation is adding operation and control as the variables of time to the design steady state base, it is significantly more complicated to handle both modeling and calculation challenges. This also requires more skills and knowledge from the user, who can more likely end up with unstable simulation and unreliable results.
Most simulation programs use steady state simulation as the basis for dynamics. For a steady state model to be able to represent the basis for a dynamic simulation, it needs to be defined in more details than when simulating steady state only. Because, in addition to the variables used in the steady state, there comes a time as variable and all parameters influenced by time need to be defined. Examples are vessel dimensions and geometry, valve characteristics etc.
Control parameters for every PID loop also need to be defined when simulating process dynamics. This opens area of process disturbance and instability and brings the user to a situation to control both– the simulation and the process.
Dynamics vs. steady state on example of a vessel
Let’s try to see the difference between a steady state and dynamic model at an example of a simple vessel. When defining a vessel in a steady state model with typical simulation software, usually there is only a simple specification which needs to be done. Sometimes even none. Because, the vessel model will use definitions of inlet streams (e.g. temperature, pressure, flow). There are also cases when just one variable which needs to be defined, e.g. pressure.
But, when looking at dynamics of the vessel, there are more parameters that need our attention:
- First of all – dimensions and geometry! Because, with the time as the variable, there is also accumulation involved. So, there and parameters which need to be defined to understand how fast or slow the vessel is being loaded or unloaded – and those are related to size and geometry of a vessel.
- Basic control. There are a few PID loops without which the vessel cannot operate safely. There would be at least two PID loops: level and pressure control. This means that at least two valves and PID controllers need to be specified together with all related parameters. If temperature or flow control is existing on the vessel, other PID loops need to be specified too.
- Solving. Solving the typical vessel in steady state is most often just a mouse click away. But with dynamics included, it can be a fun and challenging exercise. Since in dynamics you have to be able to take care of a simulation and to control the process, which means that you have to synchronize all the unsteady operation which PID loops are bringing to the vessel. So, now you will be able to practice both simulation and process control which, once you figure out all the interactions, can be a fun game.
This is just a simple example of a vessel in steady state and dynamics. From the example, it is quite obvious that the development of the dynamic simulation of an e.g. column is a real game.
From the described example, it is clear that an integration of different knowledge is employed when developing a dynamic model. First of all, you need to have the knowledge of the process and basic process operations. It is the most important ingredient to integrate and with it, it would be much easier to solve any problem which will for sure occur. Some mathematics and engineering logic will be good as well. You will have to know at least some basics of a process control. Then, having the right software and - you’re good to go! SimulateLive will be there as well....
To be able to use already a developed simulation – you can just plug and play with it. Just make sure to have a back-up of the steady state.