Optimization Capabilities for Crushing Plants

Doctoral thesis, 2022

Responsible production and minimal consumption of resources are becoming competitive factors in the industry. The aggregates and minerals processing industries consist of multiple heavy mechanized industrial processes handling large volumes of materials and are energy-intensive. One such process is a crushing plant operation consisting of rock size reduction (comminution) and particle size separation (classification) processes. The objective of the crushing plant operation for the aggregates industry is to supply specific size fractions of rock material for infrastructure development, while the objective in minerals processing is to maximize material ore throughput below a target size fraction for the subsequent process. The operation of a crushing plant is complex and suffers variabilities during the process operation, resulting in a drive for optimization functionality development. Process knowledge and understanding are needed to make proactive decisions to enable operations to maintain and elevate performance levels.

To examine the complex relationships and interdependencies of the physical processes of crushing plants, a simulation platform can be used at the design stage. Process simulation for crushing plants can be classified as either steady-state simulation or dynamic simulation. The steady-state simulation models are based on instantaneous mass balancing while the dynamic simulation models can capture the process change over time due to non-ideal operating conditions. Both simulation types can replicate the process performance at different fidelities for industrial applications but are limited in application for everyday operation. Most companies operating crushing plants are equipped with digital data-collection systems capturing continuous production data such as mass flow and power draw. The use of the production data for the daily decision-making process is still not utilized to its full potential. There are opportunities to integrate optimization functions with the simulation platform and digital data platforms to create decision-making functionality for everyday operation in a crushing plant. This thesis presents a multi-layered modular framework for the development of the optimization capabilities in a crushing plant aimed at achieving process optimization and process improvements. The optimization capabilities for crushing plants comprise a system solution with the two-fold application of 1) Utilizing the simulation platform for identification and exploration of operational settings based on the stakeholder’s need to generate knowledge about the process operation, 2) Assuring the reliability of the equipment model and production data to create validated process simulations that can be utilized for process optimization and performance improvements.

During the iterative development work, multiple optimization methods such as multi-objective optimization (MOO) and multi-disciplinary optimization (MDO) are applied for process optimization. An adaptation of the ISO 22400 standard for the aggregates production process is performed and applied in dynamic simulations of crushing plants. A detailed optimization method for calibration and validation of process simulation and production data, especially for mass flow data, is presented. Standard optimization problem formulations for each of the applications are demonstrated, which is essential for the replicability of the application. The proposed framework poses a challenge in the future development of a large-scale integrated digital solution for realizing the potential of production data, simulation, and optimization. In conclusion, optimization capabilities are essential for the modernization of the decision-making process in crushing plant operations.