Success Story

Steel Making
AI-POWERED PROCESS CONTROL

Avoiding Slopping (Overfoaming) in Basic Oxygen Steelmaking

In Basic Oxygen Steelmaking (BOS), slopping happens when foaming slag overflows the vessel, causing spills that need cleanup. This leads to metal loss, process disruptions, environmental issues, and safety risks. The equipment must be inspected for damage and repaired if needed before resuming operations. The complexity of the process and the numerous adjustable parameters make it hard for operators to prevent slopping, especially since it can only be detected a few seconds in advance, leaving little time to react.

Customer

One of Europe's largest steel makers, producing a wide range of high-quality quality strip steel products for demanding markets such as construction, automotive, and machines.

Problem

In Basic Oxygen Steelmaking (BOS), the critical problem of slopping occurs when foaming slag exceeds the height of the vessel and overflows, causing spillage that must be contained and then cleaned up.

The result is metal loss, process disruption, environmental contamination and even a threat to employee safety. The vessel and surrounding equipment must be inspected for damage. If any equipment has been compromised, repairs must be made to ensure that it is safe to resume operations.

Because basic oxygen steelmaking is a complex process with a large number of adjustable parameters, it is very difficult for even experienced process operators to understand and counteract the various factors that lead to slopping, especially since impending slopping can be anticipated only a few seconds in advance, leaving too little time to react.

Solution

The AI engine behind Process Booster, aivis®, was fed the raw historical process data of one BOF whereas the signal of a photo sensor could be used as marker to identify the slopping within the data. aivis® then performed a Root Cause Analysis (RCA) including a report that classified all slopping occurrences into two distinct scenarios. Each scenario told its own story of how the incidents built up over time, which of the hundreds of sensor signals were relevant, and how their behavior changed over time. This allowed the process engineer to interpret and understand these scenarios and ultimately develop countermeasures.

Outcome

Based on the insights provided by aivis® AI and the prediction models it created, the ability to proactively identify and resolve slopping events was increase by 54% with warnings 1-2 minutes in advance. This allowed for a major shift from reactive to proactive actions.

Excerpt from an aivis® root cause report investigating slopping in a BOF. The report reveals the most important influencing signals as well as their characteristic, unhealthy behavior prior to the disruptions.

Based on the insights provided by aivis® AI and the prediction models it created, the ability to proactively identify and resolve slopping events was increase by 54% with warnings 1-2 minutes in advance. This allowed for a major shift from reactive to proactive actions.

Impact

€ 9 Mio.

COST SAVINGS*

* Assuming rollout on 5 lines, production slowdown for countermeasure is 0.8% and 1.5t steel is lost during slopping event.

Unlock the full potential of your production line

Get in touch with us now and take the first step towards a more streamlined production process. Or do you need an immediate solution to an acute problem? See if you qualify for our Instant Solution Program!