*Bruce W. Downing, M.Sc., P.Geo. is Senior Geologist, Gamah International Ltd., Vancouver.

Editor's Note: This session discusses general sampling methods for geochemical investigations, including ARD. The session contents at the text level are authored by Bruce W. Downing.


Sample: representative fraction of body of material, removed by approved methods, guarded against accidental or fraudulent adulteration, and tested or analyzed in order to determine the nature, composition, percentage of specified constituents, etc. and possibly their reactivity. (Thrush et al. (1990))

Sample: a small quantity of material relative to the geological mass it represents, collected according to a systematic procedure, of measurable reliability, from which the acid rock drainage potential of the mass which it represents may be estimated, based upon appropriate protocols (Vallee (1999)).

The acid rock drainage (ARD) appraisal involves five work stages; namely deposit definition, project engineering, project economics, production/operation and reclamation/closure. In each of these stages, there is an element of sampling and analysis to meet environmental requirements. Inherent in sampling is quality assurance and quality control.

Poor sampling techniques and inadequate sample selection can contribute to excessive variance, difficulties in interpretation and incorrect assessment in an ARD prediction program. As a result of poor sampling there may be serious consequences and costs. Samples must be representative of all geologic, lithologic and alteration units related to the mine development plan and be representative of the relative amounts and particle size of each type of material (Steffen, Robertson and Kirsten (1992)), in other words what are the geological controls to ARD. A sampling program should be designed to be iterative in nature because the pre-sampling notion of what and how much is representative may shown by the results to be inadequate.

The construction of a sound ARD database begins with good sample collection. Appropriate sample collection, preparation, analytical and QA/QC procedures must be maintained throughout the project life. Sampling is the single most important aspect of a good survey, for without good sampling, results may not be valid and hence correct interpretations difficult to achieve. The essence of sampling is to examine geochemical variation. The proper collection and handling of field samples is the first step towards preventing environmental prosecution and liabilities. This aspect of 'legal sampling' is the focus of a training and reference CD-ROM that has been assembled by staff from Environment Canada (web site). Guidelines and manuals have been assembled by the Mine Environment Neutral Drainage personnel (MEND) (web site) and information concerning them is available from their web site.

The basic components of an ARD sampling program are:

Although sampling costs can be quite expensive, costly remediation and reclamation plans will be based on the results and interpretation of sampling and analysis, hence the importance of a correct assessment. Sampling costs should therefore not pre-determine the number of samples taken and analyzed but should be dependent upon the amount necessary to increase confidence in the data and should be considered with regard to exploration, development and production decisions that rely on sampling results. The number of samples to be taken should be premised on characterization of both the waste and non-waste material. The number of samples will also depend upon spatial variation of waste material. Steffen Robertson and Kirsten et al (1989) developed a curve (Figure 1 below) that can be used as a guideline to ascertain the minimum number of samples required to characterize each geological unit in terms of acid generation and leaching potential. The curve was developed based on a limited number of sites within British Columbia, Canada as a function of the mass of the geologic unit being sampled. There are no strict guidelines for establishing sampling protocols, but one must judge the best method(s) and be prepared to defend those methods.

Figure 1: Hypothetical curve to determine the number of samples required to characterize geological units.

Sampling is an integral part of Risk Assessment and Risk Management. The initial sampling will help determine if there is a risk, and if so what are the variables that must be mitigated (ie. high levels of elements such as arsenic). This initial sampling will also be used as part of the basis for an environmental bond. The ongoing sampling plan will determine the level of mitigation required at closure and if the risk has increased or decreased. Risk Assessment and Risk Management should always be viewed from the initial to final sampling, which means from baseline survey to closure work.

Design and Planning

The design and planning stage should begin in the early exploration phase. An orientation study is imperative. This study should pay close attention to details in the design which have a bearing on the costs of the major ARD program. Its purpose is to ensure that the proper samples are collected and proper preparation and chemical analyses are done in order to collect as much information as possible and avoid costly mistakes. The scopes of characterization and analyses for each category is shown in Figure 2.

Figure 2: The scopes of characterization and analyses for each mine component.

Continuity is important in the estimation and classification of acid generating and non acid generating material. Continuity is essentially composed of two parts; geological continuity and value continuity. Geological continuity refers to features that control mineralization, while value continuity refers to a value or measure such as neutralization potential (NP) and the manner in which this variable occurs spatially. Sampling must reflect both these aspects.

Sampling detail will depend upon whether the program is for a:

Sampling costs can be quite high, however these costs should NOT pre-determine the number of samples taken and analyzed but should be dependent upon the amount of sampling necessary to increase confidence in the data. These costs should be considered with regard to exploration, development and production decisions that rely on sampling results. The number of samples to be taken should be based on characterization of both the waste and non-waste material. The number of samples will also depend upon spatial variation of waste material. Steffen Robertson and Kirsten et al (1989) developed a curve (Figure 2) that can be used as a guideline to ascertain the minimum number of samples required to characterize each geological unit in terms of acid generation and leaching potential. The curve was developed based on a limited number of sites within British Columbia, Canada as a function of the mass of the geologic unit being sampled. In fact, there are no strict guidelines for establishing sampling protocols. One must rely on professional judgement to select the best method(s) and be prepared to defend those methods.

Sample Types and Sample Preparation

Material sampled can be divided into such categories as overburden (unconsolidated sediments; saprolite in tropical weathering environments), waste rock and tailings. Rock samples should be obtained from both drill core and outcrop. Samples to be collected should represent all lithologies (Figure 3) that are a) both barren and sub-ore (sulphidic) and b) altered and non-altered in order to: 1) determine limits of acid base accounting (ABA) for acid generating and non acid generating material, 2) determine spatial differences in lithologies and 3) generate enough data points for waste rock block modeling. Waste rock should be defined by a mining engineer. A few ore grade intervals should also be analyzed as a reference. In general, waste rock composites represent 10 to 15 metres lengths which conform to mineable bench height. Overburden samples should also be composited.

Figure 3: Note change in color due to alteration, each unit must be sampled

Sampling programs for a new mine site are likely to be different than for an existing site. The extent of sampling in an existing site will depend on the available information and the familiarity of site personnel with the geology and distribution of mine rock around the site. The objective of the sampling program will also affect the program. For instance, in an existing site, sampling may be done to characterize mine rock types as well as to determine the extent of oxidation, storage of oxidation products and rock classes within existing waste rock or tailings facilities.

The size of each individual sample is dependent on the heterogeneity (in particular the distribution of sulfide and alkali minerals) of the unit being sampled. For static testing a 1 kg sample of a homogeneous unit, at a minimum, is recommended by Steffen Robertson and Kirsten et al (1989)

Types of Sampling

When deciding on sample spacings and sampling procedures, one must be fully aware of the different levels (or scales) of variability with sampled materials and adjust sample intervals/locations accordingly or use composite samples to represent wide intervals in order to attain a measure of confidence in the intervals between sample locations.

Sampling can consist of various types:
Point samples: This is can be a single grab sample chosen to represent some mass; or it can be random samples taken from various source points, generally within a predetermined area, and can be either in two dimensions or three dimensions (ie. dump pile) and composited.
Linear samples: Continuous sampling over an interval in a line such as channel samples (Figure 4) or drill hole samples, (Figure 5) and profile sampling of overburden (Figure 6).
Panel samples: These are planar samples made up of multiple chips collected from a surface with dimensions (i.e. one by two metres, Figure 4).
Bulk samples: Sampling of a large mass of material that will be crushed and split into fractions. Samples may be taken from the various splits.

Figure 4: Linear (a) and Panel (b) Sample Examples.

Figure 5: Drill hole sampling to generate mineable blocks of ore and non ore material. Profile view.

Figure 6: Profile sample.

Sample Sources

Potential sources for samples for ARD test work at new mines could include:

In the exploration stage however, drilling programs typically focus on orebody definition and therefore additional drilling may be required for detailed characterization of representative waste rock. For existing mines, there are a larger number of sources from which samples could be taken, including:
River and Lake Sediment Sampling

In areas affected by mining, it is often advantageous to sample rivers and lake sediments. These sediments can play an important role in removal of metals and other contaminants from water. Conversely, they are also a potential source of metal release and toxicity to aquatic organisms that live and feed at the bottom of rivers and lakes (Steffen Robertson and Kirsten et al (1989)).

Another source for sampling is soil and stream banks (Figure 7). These areas may be more accessible than river and lake sediments and will reflect the geochemistry of underlying lithologies.

Figure 7: Sample banks (soil and vegetation) on either side to determine possible migration of elements

This type of sampling is typically not a regular part of an ARD sampling or monitoring program. It may however be worthwhile in areas where mine rock is being deposited directly into a water body or where drainage from a mine impacted area carries a contaminant load into receiving waters. The reader is referred to "Sampling for Water Quality" (Environment Canada (1983)) or the "Handbook of Stream Sampling for Waste Load Allocation Applications" (Environmental Protection Agency (1986)) for detailed descriptions of sampling methods for this type of material.

Sampling Equipment

The types sampling equipment are important when planning and conducting a sampling survey. These range from hand tools such as geological picks, scoops and shovels to motorized augers and drills. Different equipment will produce different sizes of sample. Be aware of contamination of samples from the use of sampling tools (e.g. lubricants, paint etc.).

Field Test Work

Field test work of the samples should include paste pH; total dissolved solids, pH and temperature of water (if present such as stream); photographs of field sites; precise location of site using global positioning system (GPS), grid or other survey method and detailed descriptions of material sampled (Figure 8).

Figure 8: Field test work.

A prior knowledge of overburden (soils) and geological units is very important in determining field sample sites. If such information do not exist, then one must be prepared to carry out field mapping.

An often overlooked field measurement is bulk specific gravity (density), which is used in estimating the total metal content of a deposit. It is important, specifically in ARD generating waste material, in determining the number of tonnes that must be removed and handled. The waste rock will have an economic cutoff grade and so waste material may contain some metals. Generally, each lithologic unit will have a different specific gravity due to variation in mineralogy. There should be enough measurements that a reliable estimate of the error on the mean value can be calculated. Bimodal or skewed distributions of specific gravity measurements indicate the presence of domains or multiple populations and the need to review the various geological categories.

Sample Blending for Waste Piles

In some places there may be a need to create blended waste rock piles/dumps in order to control acid rock drainage. Sampling of material that will be used in the dump is extremely important so as to control the correct blending of waste material. As in the case of systematic grade control, there should also be some systematic waste rock ARD control. Material sampled will most often come from blasthole cuttings. As a result of inappropriate waste rock sampling and subsequent blending practices, the resulting blended pile will be three-dimensionally heterogeneous and the composition of each pile (or sections within the same pile) will be known with poor accuracy and precision, and the composition may vary from pile to pile.

Sample Handling and Storage

Most samples when collected are put into plastic sample bags or kraft paper bag in the case of soil and sediments. However, there are times when field samples may need to be submersed and shipped in water or cold container in order to prevent any initial rapid oxidation. Such samples may also need to be stored under these conditions until such time that they are not needed. It is generally during the sample handling that contamination of the sample may occur.
Samples should be well labelled with location and type of sample at a minimum. The date and sampler's initials are also recommended. Sample locations should be noted on a site plan and staked or marked in the field where practical. With the advance of GPS units, it is possible to collect coordinates during sampling for recording in a logbook. Photographs of sample locations are also extremely useful.

Chain of Custody

A very important aspect of sampling is the chain of custody from sample collection to shipping and analysis and storage. This chain must be maintained in order that any source of contamination and/or errors can be identified and assessed.

Sample Preparation

Sample preparation is the next important step in sampling, whether it is done in the field or laboratory. It is during the sample preparation that some contamination of the sample may occur. All field and laboratory sample preparation procedures must be documented in detail in the ARD report.

Sample Composite Procedures

Chemical Analysis

Before analyses are carried out, the ARD practitioner must discuss with the analytical laboratory the appropriate chemical analyses and methods of sample preparation and digestion. The amount of sample used in the analysis is very important as variation in sample mass may cause analytical variation.

Geological Analysis

Geological sampling will include both regional and detail mapping and sampling. This should also incorporate overburden and soil mapping and sampling, as these materials may be used for construction purposes and/or reclamation.

Mineralogical Analysis

Neutralization Potential values alone can never represent the compositional, structural and textural nature of waste material. A detailed mineralogical study will provide insight into grain size distributions, mineral assemblages, spatial variations in assemblages etc. Mineralogical mapping and sampling must be carried out in conjunction with the geological analysis.

Pilot and Milling Plant Sampling

The milling plant is the best place to obtain a representative sample of both broken rock and tails. The broken rock material is usually crushed to 35 mm to 20mm before entering the grinding circuit. Tailings will be produced after the benefication stage. Samples should be taken on a regular basis so as to obtain information of material going to the heap leach pads or to the tailings impoundment.
In the pilot plant, testing of the material will achieve two purposes: 1) ARD characterization of potential leaching material and tailings and 2) determination of sampling procedures for the milling plant.

Data Presentation, Interpretation and Reporting

The various analytical methods applied to samples will produce numerous data. Extreme care should be taken in the application of statistical analysis. Poor understanding of the statistical aspects of data evaluation will lead to wrong (potentially expensive) conclusions. The question that needs to be addressed by the ARD practitioner is "How defensible are my data?".
The ARD report must include all aspects of sampling so that the regulator (and other interested groups) have some confidence in the reported results, interpretations and conclusions.

Qualified Personnel and the Sampling Program

Sample programs must be conducted by qualified people, who must have some geochemical training. This aspect is very important as all too often unqualified people take samples without any input from a geochemist.


Good sampling is the basis of good data, good interpretation and viable conclusions. Qualified people should be involved in the design, planning and supervision of a sampling program that will give regulators confidence in the results.


ŠThe contents of this web page are protected by copyright law. Please contact the authors for permission to re-use the contained information.



Bloom, L. (1999), Third Party vetting of Geochemical Programs or Return on Quality, in short course Quality Control in Mineral Exploration, presented at 19th International Geochemical Exploration Symposium, April 11, 1999, Vancouver, B.C.

Environment Canada, Water Quality Branch (1983), Sampling for Water Quality. Inland Waters Directorate, Ottawa

Environmental Protection Agency (1986), Handbook: Stream Sampling for Waste Load Allocation Applications. EPA/625/6-86/013, September 1986

Francois-Bongarcon, D. (1999), Extensions to the Demonstration of Gy's, Formula, Exploration and Mining Geology, vol. 7, nos 1and 2, pp 149-155.

Gy, P. (1998), Sampling for Analytical Purposes, John Wiley, New-York and London, 150 p.

Gy, P. (1992), The Sampling of Heterogeneous and Dynamic Material Systems, Elsevier, Amsterdam, 653 p

Gy, P. (1982), Sampling of Particulate Materials, Theory and Practice, Elsevier, Amsterdam, 431 p.

Sketchley, D.A. (1999) Gold Deposits: Establishing Sampling Protocols andMonitoringg Quality Control, Exploration and Mining Geology, vol. 7, nos 1 and 2, pp129-139.

Stanley, C. (1999), Treatment of Geochemical Data: Some Pitfalls in Graphical Analysis, in short course Quality Control in Mineral Exploration, presented at 19th International Geochemical Exploration Symposium, April 11, 1999, Vancouver, B.C.

Steffen, Robertson and Kirsten (1992), Mine Rock Guidelines Design and Control of Drainage Water Quality. Report prepared for the Saskatchewan Environment and Public Safety, Mines Pollution Control Branch, April 1992.

Steffen Robertson and Kirsten (B.C.) Inc., in association with Norecol Environmental Consultants and Gormely Process Engineering (1989). Draft Acid Rock Drainage Technical Guide. Prepared for British Columbia Acid Mine Drainage Task Force, Volumes I and II.

Thrush, P.W. et al., Dictionary of Mining, Mineral and Related Terms, US Bureau of Mines Special Publication, Maclean Hunter Publishing Company 1990

Vallee, Marcel (1999), Sampling Quality Control, Exploration and Mining Geology, vol.7, nos. 1 and 2.

Return to ARD at EnviromineHome pagee