How to implement a successful predictive maintenance strategy
Predictive maintenance offers ever more opportunities for the future in chemical industry. At Sitech, we are fully committed to this trend. We found that big data and real-time monitoring support us in implementing predictive maintenance. But how do you organize such a predictive maintenance strategy? We combine our expertise with engineers and our IT colleagues to implement this process from start to finish for our customers, with the aim of predicting unexpected failures as early as possible.
Reliability Engineers map out dominant failure modes
“The process of implementing predictive maintenance starts with determining the dominant failure modes,” explains Peter Bosmans, Plant Strategy Manager at Sitech. “First, we want to know which assets pose the greatest risk and which failure modes are most common. Once we have this insight, we can use it to determine which assets need to be monitored. To identify the dominant failure modes, we call on the expertise of our Reliability Engineers. They know exactly what is currently going on with all of the plant components, but they also know what has happened over the past 20 years and what the possible causes of failure were.”
Rotating Engineers provide the right sensors
Maurice Steffin manages the Rotating department at Sitech. His team specializes in selecting the right sensors. “We look at what needs to be measured based on the failure mechanisms that our Reliability Engineers have identified for each asset,” says Maurice. “For example, is it the temperature of the installation? Or are vibration measurements of particular interest? In our department, we select the right sensors to effectively monitor these parameters, after which our colleagues in IT Solutions ensure a connection between those sensors and the data center at the Sitech Asset Health Center. From then on, we can measure, analyze, and take the right action when needed.”
“We use this data to develop predictive models that tell us when failures are likely to occur.”
- Nickel van de Mortel, Service Delivery Manager SAHC
Data analysts monitor and interpret data
All of the data is sent to Sitech Asset Health Center (SAHC). This data is used to create predictive models. “We establish links among the vast quantities of data,” explains Nickel van de Mortel, Service Delivery Manager SAHC. “This includes data we receive from the sensors, but also from outside, for example via physical inspections. We use this data to develop predictive models that tell us when failures are likely to occur. Our monitoring dashboards then generate an automatic alarm if the value of a particular asset becomes too high. A signal is sent straight to the operator in the plant, so that they can take action immediately. We don’t just observe this process, but also advise our customers on the right action to take.”
Sitech brings together the know-how of engineers, IT experts and data analysts
Implementing a comprehensive predictive maintenance strategy requires know-how and expertise spanning various disciplines. At Sitech, we bring these specialists together, combining the experience of engineers with the know-how of our data analysts and IT experts. This enables us to give our customers total peace of mind when it comes to digitalizing their maintenance processes, from determining the scope right through to implementation, and from data monitoring to giving instructions for actions. Thanks to predictive maintenance, plants are easier to control, safer, more reliable and more affordable to run as a result of the huge cost savings that can be achieved in terms of maintenance.
