Chapter 5

From GARDGuide

5.0 Prediction

5.1 Introduction
Introduction to CMD Prediction
5.2 Objectives of Prediction Program
5.3 The Prediction Approach
5.3.1 Acid Rock Drainage/Metal Leaching Characterization
5.3.2 Prediction during Different Phases of the Mine Life
5.3.3 Water Quality Prediction
5.4 Prediction Tools
5.4.1 Introduction
5.4.2 Geological and Lithological Investigations
5.4.3 Hydrogeological/Hydrological Investigations
5.4.4 Introduction to Geochemical Characterization
5.4.5 Sample Storage and Preparation Prior to Analysis
5.4.6 Summary of Testing Requirements
5.4.7 Physical Characteristics
5.4.8 Total and Near-Total Solid-Phase Elemental Concentration
5.4.9 Mineralogical Properties
5.4.10 Net Acid or ARD Potential
5.4.11 Short-Term Leach Tests
5.4.12 Laboratory Kinetic Tests
5.4.13 Field Methods
5.4.14 Data Management
5.4.15 Quality Assurance/Quality Control
5.4.16 Screening and Evaluation Criteria
5.4.17 Reporting
5.5 Modeling of Acid Rock Drainage, Neutral Mine Drainage, and Saline Drainage for Characterization and Remediation
5.5.1 Introduction
5.5.2 Geochemical Modeling
5.5.3 Hydrological Modeling
5.5.4 Hydrogeological Modeling
5.5.5 Gas Transport Modeling
5.5.6 Statistical Evaluation
5.6 Conclusions
5.7 References
List of Tables
List of Figures
This is the First Page: Sections 5.1, 5.2, and 5.3
Second Page: Section 5.4 Prediction Tools
Third Page: Sections 5.5, 5.6, and 5.7, Lists of Tables and Figures


5.1 Introduction

This chapter presents an overview of the methods available for material characterization and the prediction of drainage water quality, with some guidance as to the usefulness and limitations of the various methods. For more detail, the reader should refer to http://www.mend-nedem.org/reports/files/1.20.1.pdf and the other references and links provided in this chapter.

Prediction of drainage chemistry is a critical part of mine planning; particularly water and mine waste management. The primary objective of mine and process water quality prediction is to evaluate the potential for geologic materials and mine and process wastes to generate acid and other constituents of potential environmental concern, and the potential to affect water resources. As an important corollary, the need for and nature of mitigation measures is determined through prediction. Material characterization and prediction of drainage chemistry needs to be synchronized with overall project planning (Price and Errington, 1998).

Prediction during exploration tends to be generic and generally avoids presumptions about future engineering and mine design. Pre-mine material characterization and prediction and modeling of drainage chemistry need to consider the specifics of engineering and mine design. Iteration may be required as results may lead to a revision of aspects of both the prediction program and the mine plan. The timing of the prediction program must be synchronized with the mine development so that the findings of the characterization and prediction efforts can be used for the mine design.

Accurate prediction of future mine discharges requires an understanding of the analytical procedures used and consideration of the future physical and geochemical conditions, external inputs and outputs, and the identity, location and reactivity of the contributing minerals (Price, 2009). All sites are unique for geological, geochemical, climate, commodity extraction, regulatory, and stakeholder reasons. Therefore, a prediction program needs to be tailored to the site in question. Also, the objectives of prediction programs are variable. For example, objectives can include definition of water treatment requirements, selection of mitigation methods, assessment of water quality impact, or determination of reclamation bond amounts.

Predictions of drainage quality are made qualitatively and quantitatively. Qualitative predictions involve assessing whether acidic conditions might develop in mine wastes with the attendant release of metals and acidity to mine drainage. Qualitative predictions have been performed for at least 40 years and although errors have been made, often due to inadequate sampling, the predictions have been successful for many mine sites around the world. Indeed, predictions of whether acidic conditions could develop for high sulphur (often acid producing) and low sulphur (often nonacid producing) are often straightforward. Where qualitative predictions indicate a high probability of ARD production without mitigation, attention quickly turns to reviewing alternatives to prevent ARD and the prediction program is refocused to assist in the design and evaluation of potential success of that program. Significant advances in the understanding of ARD have been made over the last several decades (see Chapter 2), with corresponding advances in mine water quality prediction and use of prevention techniques. However, mine water quality prediction can be challenging because of the wide array of reactions involved and potentially long time periods to cross geochemical thresholds and achieve specific conditions related to ARD, NMD, and SD generation.

The understanding of equilibrium vs. kinetic controls on mineral reactions and their effect on water quality is of particular importance when predicting mine drainage chemistry. Equilibrium conditions are relatively simple to simulate, but might not always be achieved in mine drainage waters under ambient conditions. Conditions governed by rate-limited reactions are common and more difficult to evaluate. However, through the use of state-of-the-art geochemical testing programs, both equilibrium conditions and rate-limited reactions can be assessed.

Despite the uncertainties associated with quantitative estimation of future mine water quality, quantitative predictions developed using a range of realistic assumptions and a recognition of associated limitations have significant value as ARD management tools and environmental impact assessment. From a risk-based perspective, the probability of a certain consequence (i.e., drainage quality) occurring is examined during the testing and prediction stage.

The following approaches have been used for predicting water quality resulting from mining activities:

  • Test leachability of waste materials in the laboratory
  • Test leachability of waste materials under field conditions
  • Geological, hydrological, chemical, and mineralogical characterization of waste materials
  • Geochemical modeling

Analog sites or historical mining wastes on the property of interest are also valuable in ARD prediction, especially those that have been thoroughly characterized and monitored for water quality and have many similar characteristics as the site in need of prediction. The development of geo-environmental models is one of the more prominent examples of the “analog” methodology. As described in Chapter 2, geo-environmental models of a mineral deposit are a compilation of geologic, geochemical, hydrologic, and environmental information pertaining to the environmental behavior of geologically similar mineral deposits. Geo-environmental models are a general guide that will help anticipate potential environmental problems at future mines, operating mines, and orphan sites.

A schematic depiction of the progression in prediction objectives and activities during the development of a hard rock mine is illustrated in Figure 5-1 and discussed in more detail in this chapter. More detail on the prediction of coal mine drainage (CMD) is presented here: Introduction to CMD Prediction.

Figure 5-1: Generic Prediction Program Flowchart

GenericPredictionProgramFlowchart.gif

Top of this page

5.2 Objectives of Prediction Program

The purpose of a drainage chemistry prediction program is to characterize mine wastes and walls and to anticipate problems so that, if required, impact prevention measures (see Chapter 6) can be implemented in the most cost-effective manner. The objective is to predict drainage chemistry and contaminant loading with sufficient accuracy to ensure mine and mitigation plans achieve the specified environmental objectives (Price, 2009). Adaptive management and contingency plans may be the most cost-effective approach to mitigation.

Predictions occur at different levels of complexity and for different reasons. In the context of pre-mine water quality prediction, the most important questions generally are: Without mitigation, will problematic drainage chemistry be produced from a particular:

  • Geological unit?
  • Zone of the deposit?
  • Mine facility or waste type?
  • Particular mining stage or phase?

This set of questions can be answered if an appropriate database on geochemical characteristics is available and a sound understanding of geological and mineralogical conditions has been developed. The strength of the database required depends on the variability and complexity of the contributing chemical species and minerals, the geological units, mine facilities and waste types. For example, a more comprehensive database may be required where there are significant variations in sulphur and carbonate mineral content or if the sulphur and carbonate mineral content are in close balance. The presence of elements, such as selenium (Se) and mercury (Hg), or minerals, such as Fe-carbonate and alunite, whose performance is difficult to predict, may create additional challenges.

Without mitigation, ARD will invariably produce environmental impacts. Where ARD will not occur, the potential for metal release under near neutral pH conditions must still be assessed. Special attention is often placed on trace elements that can be quite soluble at neutral pH such as zinc, cadmium, nickel, antimony, selenium, and arsenic. Whole rock analysis and laboratory kinetic tests can be quite effective in assessing potential near-neutral or alkaline drainage chemistry.

The quantitative prediction of drainage quality is more difficult than establishing whether ARD will be generated. However, in many cases, an accurate quantitative prediction of drainage quality is not required. Instead, it may be sufficient to know for design, operational, or closure purposes whether a particular drainage will meet certain water quality standards, whether it will be ARD, NMD, or SD type water, and what the overall volume will be. Therefore, all prediction efforts (and associated information needs and level of complexity) need to be tailored to the question at hand. As a general rule, the amount of information and sophistication of the water quality prediction approach used must reflect the scale at which the problem is to be addressed, the availability of information, and the level of detail, accuracy, and precision required.


Top of this page

5.3 The Prediction Approach

5.3.1 Acid Rock Drainage/Metal Leaching Characterization

Figure 5-1 represents an idealized generic overview of a comprehensive ARD/ML prediction program. Application of this approach needs to be customized to account for site-specific aspects. The program, as presented, applies to a project that advances from exploration through to mine closure.

The flowchart in Figure 5-1 assumes that ARD/ML prediction activities are performed at every stage of a project. These activities are coupled with other project planning activities and the level of detail of ARD/ML characterization activities is determined by the stage of the project. Data are accumulated as the project proceeds so that the appropriate information needed to support engineering design is available in a timely manner.

The following six mine phases are identified in the GARD Guide:

  • Exploration
  • Mine planning, feasibility studies, and design (including environmental impact assessment)
  • Construction and commissioning
  • Operation
  • Decommissioning
  • Post-closure

The flowchart focuses on the earlier stages of mine development, a critical period for proactive mine development, when the initial geochemical characterization is usually conducted. The description of mine phases in Figure 5-1 therefore differs slightly from the convention used in the GARD Guide. Both sets of nomenclature are presented.

The major “pillars” of the flowchart are as follows:

  • Typical Project Phase. Five typical major project phases of the mining cycle are included in Figure 5-1 (initial exploration, advanced exploration, prefeasibility, feasibility/permitting, and project implementation).
  • Minimum Objective of ML/ARD Program. The overall minimum objective for each project phase of the ARD/ML program is indicated on the flowchart. For each project phase, the minimum objective is typically defined based on the economic assessment of the project. These objectives are described as “minimum” requirements because project managers may choose to meet the objectives of subsequent phases to avoid delays.
  • ML/ARD Program Stage. This header indicates the level of characterization that is needed to meet the objective.
  • ML/ARD Program Activities. This element indicates the main types of prediction and characterization activities. All activities are considered cumulative. Activities occurring in earlier phases are continued here as needed to meet future objectives.

If new information becomes available during any one of the stages of the ARD/ML program (e.g., a change in mine plan, or unexpected monitoring results), re-evaluation of earlier stages may be required. These types of iterations are omitted from the flowchart in Figure 5-1 for clarity. An approach for characterization, classification and prediction adopted by Earth Systems is documented in the Characterization Case Study.

Top of this page

5.3.2 Prediction during Different Phases of the Mine Life

5.3.2.1 Initial Exploration/Site Reconnaissance Phase

During the initial exploration/site reconnaissance phase, the following activities take place: surface geological mapping, geophysical surveys, soil and stream sediment surveys, trenching, and wide-spaced drilling. The information acquired from these activities is used by project geologists to develop a conceptual geological model for the mineral prospect. In the context of managing existing sites, reconnaissance occurs at this stage to obtain historical and site layout information to define subsequent investigations.

The information collected during the initial exploration is not specifically interpreted for ARD/ML potential but becomes the foundation for subsequent evaluations. For example, geological mapping and mineralogical studies should consider the host or country rocks in addition to the ore. A core logging manual should be developed so that logs provide information that can be used for ARD/ML characterization. Core should be suitably stored to be available for future analyses. Rock samples should be analyzed using multi-element scans (including sulphur and carbon) in addition to the suspected commodity elements. Collection of environmental baseline data (soil, sediment, surface water, groundwater, and air) should begin during this phase.

5.3.2.2 Advanced Exploration/Detailed Site Investigation Phase

The advanced exploration/detailed site investigation phase usually involves additional drilling at narrower spacing and, where appropriate, underground development to improve delineation of the ore body, but normally a mine plan has not been developed during this phase. Specific ARD/ML characterization begins early in this phase. The geological model for the project provides a basis for design of a Phase 1 (initial or screening) ARD/ML static test program (Table 5-1 provides more detail on testing methods). The geological model also affords an opportunity for comparing the project to analogs, which may indicate a potential for drainage quality issues, and provides focus for the initial investigation. At this stage, water sampling in the area should include any existing facilities and natural weathering features (e.g., gossan seeps).

Table 5-1 is large enough to require its own page:
Table 5-1: Methods for Geochemical Characterization (Table 5-1 provides more detail on testing methods.)

5.3.2.3 Prefeasibility Phase

The prefeasibility phase includes development of initial mine plans (or closure plans for existing sites). During this phase, the results obtained during the Phase 1 program are coupled with the mine, waste, and water management plans to design a detailed Phase 2 ARD/ML characterization program that will lead to development of waste management criteria and water quality predictions. The Phase 2 characterization program will include static chemical and physical testing, mineralogical characterization, and implementation of laboratory and field kinetic tests specifically designed to answer questions about the geochemical performance of the individual mine and infrastructure facilities. A preliminary waste geochemical block model might be developed during this phase that can be used to initially estimate the quantities of different types of wastes.

5.3.2.4 Feasibility and Permitting Phase

The feasibility and permitting phases are not distinguished as separate phases in the flowchart because the ARD/ML characterization needs are essentially the same for feasibility and permitting, and the transition from a positive feasibility study to environmental assessment and permitting often occurs rapidly or occurs in parallel and therefore allows little time for additional studies.

The main activity in this phase is the development of source water quality predictions, which are used in the feasibility study (e.g., to determine water treatment requirements) and to evaluate the water quality effects of the project. The predictions are developed by coupling findings of the Phase 2 program with waste schedules and hydrological data for individual facilities. The predictions are used in the internal load balance for the site and as direct inputs to downstream groundwater and surface water effects assessments (see Chapter 8).

The flowchart in Figure 5-1 shows iterative loops from the source term predictions back to the Phase 2 program and show iterative loops from the effects assessment back to the source term predictions because further modeling and testing may be needed to refine water chemistry predictions. The parallel process for mine or closure planning may result in the redesign of some aspects of the mine or closure to address unacceptable effects or costs.

Following completion of an acceptable mine plan, monitoring plans are designed to inform waste management decisions (e.g., analysis of blast hole sample for waste classification) and verify water chemistry predictions (e.g., seep sampling) (see Chapters 8 and 9).

5.3.2.5 Construction, Operational, Closure and Post-Closure Phases

Prediction is a cradle to grave activity that does not finish when mining starts, but continues during construction, mining and processing, closure and post-closure. Objectives of prediction during mining and processing and each subsequent phase of the mine life are to verify, refine and fill gaps in the predictions from the previous phase. This is achieved through:

  • Material characterization
  • Monitoring of weathering conditions, drainage chemistry and loadings
  • Studies to address information gaps

This section provides an overview of these activities. A more detailed description is provided in Price (2009) (http://www.mend-nedem.org/reports/files/1.20.1.pdf). The best time for material characterization is during mining and processing when the materials can be most easily sampled and the information can be used to guide materials handling. Objectives of operational material characterization include:

  • Verify, refine and address gaps in the pre-mine characterization
  • Segregate materials requiring different disposal or mitigation
  • Create an inventory of the composition of materials and the mass and location of different types of material created by mining (e.g., mine walls and waste rock), processing (e.g., tailings), reprocessing (e.g., desulphurized tailings) or during deposition (e.g., tailings sand and slimes)

It is important to conduct operational material characterization for the same reason that mines conduct more detailed characterization to check pre-mine predictions of ore grades. Operational material characterization also fills information gaps that result from a lack of drill core prior to mining at the perimeter or at great depth, a lack of waste rock fines, differences between pilot and large-scale processing facilities, limited tailings samples, and uncertainty regarding the location of final mine walls.

Considerations in sampling and interpretation of analytical results include an identification of the reactive portion of a mine waste, whether segregation occurs during handling and deposition, and whether there is further processing, reprocessing, co-deposition or use of additives (Price, 2005b). Sampling becomes far more difficult once materials are buried (e.g., lower lifts of waste rock) or access is cut off to a portion of a project component (e.g., pit walls or backfilled underground workings). In addition to waste materials produced or surfaces exposed by mining and processing, characterization should be conducted on geological materials used to construct roads, foundations and dams, and stripped as part of mine construction. Sampling and analysis requirements for operational characterization of different materials created by mining are discussed in more detail in Price (2009), Chapters 7, 8 and 9.

Ensuring sufficient time to sample, analyze and act on the results may be a challenge where material characterization is used to segregate materials or verify that mitigation processes, such as desulphurization, have been effective before disposal can proceed. Effective communication will be needed between the parties responsible for each task where material characterization is used to manage materials that are a potential source of problematic drainage chemistry.

In an effective prediction program, in addition to permit compliance, monitoring is conducted to track trends, inform corrective actions and permit proactive resolution of problems, adaptive management and timely implementation of contingency plans. Monitoring should include measurement of properties and processes that cause mineral instability and changes in drainage chemistry and contaminant loadings. Since weathering processes such as mineral depletion or mine wall collapse may take many years to occur, long-term monitoring will usually be required.

A common target of weathering and seepage monitoring are wastes left exposed for some period of time prior to flooding that have an uncertain time to the onset of acidic weathering conditions. Periodic analysis of solid-phase samples from the surface of project components or field test pads can be used to measure mineral depletion to warn when accelerated flooding may be required. Geochemical and physical heterogeneity of project components may be a challenge when monitoring weathering and drainage chemistry. One solution to the challenge of tracking the performance of materials with different geochemical properties is to construct field test pads from each different material of concern.

Not all prediction questions can be answered prior to mining. Most mines need operational and post-operational studies to address unknowns in mitigation and closure plans. Common reasons for needing operational and post-closure studies include:

  • Relatively short-term nature of pre-mine kinetic tests
  • Differences between actual materials and weathering conditions at the site and materials and conditions in laboratory tests
  • Uncertainty prior to mining about the composition of tailings, tailings sand and slimes, and waste rock fines
  • Uncertainty prior to closure about the location of final mine walls, degree of wall collapse, reclamation plans or hydrogeology of the closed site (e.g., rebound in the water table, groundwater chemistry or the height of the water table
  • Operational changes to excavation, processing, waste handling and reclamation plans that change the composition, hydrogeology, size, and location of mine workings and waste materials

There is often great value in continuing pre-mine laboratory kinetic tests and setting up field test pads or monitoring sites on project components to study materials of concern. Prediction of post-closure drainage chemistry should be part of the first mine plan, and should re-occur at regular intervals (e.g., every five years) or whenever there are significant changes to site or project conditions (e.g., changes in drainage chemistry or mine plans). More detailed and accurate material characterization and information regarding site and project conditions at closure will become available as the project develops.

Mine closure may be a difficult time to conduct prediction work and collect data, with facilities being dismantled, staff departing, and equipment removed. Starting to address outstanding closure prediction questions early in the mine life will allow a mine to use its operating facilities, equipment and personnel when initiating and conducting studies, and provide more time to perform the studies and act on the results. Another important consideration in encouraging an early start to closure studies is reduced access after portions of the mine close (Price, 2005b).

After a mine closes, many properties and processes controlling weathering are in flux and there are a number of possible scenarios regarding future drainage chemistry. Many mines need post-closure monitoring and studies to address unknowns regarding future drainage chemistry. Post-closure prediction should continue for as long as there is significant uncertainty regarding environmental behavior of mine materials and a potential need for the proactive resolution of drainage chemistry problems.

Thorough, cradle to grave prediction of drainage chemistry is a relatively new phenomenon. Many older mines lack comprehensive information on operational material characterization of tailings and waste rock, and have no record of the magnitude and disposal location for material with different geochemical properties. Another common omission is a lack of long-term kinetic tests or well-characterized kinetic test samples.

It may not be possible to collect all the missing information and resolve the uncertainty regarding future drainage chemistry. For example, it is generally not feasible to collect an intact sample of the finer size fraction of waste rock buried within large dumps built in several lifts.

Top of this page

5.3.3 Water Quality Prediction

It is important to determine the objectives and the manner in which data will be interpreted when designing a prediction program. Figure 5-2 provides a generalized flowchart that shows the objectives and use of analytical and test results for the prediction of potential water quality effects (Maest and Kuipers, 2005).

Figure 5-2: Generalized Flowchart for the ARD Prediction Approach at Mine Sites
(after Maest and Kuipers, 2005)

GeneralizedFlowchartfortheARDPredictionApproach.gif

The first step in water quality prediction is to determine the prediction objectives, the importance of which is discussed in the Section 5.1, and set up the site conceptual model discussed in Chapter 4. As site characterization progresses through collection of data (geology, hydrology, mineralogy, and mineral extraction/processing), the conceptual model continues to be refined, and may change as more data become available (Younger and Sapsford, 2006). The core of the conceptual model should be a schematic that shows the major sources of contaminants (e.g., mine portals, open pits, tailings, waste rock piles), the main means of transport (e.g., wind, surface water, groundwater), and the receptors (e.g., atmosphere, lakes, reservoirs, streams, rivers, soils, aquatic biota, terrestrial flora and fauna). Figure 5-3 is an example of a conceptual model in cartoon format, developed for the Iron Mountain Mine (California) and its receiving environment. Figure 5-3 can be made into a schematic (flowchart, flux chart or reservoir chart) with the size of the arrows proportional to flow as shown in Figure 5-4.

Figure 5-3: Conceptual Model Showing Metal and Acid Source Regions at Iron Mountain
and Downstream Transport Pathways to the Sacramento River

ConceptualModelShowingMetalandAcidSourceRegionsatIronMountain.jpg

Figure 5-4: Flowchart for Metal and Acid Source Regions at Iron Mountain and
Downstream Transport Pathways to the Sacramento River

FlowchartforMetalandAcidSourceRegionsatIronMountain.jpg

Each reservoir contains a certain mass amount and average concentration of the parameters of interest (acidity, metals, and sulphate in the case of ARD) and each arrow represents a given flux (or load) of those parameters from one reservoir to the next. Because the rates may change (e.g., with hydrologic conditions, irrigation needs, or other uses), a different set of conditions can be shown by both a range of values and a different flowchart with different values for different times of year.

Within each reservoir and flux, geochemical processes, such as precipitation or sorption of metals, result in more dilute solutions. It is within these parts of the flowchart that static/kinetic tests and geochemical modeling can be helpful. For a complex mine site with an open pit, underground workings, waste piles, diversions, and tailings piles, each one of these units should be identified, their rate of weathering and water transport quantified, and the consequences for receiving water bodies determined. A water balance (i.e., a numerical representation of the flowchart) should be developed for the system that takes into account precipitation, infiltration, and evapotranspiration. The effect of extreme events, such as floods and droughts, might also be assessed. For example, the timing and volume of infrequent high precipitation events are important in predicting drainage quality and quantity in quite arid environments.

All geochemical reactions of relevance to water quality prediction should be placed in a hydrogeological context through the flowchart. The main transport pathways can be shown by arrows and by flux numbers where available. Selection of the model to be used for water quality prediction (Figure 5-2) should take into account the prediction objectives.

The hydrogeochemical modeling is conducted using site-specific information to the maximum extent possible. This hydrogeochemical modeling results in prediction of contaminant concentrations at a number of predetermined locations (e.g., compliance points) or receptors. Through use of multiple input values, sensitivity analyses, and “what-if” scenarios, a range of outcomes is generated, bracketing the likely extent of water quality compositions and potential impacts.

Through a comparison of water quality predictions against relevant water quality standards, the need for mitigation measures or redesign of the mine plan can be identified (Figure 5-2). If predicted concentrations meet standards, additional mitigation measures will likely not be required. If, however, predicted concentrations exceed standards, mitigation measures will be necessary and their effectiveness should be evaluated using predictive modeling and active monitoring during and after mine operation. If the proposed mitigation measures are deemed inadequate for meeting standards, a reassessment of mitigation measures and possibly even of the mine design may be required. The prediction process then repeats itself, possibly including development of an improved conceptual model and additional data collection. Clearly, mine water quality prediction is an iterative process that can take place on an ongoing basis throughout the life of a mine, from the exploration phase through post-closure monitoring.


Next Page (Page 2) of Chapter 5

Last Page (Page 3) of Chapter 5


Top of this page