Training in process industries: immersion is the word. You train personnel by immersing them in the operations that they must master. Earlier methods, such as on-the-job (OJT) training and old-school methods—lectures and tests—have worked for centuries. But OJT takes time, and unless there’s a direct link between theory and action, old-school may not cut it. As a direct-experience alternative, simulation offers ways to exploit not just conventional repetition for learning but also rich types of action training for specific scenarios.
At its simplest, simulation is play-acting. In matters of real consequence, such as learning what actions can result in destruction or death, simulation’s chief benefit is that it’s not reality. You can crash a flight simulator all day without fatalities or loss of multimillion dollar aircraft. You can close or open a valve and trigger a simulated melt-down/runaway/explosion without having to reimburse anyone in the neighborhood for hospitalization or window glass.
At its best, simulation-based training helps trainees see how unhappy outcomes happen, learn directly how to prevent them, and lock in on how to fix them.
In the process world, simulation first proved its worth not in training, but in research and development. New equipment design now benefits from decades of experimentation, scientific enhancements and increasingly focused—and increasingly sophisticated—applied mathematics. Variables of a process can be fed into systems that simulate outcomes quickly, packing hours of activity into seconds, days into minutes, and weeks and months into a few hours. The result: faster, more focused equipment and procedure designs.
Time compression is one of the benefits of applying simulation to training, and anything that saves time is good. “Our industry is facing training needs like never before,” says Jerry Milheiser, technology training supervisor, Yokogawa Corporation of America (www.yokogawa.com/us). “In the past, OJT training worked well and was fairly cost effective. Unfortunately, it takes several years for OJT to sink in. We don’t have years available to us now—many experienced employees will be retiring near-term, and we don’t have enough time for OJT to become effective. We need operators trained at a faster pace. Simulation-based training helps overcome this time crunch.”
Simulation lets an operator experience how actions play out with as much fidelity you want. In the process world, there are two broad categories, 2D and 3D. Simulation in 2D presents the world of computers and screens (for example, a control room mockup), and the outcomes of a trainee’s keyboard or mouse inputs can be played back at any speed, from ultra-slow motion to ultra-fast forward.
More intense 3D systems strive to mimic the physical forms and attributes of process equipment in action. It allows trainees to view—even walk around—images of the physical plant as the various structures, pieces of equipment, gauges/screens, noises and alarms reflect the training scenario. This type of simulation can range from high-resolution screen representations to 3D video glasses-plus-motion gloves to full-scale physical mockups that can be enhanced with sounds and motion feedback.
2D winning the race
The adoption rate of 2D simulation for training programs in process industries currently greatly outpaces that of 3D simulation. Industry adoption of 2D simulation over the last five years also has been far more than suppliers forecast. “The upswing in
North American capital investments in petrochemical over the last few years has been phenomenal,” says Martin Berutti, president and COO of Mynah Technologies (www.mynah.com). “Oil, gas, mining—it’s all showing great growth. That growth has generated increased demand for training—which in turn has fueled greatly increased demand for simulation.”
“We’re seeing more people interested in 2D simulator-based training than ever before,” agrees Martin Ross, product manager, UniSim Operations for Honeywell (www.honeywellprocess.com). “Simulation is just about the minimum if you want to comply with industry and regulatory standards—it’s become an engineering best practice.”
Some users have standardized globally on simulator-based training; witness ExxonMobil’s 2005 adoption of Invensys Dynsim software and project services to build operator training systems. “They married their training strategies to their global process and business strategies,” explains Tobias Scheele, vice president, advanced systems for Invensys Operations Management (IOM, iom.invensys.com). IOM offers the SimSci-Esscor family of simulation products.
A prime application of 2D simulation is training for control room operations. After all, most of the real-world action there plays out on (2D) computer screens. Many argue that control room operators must therefore learn to see events as they appear on screens—in other words, their training should focus on system dynamics exactly as they would be experienced by those in the room.
A control room is a far cry from a front-line mining operation, however. There, “you’d want a new operator to sit in a mocked-up cab, working it like a flight simulator,” says Honeywell’s Ross. “Classic [2D] simulation is the norm in the control room, since that doesn’t partake of machines in motion. Plus remember, we operate in a conservative industry and 3D simulation is still pretty new.”
To be fair, proponents of 3D simulation point out that a highly effective training regimen can be built on the dramatic presentations of the consequences of control room interventions. The immediacy of three-dimensional representations of a faltering plant and equipment helps drive the point home.
While this approach is often called “immersive” (with the suggestion that the deeper the immersion, the more effective the learning), that term is not a reliable label. Immersion in 3D simulation may or may not be more effective than immersion in 2D.
A conscientious, pressure-packed attempt to manage runaway processes via control-room computer screens may well be a more effective learning tool than an excursion with glasses and gloves when the latter is perceived by trainees as no more than a computer game session. Keep in mind that we’re talking about the need for training the new generation, and that’s the generation most likely to trivialize computer-based game play.
Of course, the intensity of a given training session is a matter of management—a good leader will get trainees immersed regardless of technology type. And to be fair, 3D simulation offers more pressure points and potential triggers to involvement than 2D. Why, then, is the adoption of 3D simulation for training in the process industries so low (see Automation World, March 1, 2013, “The Slow Advance of New Training Methods”)?
We’ve given one reason for the control-room world: operators work in a 2D world. But for front-line, equipment operators, why hasn’t 3D simulation taken off?
Ironically, reality is part of the answer. A major industry pressure point is the need to add people to operate the equipment going into new build-out, whether greenfield or brownfield. Why simulate when the real thing can be explored as it’s being built, without danger of consequences?
“We considered 3D simulation,” says Dylan Charles, a project engineer for U.S. Nitrogen, an Austin Powder Company. “But we decided to dovetail personnel hiring and training with the equipment installation. Our people are in training while the facility is being constructed—so, trainees are seeing a whole plant being assembled in front of them. They’ll be gaining familiarity with operations before any fluid fills the pipes.”
When U.S. Nitrogen finishes constructing its new, high-capacity ammonium nitrate plant in Greeneville, Tenn., Charles will become the facility’s electronic and electrical instrumentation control supervisor.
Granted, this requires exquisite timing—but it’s no more challenging than the timing of the installation itself. “Our implementation and the efficiency of our training still has unknowns,” Charles says. “We won’t have an experienced crew—our labor pool is in industry around the site, which is mainly discrete manufacturing.”
The company has 2D simulation—MiMiC from Mynah—in conjunction with DeltaV-based Emerson (www.emersonprocess.com) controls around the control room. “We’re using that to train board operators on navigating the HMI, handling startup procedures, and on communications between the board and the field operators. We’re able to throw malfunctions into the simulation and see how operators progress with trouble shooting,” says Charles.
The resource issue
The second reason for slow uptake of 3D simulation for training—the level of resources required—may be more perception than reality. “The technology price point is not really the big issue it used to be,” says IOM’s Scheele. “Hardware costs have dropped 80 percent compared to five years ago, and peripherals for virtual reality, multicore processing and 3D visualization are increasingly over-the-counter and cost-effective. We’re seeing significant interest among the early adopters, and I’m convinced 3D will move into the mainstream, but the adoption curve is still in its early stages.”
Honeywell’s Ross agrees: “A lot of people think cost when they think 3D, but cost may no longer be the real barrier. We’ve seen 3D simulation move forward in parallel industries for everything from database mining to virtual plant walk-throughs, and those industries are just as cost-conscious as process manufacturing.”
Ross says Honeywell is actively exploring the technology with 3D simulation software maker Virthualis and “a few early adopters have made the leap. We think it has a lot of applicability. This ranges from physics, for example, simulating the effect of radiant heat from a fire on operators, to incident management, where it’s important to experience how physically accessible equipment is, what it’s like to climb ladders, that sort of thing.”
Yokogawa’s Milheiser brings up an important point: “Hidden or ongoing costs [for 3D simulation] can stress the budget—you can train only a few at a time, and the duration of training can be weeks. Just the cost in manpower for a training cycle of a plant can be a stopper.”
Yokogawa has assembled a multi-tier program that, for any given end-user, may or may not culminate in 3D simulation. The program moves from generic low-resolution modules on plant functionalities (“such as a refinery top side,” Milheiser explains), to plant-specific but still relatively low resolution modules that deliver self-paced training. Whether an end-user’s program moves to higher resolution simulation depends on needs. Candidly, Milheiser says, “We think the best use for high fidelity simulation is in engineering, to test process changes, ingredient reactions and materials or supply changes, as well as long-term viability of proposed capital expenditure.”
All the many talk points aside, training via simulation is not just here to stay, but has become an unstoppable force, whether 2D or 3D. “When knowledge walks out the door—and keep in mind, within three years, half the workforce will retire—20-somethings will have to become instant experts,” says Invensys’s Scheele. “If you want that next operator to know where a certain pump is, simulation gives you the quickest way to put that location into the person’s head.”
It’s much faster than having an experienced mentor, assuming you can find one, walk a new hire around a real plant again and again, Scheele adds.