Pitfalls in Simplification of Dynamic Simulation

by | Apr 15, 2014 | Simulation

Mart Berutti

Vice President, Process Simulation, Emerson Automation Solutions

The selection of the correct level of model complexity or simplicity is an important part of every dynamic simulation project. Two experts in process modeling, Zach Sample and Don Sengur, have teamed up to discuss the dangers in misapplying simplification to building dynamic simulation.

Don Sengur, Lead Project Consultant, and Zachary Sample, Simulation Project Engineer

“Over the years, our engineers have become well practiced in process analysis and translation into simulation models. As with all engineering principles, process simplification is a necessary tool in simulation used to quantitatively describe the complexities of a real phenomenon. While simplification can be advantageous in process simulation by improving maintainability or reducing project completion time, more sophisticated processes may not be easily simplified. As modern technologies become common in industrial processes, this issue is compounded by an influx of clients with sophisticated or specialized unit operations. While simplicity is favored in practice, simplification is limited in complex systems due to an inability of simplistic models to capture the complexity of real process behavior and control response.

Examples of this practice have been encountered repeatedly in simulation development. For example, batch processes such as bioreactors and pulp digesters not only require rigorous material and energy balances for the process to reach a realistic target value, but these batch operations also require the system to respond with precise timing and process dynamics in order to appropriately capture the process behavior without tripping safety alarms or failures. Time sensitive processes and controls such as these may require a more complex model to simulate reality.

Processes such as those found in coal liquefaction that have highly endo/exothermic reactions or short residence times present their own challenges. These processes tend to have drastic fluctuations of flow, composition, pressure or temperature over shorter periods of time than the normal simulation execution time. In these cases, using a normal execution time could cause the simulation of these processes to be chaotic and uncontrollable. While it may be tempting to simplify the model to either avoid the need to increase the model execution speed or prevent overloading of system resources at faster execution speeds, a simplistic model may be unable to reflect the critical behavior of the real process. In order to appropriately simulate these fast processes, a sophisticated simulation model may be necessary.

While the complexity of sophisticated systems such as those containing bioreactors or hydrocrackers can be easily anticipated, processes that are seemingly easy to develop can be quickly complicated upon integration with the control system. For example, a simplified conveyor model, used extensively in mining industry, can be created with a number of fairly straight-forward approaches. However, the real-world conveyor introduces a large, variable process dead time in addition to conditional behaviors during a loaded startup and shutdown. Since these behaviors are accounted for in the control system, the modeled conveyor performance must be built with enough complexity to match reality. Otherwise, the control system may be unable to respond appropriately causing potential system instability or failure of associated equipment such as valves, pumps or compressors.

Complex unit operations like columns may be solved with simplified models like the one shown above or could require more complex models.

An important consideration in model simplification is that model complexity does not necessarily correlate to configuration difficulty. In fact, packaged models of greater complexity and resolution are often significantly easier to configure and integrate into the simulation. This has been made evident by recent interns that have been able to model complex systems using packaged models such as Mimic ‘s boiler object after just a few weeks of modeling and training.

This is not to say that model simplification is not beneficial in practice. In fact, simplification is necessary to meet deadlines and budgets when modeling complex processes. Furthermore, model simplification is not limited to process boundaries or unit operations with limited impact on other process equipment or controls. When careful consideration has been made to avoid potential pitfalls, simplified models can be created to simulate major process operations. However, if the consequences of oversimplifying of a process is not considered prior to model development, the amount of work required to address issues with control system integration may greatly surpass the effort required to create a complex model.”

I look forward to your comments, questions, or suggestions.

Hope to hear from you soon.




Popular Posts


Subscribe for Updates

Follow Us

We invite you to follow us on Facebook, LinkedIn, Twitter and YouTube to keep up to date on all the latest news, events and innovations to help you take on and solve your toughest challenges.

Want to re-purpose, reuse or translate content?

Please do, Just link back to the post and send us a quick note so we can share your work. Thanks!

Our Global Community

Emerson Exchange 365

The opinions expressed here are the personal opinions of the authors. Content published here is not read or approved by Emerson before it is posted and does not necessarily represent the views and opinions of Emerson.