Greg: In September, I talked with Marsha Wisely, president of PlantWise Industrial Consulting, about some of the value of process simulation. In that article, we explained that you don't always have to go with a complex or high-fidelity model to get value. However, your process simulator must match production, so any testing, training and optimization of the system is relevant to your current production.
Throughout my career, I adopted and advocated for process simulation, authored books, and improved many industries by deploying process simulation. Yet, despite the value that the automation industry has seen for years, it’s challenging to maintain simulations.
Get your subscription to Control's tri-weekly newsletter.
This month, Marsha and I explore dynamic process simulation system maintenance. Her experience spans more than a decade working as a process simulation engineer, automation engineer and consultant on process simulation projects. Her leadership at PlantWise gives her unique insight that we’ll tap into this month.
Maintenance is often seen as one of the key barriers to adopting process simulations. With all the innovation and transformational technology today, why don’t simulation systems automatically sync with production?
Marsha: Fair question. New tools are starting to be developed to identify discrepancies between simulation and production, which is great. But these tools compare the emulated control system to the production control system. They don’t look at the process model.
Greg: So, it’s getting easier, but it’s not perfect.
Marsha: Before we give up, let’s remind ourselves why we might want to deploy a process simulation in the first place. The value of process simulation is it’s a virtual replica of a control system connected to a virtual process. With this combination, we have a sandbox that lets users perform automation testing, training and optimization in a safe, offline environment, reducing the risk to production when changes are made. To realize this value though, you have to change the offline environment first.
Greg: If this isn’t what process and automation teams have been doing, they may not think to use process simulation.
Marsha: Exactly. If you deploy a process simulation but don’t change your workflow, you’ll still make changes to production, and then force the simulator to match. With this legacy workflow, you don’t see any value from the simulation, and it would seem like a burden. But really, you havn’t fully adopted it yet. Greg, you were a process control engineer, who had great success using process models. Can you describe your workflow, and how using process simulation enhanced your projects?
Greg: I first developed and ran the model with all the process and instrumentation dynamics and advanced regulatory control strategy on my computer, which nowadays is a digital twin. I found and verified process control improvements. They were incorporated into the plant dynamic simulation that was used, not only for training operators, but also for educating process, automation and maintenance engineers. If the digital twin and actual control system had the same setpoints and tuning, model fidelity was seen in how well the manipulated variables in the digital twin matched those in the actual plant. Process model parameters were adjusted to improve the match.
Marsha: That’s a great example of getting the full value of process simulation. By embedding process simulation into your workflow, your testing was enhanced, and you reduced the risk to production when deploying your optimization project.
Greg: These systems still fall out of sync. Why is that?
Marsha: Depending on staffing at a facility, it may be difficult to justify spending time learning a new workflow, when the old workflow seems so much faster. This is why I advocate right-sizing simulation for where you are in your digital transformation journey. It’s much easier to make small adjustments than big adjustments. Simpler, lower-fidelity, process-simulation systems are easier to learn.
Greg: You start by adopting lower levels of stimulation, adding to them, and enhancing where you find you’ll see the most value. This lets teams adapt their change-management process with a simulation that’s easier to maintain. It also forms good testing habits in the process-simulation environment. Once you have that workflow ingrained in your organization, teams can evaluate where to get the most value for higher fidelities in the relationship, and strategically expand the scope.
Marsha: The change in workflow isn’t complicated. It’s just a change that needs to be fully adopted. The end users I work with in the pharmaceutical industry, where qualification and validation are already a part of their process, see tremendous risk reduction from enhancing their testing protocols. You must be consistent though, including when you hire contract engineers to perform work in your facility. I’ve seen contractors remove simulation from scope to keep their bid lower than the competition. If they don’t update your simulation as part of the automation project, you end up having to reconcile as a separate project later.
Greg: As an automation and simulation engineer, who has worked with contractors, I wanted to rigorously test someone else’s work against my simulation to ensure it meets the specs. It helps keep everything in sync, and I gained confidence in their deliverables.
Marsha: Greg, as a successful process simulation adopter and longtime advocate, I’m curious what skills you think someone needs to be able to own and manage a dynamic process simulation system.
Greg: Think through the components. You have an emulated version of the control system that’s connected to the process simulation. The control system is identical to the production system, so making changes needed in the control system requires knowledge of it. If you model your process so it reflects production by using a first-principles model, with all the thermodynamic calculations and material balances and energy balances and reaction kinetics, then updating that process model requires deep process knowledge.
Marsha: I agree, and that may mean you need collaboration between process and control to update the dynamic process simulation system at your facility. This again shows why it's so important to have a workflow that supports the maintenance and use of your process simulation. You must work across teams to ensure everything is up to date, unless you have a Greg McMillan on staff, who has both skill sets and is a process control specialist.
Greg: Well, I’m certainly not alone. There are consulting companies, who have the combined skillset to participate in your capital and maintenance projects, and manage the process simulator scope, so you can use it for testing and training prior to changing production.
Marsha: By having an external contractor own maintenance, your process simulator isn’t competing for time with production. Process and automation engineers can focus on day-to-day operations, and contractors can be integral to onsite change-management.
Greg: Of course, hiring a contractor has costs associated with it. However, having a few hours on each project and getting maximum value out of this tool you’ve invested in will ensure your simulation is always ready for the next project, and you’re maximizing its return on that investment.
