Appendix G. Extreme Events
This section provides an overview of extreme events that may impact on mining operations, and subsequently on nearby streams. There is limited research on the impact of extreme events on streams, and these impacts will be highly variable for the same type of event. Similarly, it is often difficult to gauge the likelihood of the various events occurring. For these reasons, this section is based on the current state of knowledge and is presented in a qualitative rather than quantitative format.
Extreme events that may result in additional impacts on aquatic systems from gold and coal mining can be divided into events directly associated with mining activities (e.g. treatment system failure, tailings dam failure, cyanide spill) and natural events (e.g. extreme high rainfall events, landslide, drought). This appendix provides a discussion of the likely impacts arising from various extreme events, prevention and treatment activities, as well as consideration of how extreme events can be taken into account during mine planning.
G.1 Mining-related extreme events
G.1.1 Treatment system failure
Treatment system failure may occur as a result of poor fit of treatment system type to site, under-design of treatment system, mechanical failure, or inadequate operation and maintenance.
Selection between active and passive treatment and design of system within each category are critical to achieve required water quality. Active treatment systems are much more forgiving to inadequate design, as modules can be added to the system to correct most shortcomings. Passive systems, however, are much more difficult to rectify if performing inadequately since the system type might not be appropriate for the site and an entirely different system may need to be constructed. For example if an anoxic limestone drain is constructed for an AMD with high concentrations of Fe3+, armouring of limestone and plugging of pore spaces with precipitates can result in complete failure of the system and a new system, such as a vertical flow wetland (VFW), may need to be constructed. Poor fit of passive treatment system type to site chemistry, flow rate, and topography is the most common form of treatment system failure and is an area of active research.
Mechanical failure is more likely in active treatment systems rather than passive systems since passive systems generally do not have any moving parts. Active treatment involves the operation of many pumps, valves, metering devices, mixing paddles, etc., each of which has a definite lifespan and can fail prematurely. Passive systems usually have piping and valves to convey water to and through the systems. Examples of mechanical failure in passive treatment systems include build-up of rust or precipitates on valves limiting operation, build-up of precipitates in piping clogging passageway, and breaking of piping due to excessive overburden pressure.
Inadequate operation and maintenance can occur in both active and passive treatment systems. Active systems are much more labour intensive than passive systems and typically have personnel on site on a daily basis. As such, treatment system failure due to inadequate operation and maintenance can usually be rectified quickly. Passive systems, however, may not be visited for up to a month or more, so a failure may have a much greater impact. Typical operation and maintenance requirements for active and passive systems are included in Chapter 8 (Decision making and monitoring).
Treatment system failure for active treatment systems can result in discharge of water with chemistry similar to untreated water or can have worse water chemistry. If excessive neutralising agent is added, the pH can be elevated much higher than 7 and unreacted neutralising chemicals can be released in the discharge water. Depending on the chemicals used, this can result in excessive sodium, ammonia, or chloride in the water. Excessive flocculant or coagulant can occur in the discharge water as well.
Failure of passive systems usually results in discharge with chemistry similar to untreated mine discharge, but in some cases can have worse water chemistry. Some systems, such as slag leaching beds (SLBs), can raise the pH higher than 7, and some systems, such as SLBs and VFWs, can release other elements into the water. SLBs can contain numerous elements that were impurities in the iron ore and VFWs can contain various pesticides and metalloids (such as As) in compost.
Impacts on stream communities
The impacts on stream communities will depend on what the treatment system was treating, for example, suspended solids, pH, or trace elements. The release of sediment from suspended solid treatment-systems may result in physical choking of the stream and its biota when deposited. Sediments with the finest particles, when dried on stream banks, can form a hard deposit that radically changes the nature of the streambed. Fine particles also infill spaces between large substrate within the wetted streambed, reducing the quantity and quality of habitat available for macroinvertebrates.
Prevention
The best prevention is to select and design the most appropriate treatment system for the site and to conduct adequate operation and maintenance. Treatment systems should have alarms to notify operators of a failure and, if possible, excess capacity should be built into the system to contain untreated water on site.
Treatment
Treatment after treatment-system failure depends on what water chemistry was discharged from the site. If metals or metalloids are discharged only in dissolved form, then there are no remedial activities that can be conducted in the receiving stream. However, if particulates are discharged into the stream, these may settle on the streambed and alter the aquatic ecology. Artificial or natural treatment may be required.
Artificial treatment involves collection and removal of the precipitates to a repository or back into the treatment system. This process can be disruptive to the stream, as access roads are needed and the streambed must be disturbed with earth-moving equipment.
Natural treatment consists of allowing rain events such as regular floods to remove the precipitates from the stream. Remnants of precipitates are progressively colonised by the stream biota.
G.1.2 Tailings dam failure
Tailings dam failures are surprisingly common around the world, so possible failure of a tailings dam on a New Zealand mine site has to be taken seriously. Examples of tailings dam failures, the reasons for the failures, and consequences, around the world are documented on the following website: http://www.wise-uranium.org/mdaf.html.
The tailings dam failure event that has attracted the most scientific study is that of the Los Frailes mine, Aznalcollar, Spain, in 1998. An extensive scientific literature is available on this event (starting with The Science of the Total Environment, volume 242, 1999, Elsevier), and new research is still being published.
Failure of a tailings dam will typically involve dispersal of large volumes of tailings as solid-rich slurry into the nearest stream. The tailings settle progressively on the bed and banks of the stream, with coarsest material (sand and larger particles) near the dam, and finer material (silt and mud) further downstream. Sediment transport can be for tens of kilometres.
Impacts on stream communities
The sediment, when deposited, may result in physical choking of the stream and its biota. Sediments with the finest particles, when dried on stream banks, can form a hard deposit that radically changes the nature of the streambed. Fine particles also infill spaces between large substrate within the wetted streambed, reducing the quantity and quality of habitat available for macroinvertebrates. Alternately, chemical dissolution of elements such as As may occur, or if the tailings contain pyrite, oxidation of that pyrite may result in AMD issues.
Prevention
Prevention of tailings dam failure should be a key aim in original design. The following points are significant in this regard:
- Construction material. Many overseas tailings dams are built with dried tailings, and these are most prone to failure. Construction of dams with a core of waste rock is more likely to produce a long-term stable structure.
- Site selection. New Zealand has steep topography, and many hillsides are underlain by active or inactive landslides. Loading of these hillsides with dams and their contained tailings, and earthworks associated with site construction, can activate or enhance landslide activity. The Golden Cross gold mine (Coromandel, NZ) was found to be sited on an active landslide, and major treatment work was required to stabilise the structure to prevent dam failure in the 1990s. This work is ongoing, after the mine has closed.
- Water management. Many tailings dam failures result from water erosion and related phenomena during extreme rain events. The Opuha irrigation water dam (South Canterbury, NZ) failed in a major rain event during construction in 1997, suddenly releasing large volumes of water downstream (Lees and Thompson 2003). Diversion channels should be designed to cope with such extreme rain events.
- Height additions. Most tailings dams, unlike dams for irrigation and hydroelectricity generation, are built progressively, with additional height added as the tailings levels rise. Problems with construction of these additions have caused many tailings dam failures.
The International Commission on Large Dams has a large number of technical publications related to all aspects of dam design, construction, operation, monitoring and maintenance, including UNEP (2001) and USEPA (1994). These documents are available for purchase at http://www.icold-cigb.org/.
Treatment
Treatment activity after a tailings dam failure is dependent on the scale of the failure, the thickness of sediment resulting, and the potential chemical effects of that sediment.
Artificial treatment. This typically involves collection and removal of the tailings to a repository, normally the repaired dam from which they came. The treatment process can be disruptive to the stream, as access roads are needed and the streambed must be disturbed with earth-moving equipment.
Natural treatment. This mainly depends on natural rain events such as regular floods to remove the tailings from the stream. Remnants of tailings are progressively colonised by the stream biota. The process can take decades (see Black et al. 2004).
Monitoring of the success of treatment of streams after a tailings dam failure is greatly facilitated if the geological, chemical and biological state of the affected stream was known before the tailings arrived. This background information provides a baseline against which the treatment activity can be measured. Many mines are developed in areas where mining has occurred historically, and there are natural geochemical anomalies around such sites, so these streams are typically different from surrounding streams. Hence, clear definition of background data for streams that could be impacted by tailings dam failures constitutes important insurance for mining companies and regulatory bodies.
G.1.3 Cyanide spill – hard-rock gold mining
Routine use of cyanide in a hard-rock gold processing plant has little downstream environmental significance because cyanide residues in tailings are at low levels (see section 6.2). However, concentrated cyanide compounds are transported to the mine site and regularly handled on-site. Hence, there is potential for accidental spillage of this material because of:
- Transporting vehicle accident;
- Cyanide container leakage;
- Cyanidation-plant pipeline failure (see section G.1.1);
- Failure of ‘cyanide kill’ system, leading to high-cyanide tailings (see section G.1.1); and
- Tailings dam failure (see section G.1.2).
If this spillage involves discharge to the environment, an extreme ecological event is inevitable. Cyanide spills occur regularly around the world, with more than 30 significant incidents since 2000, including one in New Zealand (http://www.rainforestinfo.org.au/gold/spills.htm).
The most famous and most thoroughly documented recent example is the Baia Mare event (Romania) of 2000 (UNEP 2000).
Impacts on stream communities
Cyanide is highly toxic to most animals, so a cyanide spill into a stream is likely to kill all or most species present almost instantly. This will occur for several kilometres downstream until sidestream dilution lowers the cyanide concentration sufficiently for animal survival. The cyanide effect will persist in the stream until:
- clean water from upstream or rainfall dilutes the cyanide, and/or
- the cyanide decomposes in ultraviolet light (sunshine).
These mitigating effects can take days to weeks to occur. Ecosystem recovery after a cyanide spill can take weeks, months, or years, depending on the severity of the spill.
Mitigation options
The effects of a cyanide spill can be decreased by use of ‘neutralising’ chemicals that decompose the cyanide faster than UV light. Oxidising agents such as sodium hypochlorite are most commonly used for this purpose. These sorts of chemicals should already be on-site as part of routine cyanide management. Use of these chemicals is likely to be most relevant to an on-site spill or a local stream spill. There is potential for this ‘cure’ for a cyanide spill to be as bad or worse for an ecosystem than the cyanide itself.
It is important that a site-specific cyanide spill contingency plan be developed for the whole cyanide pathway, wherever cyanide is to be transported to a mine site and used on-site (ICME 1999; ICMI 2002, 2005). This plan should involve maintenance of adequate stockpiles of cyanide decomposition chemicals.
G.2 Natural events
For the most part no prevention or mitigation options are suggested for natural events, except for extreme high rainfall events. It is expected that obvious precautions, such as not locating tailings dams or treatment systems in earthquake-prone or landslide-prone areas, would be taken.
G.2.1 Extreme high rainfall events
Extreme high rainfall events are a natural phenomenon, and depending on their frequency, stream systems may be more or less adapted to them.
Impacts on stream communities
Extreme rain events causing flooding may have effects independent to mining, but can also interact with mining impacts to alter the severity of outcomes.
Natural effects on organisms
Stream communities are strongly influenced by natural flood disturbances. Floods have direct impacts on the diversity and density of organisms through removal of individuals, or indirect effects through influencing their habitat or food supply (e.g. removing algae and leaf litter). Consequently, stream communities often undergo a process of recolonisation after flood events. The severity of these events relates to the intensity of the flood, the frequency of floods, and the recovery time of fauna (driven by colonisation rates and the extent of changes to physical habitat). Thus, in frequently disturbed streams, communities may be in a constant state of recovery since the last disturbance event. High-flow events are relatively common in the headwaters of West Coast streams and often have a strong impact on invertebrate density. This physical template acts on top of mining activities, and can potentially confound the detection and quantification of AMD impacts.
Mining-related effects
Flood events may generate a pulse of low pH water with high metal concentrations through an increase in mine discharge. Acute toxicity during flood events may cause mortality that results in more severe impact outcomes than predicted by water quality under normal flow levels. These discrete events are often difficult to detect as water quality sampling is usually taken as spot measurements.
Alternatively, high-flow events may actually decrease mining impacts if increased stream volume dilutes an existing source of AMD. This should result in a temporary increase in pH and decreased concentrations of dissolved metals. However, the benefits arising from pulses of less impacted water is likely to be overwhelmed by the more severe, chronic toxicity under baseflow conditions.
Prevention of additional mine-related impacts
Tailings dams or treatment systems should be appropriately designed so that they can cope with significant rainfall events (World Meteorological Organization 1986).
G.2.2 Landslides
Landslides are most likely to occur on the West Coast, due to the mountainous terrain and steep bedrock underlying the topsoil. Landslides may occur as a result of an earthquake, a high rainfall event, or as a result of destabilisation by mining and/or related roadworks. Depending on where the landslide occurs, it may result in either the deposition of rocks and sediment directly into a stream, or high amounts of sediment being transported to the stream and resulting in highly turbid waters.
Impacts on stream communities
Where rocks and sediment are deposited directly into a stream there is an obvious catastrophic effect on the benthic stream community in that location through smothering. Downstream effects as a resulted of increased turbidity and sediment deposition will also be observed. This sedimentation can persist for some time if the exposed bank or debris continues to be eroded by stream flow. The effect of increased turbidity of the water will depend on its duration. The impacts of turbidity on stream communities occurs through the direct effects of physical abrasion, smothering and irritation (e.g. fish gills), and indirect effects of reduced food quality and quantity, and altered species interactions such as predation (see section 4.3 and Appendix D for more detail).
Prevention of additional mine-related impacts
Ensure that tailings dam or treatment systems are not located in landslide-prone areas. Ensure that appropriate stabilisation works are undertaken during mining operations.
G.2.3 Droughts
Droughts result in the reduction of stream flow, with extreme cases leading to the cessation of flow and the loss of surface water. Droughts cause effects independent from mining, as well as effects that interact with mining to increase the severity of outcomes.
Impacts on stream communities
Natural effects
The severity of drought impacts on stream communities depends on the extent of the reduction in stream flow. Reduced flow causes stress as a result of increased water temperature and reduced dissolved oxygen (DO), potentially impacting less tolerant taxa. In addition, the contraction of wetted habitats intensifies interactions between species, which may reduce the densities of prey species or poor competitors.
When surface flow ceases, isolated stagnant pools form refuges for fish and invertebrates. However, in the absence of flow, animals in refuge pools will experience considerable stress from reduced DO and increased temperature. Salmonid fish and EPT taxa are likely to be most vulnerable to these conditions.
In extreme cases, surface water may dry completely. The resulting effect on stream communities depends on the duration and harshness of drying. Many stream invertebrates and some native fish (non-migratory galaxiids) can burrow into the substrate to reach permanent water underground, and some can even survive for days to weeks in moist conditions under stones. Streams with permeable gravels are likely to have more of these refuges than streams with lots of bedrock or those that are choked with fine silt. Many West Coast headwater streams have large, armoured substrate as frequent floods remove smaller particles. Therefore, West Coast streams are likely to be limited in drought refugia. In contrast, many Southland streams contain loose gravels that provide good access to groundwater zones.
Mining-related effects
Droughts may reduce the dilution of AMD, especially if unimpacted tributaries normally dilute a constant mine discharge. Thus, during low-flow conditions, impacts may increase in severity, and be transmitted further downstream. In addition, stress from increased temperature and low DO may increase the vulnerability of organisms to low pH or elevated metals from AMD.
G.2.4 Earthquake
Earthquakes may result in landslides, and destruction of tailings dams and treatment systems (see Sections G.1.1 and G.1.2). Earthquake damage mitigation is a key part of design of all engineering works in earthquake-prone parts of New Zealand.
Impacts on stream communities
Where rocks and sediment are deposited directly into a stream there is an obvious catastrophic effect on stream community in that location. Downstream effects as a result of increased turbidity and sediment deposition will also be observed. The effect of increased turbidity of the water will depend on its duration.
G.3 Consideration of extreme events during mine planning
The occurrence and potential impact of extreme events should be assessed during the planning of mining operations to ensure appropriate design and operation to minimise potential negative impacts. Further, most of these issues are likely to be required to be considered during the bonding process. As part of the bonding process, a risk assessment of events leading to negative environmental outcomes during mining operations and after mine closure is generally required. This risk assessment can be completed through an appropriate forum of experts, with appropriate background data. A risk matrix can include information on the type of risk, consequences resulting from the realisation of that risk, likelihood of occurrence, and options for prevention and/or treatment.
Extreme events are by their nature unpredictable. Thus, only qualitative assessments, mostly based on professional opinion, of the likely consequences can be made. Similarly, often there are only limited data with which to estimate the likelihood of occurrence and this too will mostly be based on professional opinion. For some events, such as extreme high rainfall events or droughts, a reasonable estimate of the likelihood of occurrence could be obtained from historical rainfall data.
Examples of a risk assessment matrix outlining the likely extreme events relevant for coal mine on the West Coast, and alluvial and hard-rock gold mines in Southland are shown in Tables G1-G3. Additional less ‘extreme’ events that may result in environmental impacts, such as failure of revegetation or spillage of hazardous substances, may also be included in a full risk assessment used for bonding purposes. Bonds will also need to be sufficient to ensure a mine site is appropriately restored and rehabilitated should the agreed remedial measures not be undertaken.
Table G1. Risk matrix of ‘typical’ extreme events associated with a coal (NAF or PAF) mine on the West Coast.
|
Risk
|
Likely consequences1
|
Annual frequency (1 in x years1
|
Annual probability1
|
Prevention, treatment, or mitigation measures1
|
Most likely treatment measure
|
|
Localised Instability
|
Slumping of material into stream
|
5
|
0.20
|
Move access road
|
Excavators and dump trucks to remove sediment
|
|
Global Instability – mine wall, overburden stack
|
Slumping of material into stream, mobilisation of sediment
|
100
|
0.01
|
Re-establish access road and stabilise rocky bluffs. Recontouring of wall/stack.
|
Excavators and dump trucks to remove sediment
|
|
Instability (from earthquake or storm event)
|
Slumping of material into stream, mobilisation of sediment
|
1000
|
0.001
|
Move access road.
|
Excavators and dump trucks to remove sediment
|
|
Reduction in water quality of mine drainage (AMD) – treatment system failure
|
Unacceptable offsite sediment and chemistry discharge
|
100
|
0.01
|
Repair or reinstate treatment system
|
Repair or reinstate treatment system.
|
|
Reduction in water quality (chemistry and TSS) - failure of sediment retention works during storm or instability event (ponds)
|
Unacceptable offsite sediment and chemistry discharge
|
100
|
0.01
|
Reinstate ponds and drainage.
|
Reinstate ponds and drainage.
|
1Actual items will be specific for individual operations, but examples are included to provide an indication of how different events may impact and the relative likelihood of occurrence.
Table G2. Risk matrix of ‘typical’ extreme events associated with a hard rock gold mine in Southland
|
Risk
|
Consequences
|
Annual frequency (1 in x years)
|
Annual probability
|
Treatment, mitigation or avoidance measures
|
Most Likely treatment measure
|
|
Tailings dam failure during a 100-year flood event. Water overtops dam and erodes the structure causing a breach which then discharges tailings
|
Discharge of arsenic-bearing water and sediment into downstream catchments.
|
100
|
0.01
|
Remove sediment from accessible reaches downstream, and redeposit it in the repaired tailings dam. Water and remaining sediment will be flushed over 10-50 years.
|
No action for minor discharge of tailings.
|
|
Large cyanide spill from processing plant during a major rainstorm is washed off site
|
Cyanide kill of most aquatic species several kilometres downstream until dilution reduces the toxicity.
|
50
|
0.02
|
Immediate addition of cyanide-neutralising agent, stored on-site and accessible. Warn nearby population of risks of toxicity.
|
1000 litres of neutralising agent. Regular monitoring of downstream water quality between mine site and the sea until cyanide is no longer detected.
|
|
Waste rock pile collapse during a major rain event
|
Debris enters nearest creek, and sediments mobilised downstream catastrophically.
|
50
|
0.02
|
Remove sediment build-up from accessible reaches onto nearby banks or other disposal site.
|
Excavators and dump trucks to remove sediment
|
|
Earthquake greater than Magnitude 7
|
Instability and possible breaching of tailings dam; landslides from valley walls into tailings dam; failure of waste rock piles; debris enters nearest creek, and sediments mobilised downstream catastrophically.
|
200
|
0.005
|
Local and regional infrastructure damage will preclude immediate action on environmental issues such as arsenic discharge. Human life and access will be top priority.
|
Excavators and dump trucks to remove sediment
|
Table G3. Risk matrix of ‘typical’ extreme events associated with an alluvial gold mine in Miocene quartz gravels in Southland
|
Risk
|
Consequences
|
Annual frequency (1 in x years)
|
Annual probability
|
Treatment, mitigation or avoidance measures
|
Most Likely
|
|
Major rainfall event causes overtopping of tailings dam (silt pond)
|
Discharge of suspended sediment into downstream catchments
|
100
|
0.01
|
Remove deposited sediment from accessible reaches downstream to clear waterway.
|
No action for minor discharge of tailings
|
|
Failure of mine water containment structure during drought
|
Discharge of acid water with elevated Al, Fe, Ni, As
|
50
|
0.02
|
Flush immediate streams with water from a non-acid source.
|
Water tanker loads
|
G.4 References
Black A, Craw D, Youngson JH, Karubaba J 2004. Natural recovery rates of a river system impacted by mine tailing discharge: Shag River, East Otago, New Zealand. Journal of Geochemical Exploration 84: 21–34.
International Council on Metals and the Environment (ICME) 1999.The management of cyanide in gold extraction. http://www.icmm.com/document/124
International Cyanide Management Institute (ICMI) 2002. Cyanide facts: Cyanide sampling and analytical methods for gold mining. http://www.natural-resources.org/minerals/cd/docs/unep/cyanidecode/sampling.pdf
International Cyanide Management Institute (ICMI) 2005. International cyanide management code. http://www.cyanidecode.org/become_implementation.php
Lees P and Thompson D 2003. Emergency management, Opuha Dam collapse, Waitangi Day 1997. In: IPENZ Proceedings of Technical Groups 30/2. Available at: http://www.ipenz.org.nz/nzsold/2003Symposium/LargeDams2003pages84-104.pdf#search=%22opuha%20dam%20collapse%22.
United Nations Environment Programme (UNEP) 2000. Cyanide spill at Baia Mare. Final report. Geneva: ENEP/OCHA. www.mineralresourcesforum.org/incidents/BaiaMare/docs/final_report.pdf
United Nations Environment Programme (UNEP) 2001. Tailings dams: risk of dangerous occurrences – lessons learnt from practical experiences. http://www.mineralresourcesforum.org/docs/pdfs/Bulletin121.PDF
U.S. Environmental Protection Agency (USEPA) 1994.Technical Report: Design and evaluation of tailings dams. http://www.epa.gov/osw/nonhaz/industrial/special/mining/techdocs/tailings.pdf
World Meteorological Organization 1986. Document manual for estimation of probable maximum precipitation. Available for purchase from: http://www.wmo.int/e-catalog/detail_en.php?PUB_ID=377&SORT=N&q=
◄ previous page │ next page►
