I recently saw the movie “Avatar”, with all its 3-D graphical splendor, and an idealized world of happy people communing with nature on the forest-world of Pandora. Knowing that practically the entire movie was computer-generated, I was reminded of the role that simulation plays for process control.

In an effort to be a better process control guru, we should all make an attempt to understand how things “should” work. And for that, simulation is a terrific tool. There are many types of simulation tools available, and you should choose one that fits your specific needs.

First, decide if you need a static or dynamic simulator. Static simulators handle steady-state conditions, and can sometimes be accomplished within an Excel spreadsheet. But most control problems will require you to deal with dynamics. This will enable you to study how controllers respond to changes and manage to get the process in control over time. Here’s a link to a simple dynamic simulator: http://www.simapp.com/simulation-software-description.php

Next, you can determine if you are going to develop your own process models, or if you need pre-defined models for complex processes, like distillation columns. For years, students have been developing their own models using software like MATLAB. Industrial versions of MATLAB adds more pre-configured simulation object libraries.

The next step is to determine if you need the simulator to include a physical properties database. With this feature, the simulation can automatically determine boiling points and even chemical reactions without special programming on your part.

Lastly, determine whether the simulation will run stand-alone, or if you need it to connect to something in real-time. The latest trend in simulations for operator training is to connect the simulation to a “virtual control room”, allowing the operator to practice on your simulated plant, using an HMI that looks exactly like the real plant.

As you add complexity and fidelity to a simulation, the cost can rise dramatically. A simulation of a full plant, integrated to HMI, with training materials can reach the $1 million mark. With such a simulation, operators can be trained to face both normal and abnormal operating conditions, reducing the risk of errors in the “real” plant.

Just like in Avatar, a good simulation starts to blur the lines with reality, eventually mimicking the real world with such accuracy that it can be hard to tell the difference. In the end, here’s my advice: every process control engineer should do SOME simulation. Whether you are building a high-tech operator training room, or simply predicting a control loop’s recovery time, simulations can provide a better picture of how the world “should” be.

Advertisements