3.2 Life-of-mine planning and management

Life-of-mine planning for monitoring requires a predevelopment impact register to be formulated and risk assessment procedures to be carried out, as described in Section 3.1. Once all potential future impacts have been anticipated, monitoring systems can be designed and put in place to take account of them.

3.2.1 Baseline monitoring

Where it is possible to incorporate baseline monitoring (for example, in greenfield projects and expansions to mines), such monitoring is a critical component of leading practice monitoring programs. Baseline monitoring should commence at the pre-feasibility stage and include all relevant environmental, economic, and social issues identified in risk planning. Typical elements of monitoring programs are listed in Appendix 2.

In most cases, the baseline monitoring system will become incorporated into later monitoring over the life of the mine so that repeat assessments can have a consistent basis of comparison. This will provide essential data on several aspects that are not necessarily related to impacts of the operating mining project but serve as a reference such that:

  • the extent of natural system variability can be quantified over time and space
  • the extent of pre-existing impacts from previous or current mining projects can be used to place the contribution of the operation into context
  • the effects of other causes (such as urban or agricultural inputs) can be distinguished from those of the mining operation.

Baseline data, together with ongoing monitoring of reference (or control) sites, is essential for being able to correctly interpret the results from monitoring programs that have been designed to assess the extent of mining project–related impacts during mine life and the extent of recovery or improvement following control of the impact or rehabilitation.

3.2.2 Design principles for monitoring

A common and recommended approach for assessing impacts and recovery is the use of the ‘before– after–control–impact’ (BACI) approach to monitoring design (Quinn & Keough 2006) and derivatives of it (such as modified BACI). The ‘before–after’ component refers to measurements conducted before and after any change that might cause an impact. ‘Control–impact’ refers to measurements conducted in areas assumed to be unaffected (‘control’) or potentially affected (‘impact’) by the project. It is important to note that impacts can include direct impacts, secondary impacts and cumulative impacts.

The principle of the BACI design is that the operation’s effect on the environment is assessed by determining the difference between results measured before and after the impact at sites potentially at risk, and comparing this with similar sites not likely to be at risk. It is focused on the relative differences of the control and impact sites before and after operational changes occur, not on trends at individual sites. This approach provides much-improved statistical power to differentiate project-caused environmental responses from other sources of variability in the measured indicators.

Note two points:

  • While the ‘control’ and ‘impact’ sites should be similar in their physical and ecological characteristics, it is not necessary and not always possible for them to have identical attributes. However, the differences between the sites must be able to be measured both before and after the possible impact.
  • When comparing before and after data, if there is an increase in the differences between the ‘control’ sites and the ‘impact’ sites, that may indicate that the project has had an impact. It is this difference that can be measured and used to statistically determine whether an impact has occurred. Ideally, this should include the measurement of baseline differences before the mine is developed, but the principle can be adapted to later assessment of any changes in the differences between ‘control’ and ‘impact’ sites.

In practice, the BACI monitoring design that is needed for a specific project can sometimes be more complex than this, although the principles remain the same. Further details for the design of a conventional BACI monitoring program are given in Quinn & Keogh (2006) and Underwood (1991), while design of the modified, multivariate BACI approach is described in Humphrey et al. (1995), Humphrey & Pidgeon (2001) and Faith et al. (1995).

Although well established as a robust, statistically based monitoring approach decades ago, BACI remains underutilised in most monitoring programs despite its clear analytical advantages. It remains the leading practice monitoring design where it is feasible. However, where practical, using conceptual models, as opposed to relying only on statistical models, can also prove very useful.

Other impacts that are not necessarily directly related to mining operations, but that may subsequently occur as a result of increased population in the vicinity of the mine, may need to be taken into account in monitoring design. They can include slash-and-burn agriculture and artisanal alluvial mining (both of which may increase sediment load upstream of the mine); other industrial development; dust storms, bushfires or forest fires affecting air quality; or previous logging, hunting or clearing activities affecting biodiversity. Consideration of these impacting processes will require some type of comparison with control and/or reference sites, even if not in a formal BACI framework.

Regardless of whether or not the BACI approach can be used, leading practice monitoring programs should be designed to be cost-effective and be based on sound statistical and social science research principles. The key to good monitoring design is to base the design on statistical principles, rather than trying to fit statistics to the design (but note comments above on the role of conceptual models in some instances). This will help to avoid bias in sampling and enable appropriate sample sizes and sampling frequencies to be calculated and optimised in advance. Leading practice monitoring programs commonly take statistical power into account, thereby ensuring that, if an effect occurs, there is a high probability that it will be detected at a meaningful early-warning effect size—not only after substantial environmental detriment has occurred.

While parametric statistical analysis using normally distributed data is preferable when determining whether impacts have occurred, high variability or low sample sizes (such as when monitoring rare or threatened species) may prevent its use in practice. In such cases, the monitoring program may be designed according to non-parametric analysis procedures, making use of modern robust, generalised or Bayesian statistical procedures that are more able to reliably analyse limited datasets and datasets that otherwise do not comply with the underlying assumptions of standard classical parametric statistical analysis procedures. In all cases, visual inspection of trends in data is very important, and in some instances it may be possible to do no more than observe trends graphically using a control charting approach.

However, this can prove to be a very useful practical tool in understanding what is happening, initiating management action and communicating the results of monitoring to the community. The development of R as a free statistical analysis tool and the wider availability of Bayesian analysis tools have made the adoption of more advanced statistical analysis of monitoring data a relatively simple matter.

Whatever the case, it is essential to include consideration of what analyses will be carried out when designing the monitoring program. Green (1979) provided a list of 10 principles that should be taken into account (see Appendix 1) and noted that ‘if you have delayed seeking expert advice until you can only ask “what can I do with my data”, you richly deserve, at that point, any answer you get!’ This remains true three decades later. Consideration of data analysis requirements at the design stage can result in much more cost-effective monitoring programs by providing guidance on sampling locations; the intensity, frequency and duration of sampling; the amount of replication; and other key aspects. Importantly, it will prevent wastage of funds spent on monitoring that has low statistical power.

Mine planners must be consulted when designing monitoring programs, and in some instances they should participate in the design of the monitoring that is required (where it is part of mine operations).

The environmental monitoring program and the resulting output data should be linked to the mine’s spatial or geographic information system (GIS) and should be accessible and highly visible on the system. In this way, when mine plans change (as they frequently do), those responsible for the monitoring system have early notice and can take action to ensure that impacts due to changes in mining operations continue to be appropriately monitored and managed. A particularly important example of this is where monitoring sites are damaged or destroyed by mine development and continuity of the data record is lost. Good knowledge of the importance of such sites ensures that it might be possible to maintain them by proactive mine planning or, if not, new sites of similar monitoring effectiveness can be established as soon as possible. Ideally, the new sites should be established and running before the original sites are lost so that it is possible to have a period of monitoring overlap. A good GIS should not only store locations of monitoring sites, but also be capable of displaying overlays of monitoring trend data.

At some sites, operators decide (or are required) to develop or modify a monitoring program without the benefit of baseline monitoring and a predevelopment risk register. This can happen when a company acquires an existing operation that does not have that data, when mining operations recommence in the vicinity of old abandoned mines, or when a decision is made to significantly modify or upgrade the existing monitoring program in line with current regulatory and community expectations. In these instances, careful thought will need to be given to the design of the monitoring program using the principles described above where possible. Approaches such as monitoring nearby reference and/or control sites, monitoring upstream versus downstream, and determining whether previous owners or regulators have conducted monitoring can help in designing an appropriate monitoring program. In situations where a number of mining operations are present, including closed or abandoned mines for which no company has ongoing management responsibility, monitoring in conjunction with regulators may be needed to distinguish the impact of the active mining operation from those other (cumulative) sources.

The monitoring program should carry on through the entire life of the mining project, including rehabilitation and closure (as discussed above, the detailed content of the program will change through time). Post-closure monitoring will also be required where impacts have the potential to be high risk, long term, or both (for example, where drainage from the mine may be acidic or contaminated with metals).

The design and duration of post-closure monitoring and responsibility for conducting it should be determined by agreement with relevant regulators. Once the mining company has demonstrated that the rehabilitation has been completed satisfactorily and is performing as required, it usually requests the relinquishment of the lease back to the state (note that Western Australia and the Northern Territory have introduced levies that may partially contribute to remediating existing and future mining legacies). Once relinquishment has occurred, post-closure monitoring may be carried out by regulators rather than by the mine or consultants, provided an agreed source of funding is available.

Overall, it is essential that monitoring programs be designed according to the defined risk and potential impact of the project, and be capable of detecting all relevant impacts, including those that are positive. They must have clear objectives and, where possible, should be quantitative or else incorporate qualitative data that complies with sound statistical design and analysis principles and is in a form that can be replicated in each stage of the mine’s life.

For social performance monitoring, objectively measurable datasets, such as local employment statistics, changes in regional health profiles or surveys of household income and expenditure, may be complemented by well-designed qualitatively focused monitoring. For example, monitoring could include tracking career progression for indigenous employees and the factors that influence employment outcomes, the essence of which cannot be captured in quantitative data or statistical analysis alone.

Share this Page