Thickener Control via Predictive Mining Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current thickener systems in mineral processing face challenges in achieving optimal water recovery and solid concentration due to slow settling rates and poor overflow clarity, largely because of the complex interactions between lithological composition, flocculant dosage, and upstream process variations, which existing control methods fail to accurately predict and manage.
Innovation Solution
A mineral recovery system incorporating a thickener controller and a mining operations model that combines physical, statistical, and machine learning models to predict future thickener states and optimize control settings in real-time, utilizing data from upstream processes to adjust flocculant and water inputs, thereby improving settling rates and solid concentration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional thickener control methods are used, then operation simplicity is maintained, but water recovery and solid concentration fall below design values due to slow settling rates
Solution Approach 1:
The system performs preliminary action by predicting future thickener states and optimal control settings before actual process changes occur. The predictive model anticipates settling behavior based on upstream process conditions, allowing advance adjustment of flocculant dosage and underflow rates to maintain optimal settling rates and water recovery.
Solution Approach 2:
The system implements feedback by continuously monitoring actual thickener performance against predicted states and using this information to adjust control settings. The control system compares predicted versus actual settling rates and adjusts flocculant dosage and underflow rates dynamically to optimize water recovery and solid concentration.
2Manufacturing precision
If conventional rule-based control is used, then system complexity is minimized, but manufacturing precision of underflow solids concentration deteriorates due to inability to predict thickener behavior
Solution Approach 1:
The system replaces conventional mechanical rule-based control with a computational predictive model. Instead of using fixed rules and manual adjustments, the system employs a predictive model that calculates optimal control settings based on predicted thickener states, upstream process conditions, and material properties, significantly improving underflow solids concentration precision.
Solution Approach 2:
The system changes parameters by dynamically adjusting flocculant dosage and underflow rates based on predicted thickener states. The predictive model determines optimal parameter values that account for varying upstream conditions and material properties, maintaining precise underflow solids concentration control despite changes in feed characteristics.
3Adaptability or versatility
If no predictive modeling is used, then system simplicity is maintained, but adaptability to upstream process variations deteriorates, causing poor overflow clarity and low throughput
Solution Approach 1:
The system performs preliminary action by predicting how upstream process variations will affect thickener performance before these changes fully manifest. The predictive model anticipates the impact of varying feed rates, material properties, and upstream process conditions, allowing advance adjustment of control settings to maintain adaptability and throughput.
Solution Approach 2:
The system implements dynamics by making the control system adaptive and responsive to changing conditions. The predictive model continuously updates predictions based on current upstream process states and material properties, dynamically adjusting flocculant dosage and underflow rates to optimize thickener performance under varying operating conditions.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables more accurate prediction and control of thickener performance, leading to enhanced water recovery, increased solid concentration, and improved thickener stability, overcoming the limitations of previous methods by integrating predictive models and real-time data analysis.
Implementation Method 1
a flocculant is added to the thickener feed which joins solid particles together and increases the rate of settling
Implementation Method 2
When applied in tailings dewatering and liquid ore concentrate extraction thickeners separate solid and liquid fractions. Solid particles settle to the bottom of the thickener
Data Source
AI summary
A mineral recovery system for use in a mining operation is described. The mineral recovery system a thickener includes a process water input, an underflow output having an underflow controller configured to adjust outflow of thickened slurry from the thickener, an overflow output configured to dispense clarified water from the thickener; and a flocculant input and a flocculant dilution input, a thickener controller configured to control an operation of the thickener; and a processor executing a mining operations generated model to issue commands to the thickener controller, based on inputs of sensed conditions, wherein the mining operations model incorporates a thickener sub-model and a material sub-model, wherein the mining operations model is employed to predict a future state of a thickener based on inputs of sensed conditions in the thickener and predictions made by the thickener sub-model and the material sub-model in real time.