Topic Centers

Diversified Industrials: Horizons

Digitalization. Complex supply chains. Growing regulation and compliance. Conglomerates and industrials operate in an increasingly dynamic environment – one which requires creative solutions and forward thinking.

Predictive maintenance: opportunity or dead end for the chemical industry 4.0?

By Tobias Faber and Johannes Groß

Data, machine-to-machine communication, and the Internet of Things (IoT) are buzzwords describing the fourth industrial revolution. In the past few years, there have been claims that the chemical industry is falling behind on digitalization issues. The question then arises as to how the chemical industry can succeed in the transformation to 4.0? A first step to be taken in this process could be the implementation of predictive maintenance models into the production cycle of a chemical production plant.

Case study

The production line of a chemical plant uses diverse raw materials. These raw materials should be transported via pipelines and combined in precisely weighed quantities at the right time in order to form an individual product. An empty storage tank of the materials can have severe impacts on the entire production process. Defects or missing resources can only be detected retrospectively in the event of a “sudden” failure. The resulting downtimes could often entail considerable additional costs for the operator of the chemical plant. Consequently, “traditional” maintenance comes either too early (i.e. the components do not yet require replacement) or too late (the component is already worn down).

The more efficient way – predictive maintenance

Predictive maintenance follows a different path. Unlike reactive maintenance, predictive maintenance is carried out “on-demand” based on the real-time performance data of the components.

The chemical plant is equipped with sensors (so-called cyber-physical system chips) that are connected to a local network or the internet. Once this data is analyzed, the operator of the chemical plant can create a real-time representation of the chemical plant (digital twin). If the data is analyzed over a certain period or compared with data from other plants with identical components, the operator can make predictions about the remaining lifetime of a specific component. Predictive maintenance offers as key innovation that a component can be maintained “precisely” due to its maintenance need based on the available real-time data. This can reduce the overall maintenance costs of the chemical plant as well as increase the efficiency of the plant.

Infrastructure as a service

The operator of a chemical plant can subcontract the analysis of the data and the maintenance of the chemical plant to a IoT-service provider. The producer of the machines in particular is to be considered as such IoT-service provider, given the existing valuable know-how from the production of the machines used in the chemical plant. If the IoT-service provider is subcontracted for more than one chemical plant, the combined data of all chemical plants could assist in improving the quality of the predictions even further. If they feel comfortable enough, the IoT-service provider could even consider offering their machinery to the chemical plant as a service. The IoT-service provider is then responsible for the “provision and operation” of the plant and thus bears the risk of defects.

Challenges of implementation and operation

One of the main challenges of implementing predictive maintenance into the operation of a chemical plant is the large volume of data created and obtained in real-time that can become almost unmanageable. This data needs to be of high quality in order to be analyzed. Storing, updating, processing, and transferring this data is therefore a central challenge for predictive maintenance applications and requires a stable and powerful network connection.

In addition, a reliable protection against cyber-attacks becomes increasingly important. Successful past cyber-attacks have shown that companies should not underestimate this danger and should implement state-of-the-art standards and preventative measures.

The successful implementation of predictive maintenance business models is a headstone of IoT and digitalization in the chemical sector. With the right know-how, it is – despite initial costs – possible to exploit great potential for added value. Innovative companies can play an essential pioneering role in this regard and define standards themselves, thus helping to shape the future of the chemical industry.

Loading data