CHAPTER 10: SAMPLING

Sampling can also measure how the environment is changing over time. Data from sampling is necessary to understand how mining activities and mine waste changes the environment. Sampling is designed, carried out and assembled to structured information by skilled professionals. Sampling results are interpreted into information that can be shared with all stakeholders at a mine site. Data derived from poorly planned or inadequately carried out sampling can misrepresent and misinform. Poor sampling can hinder prediction and monitoring of environmental impacts.

Resources and transparency

Economic or human resources available for data acquisition and sampling are sometimes insufficient. The lack of resources can be due to a range of reasons, from national budget deficits to shortcomings in environmental legislation or defunct and abandoned mine sites. To what degree resources are limiting data collection can vary but the design of a sampling campaign is always a balance between data requirements and of warranted costs. Not only economic resources are required to collect, organise, and analyse sampling data. Experienced professionals and freedom to use, analyse and share data is also necessary to extract and share information from sampling.  The design, execution, and analysis of sampling, including sample results, must be transparent, available for third-party review and ideally available for research and long-term evaluation.

The value of small sampling campaigns

When resources are insufficient also small data sets from limited sampling can be of great value, but all limitations of the sampling data must be clearly stated. Limitations of sampling data can be sample representativity, sampling method or sample series continuity. Smaller sampling campaigns can add relevant information if limitations of collection or analysis are known and described. When resources are scarce data collection can include basic techniques such as well documented field visits, visual inspections, and photography. Simple methods of data collection will add to the information if observations are structured, stored, and analysed. The aim of sampling is often to study and document changes over time at a location. Recurring and consistent observations and sampling is of value whether the sampling program is large and complex or if sampling is limited. At some point however the lack of quality data makes predictions or monitoring very unreliable.

Sampling plan

A sampling plan should be in place before carrying out sampling in the field. A plan for sampling need to include:  identified involved parties (stakeholders), objective and purpose, background, extent, sampling strategy, sampling methods, sampling handling, sample analysis, quality insurance, documentation and working conditions.

Stakeholders; operators, consultants, regulators and people living close to or at the site, should be identified in a sampling plan. Effective and reliable sampling can depend on how and with who field work is conducted who the end user of sample results are.  A definition of general objectives and purposes (i.e. water treatment) will decide the quality and extent of the information needed. Objectives could be multiple during a Life of mine (LOM). In any way, it is of great importance to collect background information and undertake field inspections for the development of valid sampling instructions. This information could include valid prior information i.e. data from previous exploration work, former land uses, background concentrations and the evaluation of old data (type and consistency of sampling, analytical procedures, objective of previous sampling). A field inspection will serve to get information on local conditions, detect physical restrictions (accessibility, i.e. rain season), relevant environmental factors (precipitate colour, seepage, erosion, dust, smell) dependent. Sampling during the stages of a mining project can include:

Exploration and mine planning phase:  baseline study, screening, characterization, supporting EIA, supporting waste management and rehabilitation plans.

Operational phase: continuous characterization of mine waste, general environmental monitoring for compliance, evaluation and refining of EIA parameters, general environmental monitoring of air, water, soils, sediments and organisms, evaluating and optimising waste management and rehabilitation plans.

Mine closure and Post-closure: continuous mine waste characterisation,  establishing post mining water balance and water quality, sampling for compliance, sampling to measure if short-term and long-term objectives are met.

Strategy for sampling of mine waste and mine water

During exploration and mine planning there are multiple sampling objectives for the ability to control and identify:

• Hydrology (surface water and groundwater, up-/downstream and recipients), habitat control (biological measures), water balance for the site (flows, pump rates, precipitation, evaporation, infiltration)
• Mineralogy (drill cores, exploration drilling datasets, pilot scale processing, possible formation of ARD)

Samples of water should be collected from all water bodies potentially affected by drainage water from the mining activity. Analysis of pH in the water should be conducted in the field, and in filtered and unfiltered samples. Sampling extent is to be evolving by dynamic evaluations. The number of samples should increase by complexity of ore and the host rock geology.

During operation, sampling results from mine planning and exploration should be used, to increase validity of the sampling programme. As the mine is in operation, these sampling efforts are needed:

• Sampling of ore, waste rock, mine water and tailings: Ideally sampled for possible compliance with the results from the sampling campaigns conducted during exploration and mine planning.
• Sampling of process water (Cyanide, lime, flotation agents, toxicity)
• Process effectivity: Exchange of metals/minerals of interest, possible formation of ARD

In general, it is of major importance to control flows of water and waste at site, instrumentation and monitoring of open pits, underground mining, waste facilities could be:

• Geotechnical (Safety, environment, health)
• Geochemical (pit and underground water, seepage) - Compliance regulators
• Ongoing characterization (ABA-test, Humidity cells). Mineralogy/iron phases, grain size, material type, erosion (dust, water and air).
• Seepage, ongoing recipient control (biological receptors). Results used in closure phase

Mine closure and Post-closure: A majority of the sampling efforts conducted during the operational phase are continuing such as:

• Ongoing characterization (I.e. ABA-test, Humidity cells).
• Erosion, Mineralogy (oxidation fronts).
• Seepage from waste facilities, ongoing ground water and recipient control (biological receptors).

These sampling efforts should continue until compliance with the limit values (in recipients) stated by the authorities are withhold and no further risk to the environment are anticipated. A normal timespan for this is 30 years or more.

Representative sampling

Waste rock
Representative sampling of waste rock is challenging, there are many different approaches and suggestions. In general, it is a good idea to separate the waste rock into different particle size fractions (i.e: < 2 mm, 2-12 mm, > 12 mm) and try to detect if there are any mineralogical differences. If not, the finest particle size is sampled for analysis. Regarding number of samples/amount of waste rock, there are few guidelines for this matter, in general, samples should have a good spatial representation (vertical and horizontal) of the mining area. This kind of sampling campaigns could end up with several hundreds of samples as complexity of the ore and host rock increases. There´s a recommendation from Swedish, Canadian and Danish authorities, for representative sampling, namely:

N = 0.026 × M^0.5 (N=Number of samples, M = mass in tonnes) (I.e. 10 million tonnes = 80 samples)

In this case, the equation is adaptable if there´s no prior knowledge about the waste material. Each of these samples should have a wight of > 5 kg and contain 15-30 sub-samples.

Tailings
Sampling of tailings is often conducted by dividing the TSF into grid-sectors that are sampled. There are few guidelines for this matter, although, some studies claims that the number of samples needed for sufficient representativity could be 15-30 samples/TSF. 

Water and sediment
Baseline Studies are conducted to understand the environmental setting prior to mining, during the exploration and mine planning phase. It provides the basis for impact evaluation, mitigation measures and assessment of risk to maintain the quality of land and soils. In lakes, sediments, streams and ground water that may be affected from the mine sites, samples are to be taken:

• Lakes (depth < 15 m), at one sampling point, 3 depths (surface, thermocline, bottom)
• Lakes (depth < 5 m) surface sample at minimum 0.5m depth
• Streams: Reflect seasonal and operational differences (i.e. flow-based)
• Sediment: 5 samples in lake-areas < 1km²
• Ground water: Quarterly (early warning)

These sampling points are to be firm for long-term monitoring. Continuous monitoring of pH, E_c and temperature is often used to set sampling frequency. Sampling of water at mine sites is conducted to control diffusive flows, or the water quality at constructed discharge points. This is done at a given time-interval, often agreed upon with the authorities.

Sampling for characterization

Prior to most tests, samples of tailings and waste rock should be air- (at less than 40°C) or freeze-dried, and the moisture content and particle size distribution should be measured as part of the sample preparation. Prior to and after drying, samples should be kept cool and dry. This is however not applicable for some biological and toxicity tests and waste materials that is to be stored in an anoxic environment. For that matter, waste materials needs to be kept in anoxic conditions, before testing. Tailings sludge (preferably taken directly from the tailings outflow) should be measured for pH, electric conductivity (EC), oxygen and redox potential (Eh) in the slurry as it is.

Geochemical test program for characterization

A geochemical test program is generally conducted to predict the future quality of leachate (metals and elements that could leach out) if the mine waste could be potentially acid-generating. In most geochemical test programs, testing will start with static tests, that often include some kind of screening, to see if the mine waste could produce ARD, but also to appreciate metal leachability and chemical composition. Static tests are used in the first stage of the test programs measuring the quality and quantity of different constituents in a sample at one point in time or during a very short time span (less than 24 h).

Table 1: Overview of important static tests for geochemical test work and their typical application in mining projects.

In many cases, if the static tests implies that the mine waste is to be considered acid-generating (or uncertain to be acid-generating), kinetic tests will follow.  Kinetic tests could give information on how the material will react over time on a long term basis (i.e. reaction rates, release rates of elements, time for onset of net acid production). There´s a number of these tests, but in general, test should be tailored on conditions  specific for the mine site.

Table 2: Overview of important kinetic tests for geochemical test work and their typical application in mining projects.

However, results from the kinetic tests will not reflect field conditions, for that matter up-scaling of the test results from laboratory scale to a full-scale mining situation will have to be performed. This is typically conducted by multiplying the test results with a cumulative scaling factor (CSF), that takes into account differences in i.e. particle size, temperature, oxygen and water availability, infiltration rate. 

As the kinetic tests are performed, up-scaled test results should be included in model predictions of the mine drainage, that is to be a part of the EIA. Actual volumes of tailings, waste rock and ore for each year should be included in the model, to be able to make predictions on an annual basis during the LOM and after mine closure. Predictions should be made based on up-scaled kinetic test results (using representative samples). The predictions should reflect a most likely scenario (i.e. with a 95 % confidence interval) or a worst case scenario, based on the highest leaching samples.