Chemical makers are realizing real benefits from virtualization, an approach to automation that separates the physical and functional aspects of a control project into independent hardware and software activities. Implementation of automation projects can go more quickly, safely and efficiently, report BASF and vendors such as Siemens, Honeywell Process Solutions and Emerson Process Management.
However, using virtualization technologies can pose a number of challenges, cautions Michael Krauss, senior automation manager at BASF, Ludwigshafen, Germany.
“Cost is certainly one, along with time and efficiency, because it can be quite a lengthy process with today’s simulation products,” he says.
Engineering efforts are high at the beginning of a project and also during its lifecycle, he explains. For example, keeping the simulation of a plant up-to-date is a big challenge — doing that might require up to one full person per plant.
In addition, setting up a project simulation requires significant preparation time. Often the model might not be fully ready at the time of plant startup, especially if project details change — as often happens — along the way. “So you do need to set an earlier project freezing date for the simulation,” he adds.
On the other hand, given well-trained operators, virtualization technology helps to keep a plant online and optimized. This is very important from the point of view of overall efficiency because, for example, it can reduce the amount of off-specification product. “It also allows us to test certain things in advance using the simulator. For example, if high fidelity is in place on the simulator, even very complex processes can be modeled. So it’s a good test bed for your plant optimization work and means you can potentially do more optimization, too,” Krauss says.
To further develop its virtualization capabilities, BASF has been working with Siemens, Dusseldorf, Germany, and design, engineering and construction company M+W, Stuttgart, Germany, on how to make commissioning even safer and more efficient — while at the same time exploring new ways to test process variations without risk or disruption to operations.
The two-year project involved a pilot plant for the batch manufacturing of plant protection products at Ludwigshafen.
“BASF wanted to test whether simulation has additional value when it comes to software quality and length of commissioning,” notes Ute Forstner, marketing manager chemicals, Siemens, Dusseldorf.
For the pilot, BASF used a simplified cold commissioning model, i.e., it only simulated water flowing through the plant. “This simplification resulted in a drastic reduction of the simulation effort but still helped us to perform more tests than in a traditional FAT [factory acceptance test]. For example, we could test batch sequences running in parallel and interlock strategies. In both areas we were able to detect errors which otherwise would have to be fixed during the site commissioning,” explains Krauss.
The project, because it involved migrating from one control system to another, didn’t require the lengthy startup associated with a greenfield project. “Our main concern was keeping the automation system free from engineering errors because the new system basically had to do the same as the old system; if the errors weren’t found early enough, they would delay startup. However, operations achieved startup very quickly, days faster than expected,” he adds.
“The goal was achieved, with commissioning in under a week rather than the typical 3–4 weeks or longer, so the customer could start production earlier. Software quality was extremely high, with no further bugs found,” notes Forstner (Figure 1).
With the pilot work successfully concluded, BASF now is using the experience in its internal processes by screening projects in terms of potential benefits versus the costs involved.
BASF’s long-term vision for virtualization technologies focuses largely on increased efficiency through better integration between the engineering and the simulation interfaces. With a fully integrated tool like this, the simulator might even auto-change when engineering data are updated. “This, in turn, would reduce the costs and efforts involved in simulator setups. So, this would also make simulation much more accessible. With the advances of technologies such as the Internet of things, the need for digitalization will only grow and lead to a higher need for simulation,” Krauss adds.
Siemens has several projects, not all pilots, currently underway with other chemical companies, reveals Forstner. The emphasis of these is two-fold: first, to generate higher-quality data sooner in the project phase to reduce the duration of commissioning; and, second, to train operators much earlier before the start of plant operations.
On a broader industrial scale, the company installed 80 virtualization systems in November alone.
Its “Simatic Virtualization as a Service” aims to give easy access to advanced virtualization technology and includes setup of the virtualization server, configuration of virtual machines, installation and configuration of the operating system, turnkey installation of the Simatic software, plus technical support for all of the system components.