A Practical Guide for Geostatisticians and Industry Professionals
This book is designed for both academics and practitioners, bridging the gap between theory and application in geostatistics. Each chapter focuses on a different aspect of mineral resource estimation, breaking down complex mathematical and statistical concepts into easy-to-understand steps. Whether you are a third-year student in Geology or a professional in Mining Engineering, this book serves as a comprehensive reference that guides you through every stage of the geostatistical process.
- Step-by-step guidance: From linear to non-linear geostatistical methods, each process is explained in detail, making it easy for readers to follow along and implement the techniques themselves.
- Hands-on code demonstrations: The book includes over 10,000 lines of Python code, giving readers the opportunity to directly apply the methods discussed, reinforcing learning through practice.
- Simplified mathematical explanations: While geostatistics can be mathematically intense, this book breaks down the concepts into simpler terms, ensuring that even complex ideas are accessible.
- Real-world case studies: Numerous case studies and examples are provided to illustrate the practical application of geostatistics in real mining projects, offering readers valuable insight from both academic and industry perspectives.
- Integrated resource estimation and reconciliation: The book takes a holistic approach, covering not only resource estimation but also the steps involved in mine-to-mill reconciliation, helping professionals optimize mining operations from start to finish.
Who Should Read This Book?
- Third-year and postgraduate students studying Mineral Resources Management, Geology, Spatial Statistics, and Mining Engineering.
- Industry professionals looking for practical guidance on applying geostatistics to their work in mineral resource estimation and ore reserve conversion.
I am also working on a new book:
Book Title: Data Analytics and Machine Learning Applications in Extractive Metallurgy
Authors: Glen Nwaila; Yousef Ghorbani; Mehdi Safari
This book provides an indispensable guide for professionals and students aiming to adopt data science and machine learning into their workflows within extractive metallurgy. With its focus on Python-based applications, readers can explore how big data, and advanced analytics improve operational efficiency, reduce environmental impact, and increase profitability. Each chapter is packed with real-world examples, hands-on coding exercises, and practical workflows that can be immediately applied to mineral processing and extraction challenges. As industries increasingly adopt digital transformation, this book stands as a necessary resource for professionals eager to stay competitive and innovative within the mining and metallurgical sectors. The combination of cutting-edge technologies and proven methodologies makes this book an essential reference for both academic and industry professionals.
The book provides theoretical underpinnings as well as practical, hands-on approaches to solve real-world problems in mining, metal recovery, and process optimisation. The text incorporates cutting-edge Python algorithms and over 10,000 lines of code to offer demonstrations of how digital technologies can be seamlessly integrated into traditional metallurgical processes. With numerous case studies, practical insights, and simplified explanations of complex models, this book is an invaluable resource for students, researchers, and professionals looking to integrate artificial intelligence (AI), machine learning, and data-driven methodologies into their mineral processing workflows.
Book Features
- Step-by-step Python code demonstrations for modeling and optimisation.
- Simplified explanations of machine learning models tailored for the extractive industry.
- In-depth discussion of nonlinear processing techniques and their impact on mineral extraction.
- Practical examples and case studies from industry and academia for a real-world perspective.
- Focus on increasing sustainability, resource utilisation, and economic viability.
This book is targeted towards third-year students, postgraduate students, and professionals working in geometallurgy, data science, and mineral processing.
Chapter 1: Introduction
This chapter sets the stage for the rest of the book, introducing the importance of data analytics and machine learning in the current landscape of extractive metallurgy. It explains the challenges of conventional metallurgical processes—such as energy inefficiency, manual operations, and delayed decision-making—and contrasts them with the potential of automation, predictive analytics, and AI. Readers will gain insight into how industry-wide digital transformation is reshaping mining and processing operations for higher productivity, safety, and environmental sustainability. The introduction outlines key themes that will be explored in detail throughout the book, such as process control, optimisation using machine learning models, and the deployment of big data in the mining sector.
Chapter 2: Comminution
Comminution, the process of reducing material size through crushing and grinding, is critical in metallurgy due to its high energy consumption and impact on subsequent processing stages. The chapter explores the fundamental engineering principles behind comminution, including the physics of fracture mechanics, energy transfer, and thermodynamics. Key mathematical models such as Bond's Law, Kick's Law, and Rittinger’s Law are discussed in relation to predicting grindability and energy requirements. Different types of comminution equipment—jaw crushers, ball mills, and rod mills—are explained with a focus on their operational principles and efficiency.
Special emphasis is placed on the importance of particle size distribution in determining downstream process efficiency. The chapter also explores how thermodynamics governs the heat generated during comminution, which impacts material integrity. Fluid mechanics concepts are explored in relation to the grinding media and slurry, and the stress-strain relationships are reviewed to understand material fracture behaviour. Understanding the role of compressive forces, impact and attrition in comminution systems provides an in-depth knowledge for optimising both mechanical design and process conditions. Understanding these fundamental concepts can help minerals engineers to effectively evaluate comminution circuit efficiency, energy consumption, and the impact of various feed characteristics on downstream processes.
Chapter 3: Flotation
Flotation is a separation process widely used in the beneficiation of ores, primarily based on the differences in surface chemistry of minerals. This chapter focuses on the physics and thermodynamic principles that govern flotation processes, focusing on surface tension, bubble-particle interactions, and hydrophobicity. The engineering aspects of flotation are covered, including the design and operation of flotation cells, the types of reagents (collectors, frothers, and depressants) used, and how they affect the bubble formation, froth stability, and selectivity of mineral recovery.
The chapter highlights key mathematical models used to describe flotation kinetics, such as the first-order flotation model, which predicts the rate at which particles attach to bubbles and are recovered. This section emphasizes mass transfer principles, discussing how bubble dynamics, collision rates, and attachment efficiency play crucial roles in determining flotation efficiency. Detailed descriptions of mechanical flotation cells, column flotation, and the Jameson cell are provided, each highlighting the differences in operational mechanisms and the advantages for specific ore types.
The thermodynamics behind bubble-particle attachment is also explored, including how interfacial forces (e.g., van der Waals and electrostatic forces) impact mineral recovery rates. The chapter concludes with a discussion on fluid mechanics within flotation cells and how the flow dynamics of pulp and air bubbles contribute to froth formation and mineral recovery. The importance of optimising air flow, pulp density, and reagent dosages is emphasized, ensuring that engineers can apply these concepts to enhance flotation performance.
Chapter 4: Leaching
Leaching is a hydrometallurgical process that extracts valuable metals from ores through chemical dissolution. This chapter explores the chemical thermodynamics and kinetics that drive leaching reactions, focusing on the conditions necessary to achieve efficient metal dissolution. Fundamental equations governing reaction rates, such as the Arrhenius equation, are discussed to understand how temperature, concentration, and surface area influence leaching performance.
The chapter covers the types of leaching processes—including heap leaching, in-situ leaching, and agitation leaching—and explores the advantages and challenges associated with each. Special attention is given to the mass transfer mechanisms involved in leaching, particularly the role of diffusion through the porous ore matrix, the convective transport of leachants, and chemical reaction control. The thermodynamic equilibrium of leaching reactions is explained using Gibbs free energy and Le Chatelier’s principle to describe the factors that shift reactions toward higher metal recovery.
The chapter also explores the types of leachants used, such as cyanide for gold extraction, sulfuric acid for copper leaching, and ammoniacal solutions for nickel and cobalt extraction. A deep dive into thermodynamic stability diagrams helps explain the optimal conditions for metal dissolution and impurity removal. Chemical kinetics models, such as the shrinking core model, are discussed in relation to leaching performance and optimization. Furthermore, environmental considerations in leaching are addressed, emphasising the need for minimising reagent consumption and managing tailings effectively.
Chapter 5: Exploratory Data Analysis
This chapter introduces readers to Exploratory Data Analysis (EDA), an essential step in preparing data for machine learning models. EDA involves data cleaning, visualisation, and transformation to uncover hidden trends and patterns in datasets. The chapter covers key Python libraries such as Pandas, NumPy, and Matplotlib that facilitate data analysis. Special attention is paid to anomaly detection and the identification of correlations in metallurgical processes, where poor operational data can lead to suboptimal process outcomes. Real-world examples of EDA in mining operations are provided, where engineers use this technique to understand ore variability, grade distributions, and plant performance.
Chapter 6: Process Conditional Simulation
In this chapter, readers explore the practical applications of conditional simulation in the context of gold extraction. Conditional simulation allows process engineers to evaluate the uncertainty in ore body modeling and predict how variations in ore quality can affect downstream processes like flotation and leaching. Python codes for Monte Carlo simulations are presented, allowing users to generate multiple realisations of ore deposits and understand the range of possible outcomes. Through case studies, the chapter demonstrates how mining companies use simulation techniques to improve decision-making and optimize resource extraction strategies. The case studies also illustrate how combining simulation with machine learning techniques can enhance the accuracy of production forecasts, ensuring more effective resource utilisation.
Chapter 7: Regression and SHAP in Process Control
This chapter introduces regression models and their role in process control, focusing on how SHAP (Shapley Additive Explanations) values can be used to interpret the predictions of machine learning models. The discussion revolves around applications in ore processing, where predictive models forecast important variables such as metal recovery, throughput, and energy consumption. Readers will gain an understanding of how to use Python to develop regression models, assess their performance, and generate SHAP values that explain the contribution of each feature to the model's output. Through incorporating interpretable machine learning, this chapter helps process engineers and decision-makers understand the drivers behind the predictions, enabling more informed operational changes.
Chapter 8: Classification Techniques
Classification is a fundamental task in machine learning that is crucial for ore sorting, material characterisation, and process state identification. This chapter covers popular classification algorithms such as logistic regression, decision trees, and k-nearest neighbours (KNN), applied specifically to challenges in mineral processing. Real-world applications include sorting low-grade ores, detecting impurities, and identifying optimal process conditions. Readers are provided with Python-based implementations of classification models, allowing them to experiment with various techniques and understand the factors that influence classification accuracy. Case studies explore the use of classification algorithms in automated systems for sorting ores based on grade and texture.
Chapter 9: CNN Texture Classification
Convolutional Neural Networks (CNNs) are particularly effective in classifying complex visual data, such as ore textures. This chapter focuses on the use of CNNs for texture classification, a critical task in mineral processing where ore texture can significantly impact processing performance. The chapter begins by explaining the architecture of CNNs and how they are trained on large datasets to recognize different ore textures. Python-based implementations of CNNs for texture classification are provided, allowing readers to experiment with real-world data. Case studies demonstrate how CNNs have been used in operations to automate ore characterisation, reducing the need for manual inspection and increasing the speed and accuracy of ore sorting.
Chapter 10: Principal Component Analysis and Clustering
The final chapter introduces Principal Component Analysis (PCA) and clustering algorithms, both of which are essential for handling high-dimensional datasets in metallurgical processes. PCA is used to reduce the dimensionality of datasets, allowing process engineers to focus on the most critical variables influencing process outcomes. Clustering algorithms, such as k-means and hierarchical clustering, are used to group similar data points, helping engineers identify patterns and optimize process control strategies. Python implementations of PCA and clustering techniques are provided, giving readers the tools they need to analyse large datasets and improve decision-making. This chapter also covers how unsupervised learning and clustering techniques can be combined to deliver actionable insights in real-time, further enhancing operational efficiency.
