Аn IPCOS client, a company with more than 50 plants mostly in specialty chemicals, was struggling with the vast variability of how PID loops were operated and tuned between plants. The customer did not have standard tools or technology to monitor the performance of their control loops.
IPCOS was asked to provide a solution using standard, cloud-native technology. Our client chose a cloud-centric approach for its digitalization built on the Microsoft Azure platform. The solution provided the Central Data Scientists with a platform to move from ad hoc analytics and insights to company standard analytical methods leveraging the vast amount of data in the data lake.
Let us take you along the process of integrating control loop performance monitoring for this client in the specialty chemicals industry.
Often centrally steered programs end up being a burden for the local sites involved. Flexibility and custom site-specific technology solutions are sacrificed for standardization and centralization, hindering the site local engineers to get things done.
This time it was different. One of the site managers of our client cheered:
“This is the first time the central platform is giving us something back that can be very useful. In the past, we have spent tremendous efforts supporting the centralization effort with no real tangible benefits on-site.”
Our solution consists out of two integrated applications:
1. a plant-facing application built to engage the different stakeholders on site around the subject: product engineers, DCS programmers, site managers, process control engineers etc.
2. a corporate-facing dashboard application developed to provide management a high-level picture of the different sites related to control loop performance status and overall plant operations.
Requirements of the plant-facing solution were that it should provide useful metrics, give highlights and pinpoint changes over time. The tool that we implemented became a real game changer for the engineers involved. It allows for them to take immediate action based on an interface that connects and visualizes process data, DCS configuration data, and various other data sources. We chose simple and easy-to-understand KPIs for control loop performance monitoring, such as: How many times did the controller switch mode? What is the controller error variability? Is the controller output limited, etc.?
Moreover, building this tool in a cloud ecosystem provided agility and flexibility to all stakeholders involved: IT, OT, data scientists, on-site engineers as well as corporate decision-makers.
After some initial challenges with setting up the data ingestion and the access control, we managed to develop a minimum viable product, adhering to the existing security processes, within six months.
In the next phase (solution scaling) we rolled the tool out to all the different sites in again six months’ time.
Currently the client is already using the tool, receiving better insights on how their control loops are performing. At the same time, we are continuously improving the system and engaging the sites to participate in the adoption stage of the project.
Our solution allows for deployable interactive analytics visualizations that will span beyond the project’s current scope and will enable other advanced analytics initiatives e.g., alarm event analytics, heat exchanger fouling predictions, rotating equipment monitoring, etc.
Would you like a success story like this as well?
Get in touch with one of our digital strategy experts at IPCOS
and request a free consultation.