Reducing Mill Attrition and Energy Consumption in Cocoa Liquor Production

Smarter detection of mill wear and excessive energy use through data-driven monitoring.

‍

The Challenge

Cocoa nibs are ground in ball mills with steel balls that collide to achieve the desired liquor fineness.

As the balls wear down, the shaft speed must increase to compensate — leading to:

• Increased energy consumption

• Risk of equipment degradation

• No automatic detection of when high consumption coincides with the production of a specific liquor

• Only tank weights are available — no tag to identify which tank is being filled

‍

The Approach

The team designed a system to detect when specific products are being made under high energy consumption, using only tank weight data:

• Created formulas to calculate the rate of weight change (derivative) to infer which tank is being filled.

• Used Value-Based Searches to detect when power usage exceeded a set limit (e.g., 75 units) and what specific tanks (e.g., 6 or 9) were being filled for over 1 hour at those conditions.

• Built monitors to send automatic email alerts when the conditions were met.

‍

Calculations for Tank derivative and tank being filled in TrendMiner

‍

The Results

The following table summarizes the key performance improvements achieved:

‍

‍

The Takeaway

By creatively using existing tank weight data and formulas to infer tank activity, the site enabled real-time detection of excessive energy use and took action to prevent wear — without the need for new hardware.

Curious how your data could reveal hidden maintenance needs or efficiency gaps? Let’s bring those insights to life.

‍