Geochemistry – Part II: Predicting metal leaching and acid rock drainage (ML/ARD) chemistry

While the environmental impacts can be devastating, very costly and difficult to remediate, a scientific and multidisciplinary approach can be used to predict ML/ARD chemistry. With effective predictions, solutions can be engineered to manage and prevent ML/ARD from occurring in the first place, thereby protecting the environment, and potentially saving mining companies several millions of dollars in environmental liabilities.

A summary of the common methods and analytical techniques used in predicting ML/ARD chemistry is presented in Figure 1. These methods and techniques can be classified within the following general approaches to prediction:

• Static testing
• Kinetic testing
• Model predictions
• Iterative data review

Static tests are used to quantify fundamental geochemical parameters (i.e., metal concentration, mineralogy, risk of acid generation, etc.). These tests serve as an excellent first step to assess the ML/ARD risk caused by an abundance of information, a relatively short testing period and relatively low costs. However, while static tests are integral for prediction, they do not provide any information concerning the rate of weathering reactions, which will determine actual risk of ML/ARD on site and in the field.

• Acid-base accounting (ABA) – Analyses and calculations used to assess the risk of acid generation from the oxidation of sulphide minerals. ABA quantifies the acid potential (AP) and the neutralizing potential (NP) of a sample material. Based on the determination of AP and NP, a sample is classified as potentially acid generating (PAG) or non-potentially acid generating (NPAG).

• Bulk chemistry analysis – An elemental analysis used to screen samples for substantial enrichment in particular metals or elements by comparative analysis with background concentrations (i.e., average crustal abundances).

• Shake flask extraction (SFE) – A solubility test procedure whereby solid material is agitated with water for a given duration. The SFE method is used to determine the stored concentration of metals and elements in the extracted water.

• Mineralogical analysis – Analytical techniques such as Reitveld X-ray diffraction (XRD) and quantitative evaluation of materials by scanning electron microscopy (QEMSCAN) are used to quantify minerals contributing to AP, NP and elements of concern. These techniques may also provide insight into mineral grain size, shape and surface area exposure (reactivity).

Kinetic tests are used to quantify the rates of metal leaching, acid generation and neutralization potential. These tests provide insight into the rates of natural weathering reactions, which more accurately predict the site-specific risk of ML/ARD. The rates derived from kinetic tests are also essential inputs for site-wide water quality model predictions required for project permitting. Kinetic tests often take a long time to complete (from several weeks to years), as chemical equilibrium must be attained to accurately quantify reaction rates.

• Humidity cell tests (HCT) – Bench-scale laboratory testing used to simulate and quantify reaction leaching rates of waste rock or tailings under atmospheric weathering conditions.

• Column leach tests – Bench-scale laboratory testing used to simulate and quantify leaching rates of covered and/or underwater waste rock or tailings disposal.

• Field barrel tests (FBT) – Similar to HCT, reaction rates are simulated and quantified under atmospheric weathering conditions. However, FBT are employed directly at the mine development site with large quantities of material (hundreds of kg) per sample, and the natural climate conditions help provide more realistic predictions.

• Geochemical modelling – Model predictions with computer software, such as PHREEQC, use a combination of chemical thermodynamics and experimental results to provide insight into mineral-water interactions within aqueous environments. For example, modelling may be used to simulate the physicochemical conditions (e.g., temperature, pH, redox conditions, chemical composition) to assess the stability and dissolution of acid generating minerals in mine waste.

• Onsite data monitoring – Routine collection of data and analysis during mining operations provides iterative feedback and validates initial ML/ARD chemistry predictions. This allows for rapid identification, if engineering design is required.

Effective ML/ARD prediction starts with early integration of geochemistry into the design of a mine development plan, beginning with the advanced exploration stages and sustaining geochemical practice through to the final mine closure stages. These predictions help guide the engineering design for mine waste and water management plans by understanding the environmental compliance risk and providing insight into design cost‑benefit analyses.

As discussed, there are several tools available to support ML/ARD chemistry prediction. However, a well-rounded approach through static testing, kinetic testing and model predictions is required for confidence in such predictions. Moreover, iterative feedback from onsite monitoring throughout mining operations validates predictions and allows for early identification, if design modification is required.

This blog article was written in collaboration with Neal Sullivan.