top of page

Create Your First Project

Start adding your projects to your portfolio. Click on "Manage Projects" to get started

Operational Controls -- Linden Mining Enterprise (Linmine), Guyana

Project Type

Use of AI to enhance kiln control operations and performance.

Date

April 1997

This Successfully applied data analytics and time series control models to PLC programs that led to significant reductions in the following key performance metrics for the organization’s bauxite calcination process:
• SPOC (Specific Oil Consumption) – the number of gals of oil required to produce one ton of product, RASC calcined bauxite.
• Through-put – The number of tons of RASC produced in one hour.
• Oil flow rate – amount of oil (kg) per hour required to maintain the kiln’s temperature setpoint.
The strategy adopted was to create for each key process an ARIMA statistical AI model using data analytics along with Turtle diagrams to guide improvement opportunities.
The key processes were:
• feed quality,
• calcination,
• product (RASC) blending,
The AI models used the results of multi-variable analysis and ARIMA models to generate process control set points to minimize variations of the calcination environment and ultimately product (RASC) quality.
Process Inputs were assessed to determine the process controls & setpoints forecasting models.
The Statistica Data Mining & forecasting model -- analyzed historical data to identify trends in the inputs (Kiln speed, feed rate of raw bauxite, oil flow, kiln temperature, etc.). Predictive models (transfer functions, multivariable equations, Box-Jenkins ARIMA models) were used to forecast operating setpoints to ensure proactive measures were taken to minimize variations in the final product quality characteristic, BSG (Bulk specific gravity).
Monitoring and Improving Controls
Controls were crucial for ensuring that processes were carried out according to plan.
The AI models improved control mechanisms in the following ways:
• Regulatory Compliance Monitoring: The AI models analyzed real-time data from sensors to track environmental factors like emissions, waste discharged into the Demerara River, and energy usage.
• Automating and Tracking Activities
• The models automated data collection and tracking of process activities, providing real-time insights and supporting decision-making:
• Process Automation & Monitoring: The AI models supported the automated controls of kiln feed rate, kiln speed, and oil flow rates.
• Root Cause Analysis: In the event of non-conformance or failure, the AI models correlated data from different kiln control parameters to identify the underlying causes of issues, helping to resolve them effectively.
• Energy and Material Optimization: The AI models effectively managed the relationship between the kiln control variables to reduce waste by predicting kiln feed and oil flow demand more accurately consequently optimizing energy consumption.

60 Turtle Rock Cove

Acworth GA 30101

279 Meadowbrook Drive

Meadowbrook Gardens

Georgetown, Guyana

+592-504-6066

bottom of page