Èñòî÷íèê:
http://http://www.image-train.net/products/IT_ASC2_proceedings.pdf
This paper provides a brief overview of the impacts of mining on physical hydrogeology. For a more comprehensive account, the reader is referred to the recent text book of Younger et al. (2002). In the account which follows, the following topics are examined:
Types of Mine
Fundamentally, the mining industry distinguishes between two type of mine: (i)
Deep Mine: Any mine in which the miner and / or his machinery work beneath a cover of soil or rock (irrespective of absolute depth below ground surface) (ii)
Surface Mine: Any mine in which the miner and / or his machinery work in an exca- vated void which is open to the skies
The term "deep mine" is synonymous with “underground mine” or “subterranean mine”. (In British usage, the unqualified use of the word “mine” generally means a deep mine).
Deep mines can be further differentiated into "drift mines" or "shaft mines" depending on the mode of access to the underground workings. Access to deep mines can be either by means of a shaft (i.e. a vertical or sub-vertical tunnel) or an adit (an essentially horizontal or sub-horizontal tunnel from a hillside). In coal-mining districts, the term “drift” is often used synonymously with “adit”, hence the name "drift mine".
Other types of mine access features represent some form of compromise between horizon- tal adits or vertical shafts, and these are generally termed "declines", which refers to in- clined tunnels from the ground surface to the workings. (Synonyms for 'decline' include “in- cline”, “inclined drift”, “slope entry” or (least commonly) “dib”).
Working of deep mines
Working of all deep mines involves the following activities:
Distinctive arrangements of these activities can be recognised by the patterns of voids shown on mine plans (Figure 1). Fig. 1: patterns of underground mine voids associated with (a) various types of "bord-and-pillar" work- ings and (b) typical longwall workings.
Bord-and-pillar workings (Figure 1(a)) result in approximately rectilinear networks of inter- connecting roadways ("bords") separated by un-mined “pillars” of coal (or other mineral) left behind in order to provide roof support. This manner of workings was the main deep mining technique in Europe for many centuries. Because of the relatively low capital costs of bord- and-pillar working compared to longwall, it remains the principal technique in the USA (where most underground coal mining is undertaken by relatively small companies). It is al- so practised in situations where:
Longwall working (Figure 1(b)) involves the removal of all coal (it is rarely practised as such for other minerals) in entire, discrete 'panels', which may be up to 250m wide. The “long wall” is the long, working face of such a panel, along which a drum shearer passes back and forth. As the face is cut away, the shearer and its hydraulic supports advance towards the retreating wall, and the area behind the supports is left to fall (forming “goaf”). The longwall is usually sheared away over a distance of 500m to 1000m perpendicular to the ini- tial position of the working face. Extraction rates for longwall commonly exceed 90%, and it is the main technique in modern European coal mining.
“Shortwall” works on the same principle as longwall, but with narrower panels (as narrow as 30m in extreme cases); this is usually done to minimise the vertical extent of subsidence and associated fracturing above the panel, which may be advisable when mining below wa- ter bodies.
The term "stoping" refers to a range of techniques applied to deposits with significant verti- cal extension (e.g. vertical / sub-vertical veins, as in many European metals mines). Varie- ties of stoping include “stope-and-pillar” (analogous to bord-and-pillar, but in the vertical azimuth), “overhand stoping” (e.g. shrinkage stoping), "block-caving" and “sub-level stop- ing”. Full details of these techniques are beyond the scope of this paper, and the reader should consult mining engineering texts such as that of Hartman & Mutmansky (2002) for further details.
Surface Mines
Surface mines currently account for the bulk of world-wide mineral production (> 80% in the mid-1990s and rising). The term 'surface mine' is approximately synonymous with “open-pit mine”, “quarry”, “opencast mine”, though these terms are also used to signify specific types of surface mines. As in deep mining, there are three Principal activities in surface mining:
Types of surface mine are distinguished on the basis of how these activities are done, as follows:
Fig. 2: Opencast coal mining operations: (left) working bench in front of the highwall of a shallow pit. (right) advancing loose-wall (on left) behind working bench
Mine Wastes
It is estimated that more than 70% of all the material excavated in mining operations world- wide is currently waste. In the EU over the last few years, more than 400 Mt of mine waste has been generated per annum, amounting to nearly 30% of all waste generated in the EU.
This is a vast proportion compared to that part of the EU's economy which is currently ac- counted for by mining.
The reason for the very large volume waste production by mining at present relates to the current preponderance of surface mining. Whereas mature deep mining provides scope to dispose of waste rock within mined voids, and thus produces a tiny proportion (around 1%) of the total mass of mine waste produced annually around the world, surface mines inevita- bly disturb vast quantities of overburden. Whilst open-cast mines largely backfill their voids as they go, open-pit mines often don’t backfill at all and at the end of mining they leave be- hind both an open void plus large adjoining waste rock piles.
Mine waste disposal occurs in two principal manners, depending on the origins and grain- size of the wastes:
Implications for physical hydrogeology How mining alters natural hydrogeology Surface mining
Where surface mines are excavated into aquifer materials, they clearly remove part of the aquifer, which in itself may represent a loss of resource (e.g. increased evaporation from the post-mining pit-lake) or at least an increase in vulnerability for the surrounding aquifer resources (i.e. removal of the barrier to pollutants represented by the unsaturated zone).
Besides these obvious impacts, most other effects of surface mines on natural hydrogeol- ogy are rather subtle. A “halo” of increased permeability (? 100 times greater than back- ground values) can develop around open-pit walls, due to extensional fracturing induced by blasting and the reduction of lateral stresses. Indeed, permeability close to the void may be so high as to favour turbulent flow near the void, resulting in a near-pit water table which is much more steep than would be expected if groundwater flow remained strictly Darcian.
This phenomenon has been extensively analysed by Dudgeon (1985), and is also dis- cussed by Younger et al. (2002).
A further impact of pit lakes left behind after cessation of surface mining is that the water table tends to be steeper on the up-gradient side of the pit, and more gentle on the down- gradient side of the pit, than would be the case under natural conditions (Morgan-Jones et al., 1984). On the other hand, although many pit lakes are in hydraulic continuity with the surrounding ground water, in some cases the blinding of the pit floor with fine-grained sedi- ments can effectively “perch” water in the open-pit, with little or no interaction with the sur- rounding ground water system.
Pit lakes are complex environments from a limnological and geochemical perspective. Limnologically, the key difference between pit lakes and natural lakes can be quantified by a parameter known as the relative depth (D R ) (Castro & Moore, 2000) CP-035 (2004) D R = 100 • (z m / d) where z m is the maximum depth of the lake and d is some standardised diameter (e.g. for circular lakes, d = 2 •v(A / p), where A is the surface area). Typical D R values for natural lakes are usually < 2%, and few have values of more than 5%. By contrast mine pit lake typically have D R values in the range 10 to 40%. Such high values of D R have important hydrological consequences. Most notably a high D R means that evaporative losses as % of stored water (and therefore evaporative concentration) from the pit lake will be limited compared to a natural lake with a similar surface area. High D R also has profound effects on density stratification: with high D R values, pit lakes promote the de- velopment of three-layer systems in which the third, deepest layer is never involved in sea- sonal overturn (“meromictic” conditions). For further discussion of the importance of mixing dynamics and coupled geochemical processes for the evolution of pit-lake water quality, the interested reader is referred to the recent, excellent review by Bowell (2002).
Deep mining
As a rough 'rule-of-thumb', the caving-in of deep mines can be expected to cause fracturing and subsidence of overlying strata over a vertical distance typically half as high as the void is wide (e.g. a 200m-wide longwall panel can be expected to affect around 100m of overly- ing strata). The volume of rock affected by fracturing and subsidence above longwall panels can be resolved into 3 zones:
For further discussion of these phenomena and their practical implications, the reader should consult the recent comprehensive review of Booth (2002).
The hydrogeology of abandoned deep mines is essentially "non-standard" when compared with natural aquifers. The flow in flooded deep mines is usually highly channelised (i.e. most flow is associated with old mine voids) and very often turbulent (hence Darcy’s Law is inap- plicable). Collapse of voids and erosion by turbulent flow can lead to permeability and stora- tivity changes over a scale of days or even hours.
Records of water level rises in shafts accessing abandoned, flooding underground workings are known as "rebound curves". These typically show staged variations in rebound gradient which are particularly clear on semi-log (“Jacob”) plots. The variations in gradient record ver- tical variations in storage properties as the water level successively passes through worked and unworked zones (corresponding hydraulically to alternating unconfined and confined storativity conditions).
Modelling of rebound processes has recently been approached using a number of alterna- tive formulations (Adams & Younger, 2001). Essentially, the choice of the most appropriate model is a question of scale, with discrete pipe-network models being most appropriate for detailed analysis of small sub-areas of mine systems, and more coarse semi-distributed models (such as the GRAM model of Sherwood & Younger, 1997) being most appropriate for medium-to-large scale systems of variously inter-connected "ponds" of flooded workings.
The use of standard porous medium models may be appropriate at the very largest scales of analysis, especially for rebounded systems where hydraulic gradients are shallow once more. Once an appropriate simulation code has been chosen, the predictive modelling of rebound processes is likely still to be hindered by a number of factors, including:
Despite these setbacks, successful modelling of rebound processes has been undertaken for a number of major mined systems (e.g. Adams & Younger, 2001), and there is no doubt that further mine closures throughout Europe will necessitate more studies of this type in fu- ture (e.g. Younger, 2002).
Hydrogeology of bodies of mine waste
Spoil heap creation and hydrogeological properties
The way in which spoil heaps are formed fundamentally influences their internal hydro- geological character. Loose-tipping of poorly-sorted waste rock fragments tends to sort the sediments out into 'cobbly zones' (pebbles and cobbles) and fine-grained zones. This im- parts profound heterogeneity to the spoil, giving rise to strongly preferential flow mecha- nisms. Modes of preferential flow differ between the unsaturated and saturated zones in mine spoil heaps:
Unsaturated zone: the pore system of the fine-grained fraction of the spoil will always be at or near saturation, whereas pores in cobbly zones will usually be fully drained; hence in the unsaturated zone water moves preferentially through the fine-grained fraction of the sedi- ment.
Saturated zone: the cobbly zones are FAR more permeable than the fine-grained zones, so that in the saturated zone, water moves preferentially in cobbly zones.
Preferential flow can in some cases provoke preferential erosion of spoil heaps, leading to development of sinkholes and other features which give rise to surface subsidence (Figure 3a). Surface runoff can be channelled into such features (Figure 3b), further exacerbating erosion and also denying surface environments some much-needed water.
Geotechnical stability problems associated with the hydrogeological behaviour of waste rock depositories. (a) Subsidence 'crown hole' developed above a preferential subsurface storm- flow pathway in open-cast backfill. (b) The entire flow of a stream disappearing into the same crown hole during a storm. (c) The failure of the Aznalcollar Tailings Dam in SW Spain in April 1998 - an emblematic event in the recent history of mining environmental manage- ment in the EU.
Depending on the lithology of the waste rock, spoil heaps may be dominated by surface runoff or subsurface flow. For instance, the more shale in the waste rock, the more likely is a spoil heap to give rise predominantly to surface runoff. In areas where much of the waste rock is hard crystalline rock, infiltration and perched groundwater systems are the norm. The relative dominance of surface / subsurface flow has important implications for pollutant- release dynamics:
Surface runoff-dominated spoil heaps release their pollutants when heavy rains follow dry spells: peak flows and peak contaminant loads tend to coincide.
Subsurface-dominated spoil heaps often have dilution in wet periods, with peak contami- nant concentrations during baseflow periods. This distinction has important implications for remedial design options (Younger et al., 2002).
Tailings Dams
Tailings are usually deposited from suspension in water. Besides easing physical transport of the tailings, sub-aqueous emplacement also helps prevent pyrite oxidation. However, the mixture of fine-grained sediments and water can give rise to problems with the geotechnical behaviour of tailings dams. In particular, inadequate control of water in tailings and their im- pounding dams can contribute to rotational / sliding failures which then allow vast volumes of tailings to escape from the tailings dam, as happened at Aznalcollar, Spain, in April 1998 (Figure 3(b)) and at Baia Mare, Romania, in 1999. Such problems are avoidable, and al- though the relevant precautions are generally built-in to tailings dam designs, problems tend to arise where later raising of the dam (which might be decades after the initial design, de- pending on the fortunes of the related mining operation) is implemented in a manner which departs from the original design specifications. Changes in EU laws to prevent this sort of calamity occurring in future are currently very much on the agenda (e.g. European Commis- sion, 2003).
Conclusions
Mining affects natural hydrogeology quite profoundly, in ways which are often not readily amenable to analysis using conventional methods of groundwater hydrology. Similarly, bod- ies of mine waste have distinctive hydrological characteristics which complicate their analy- sis and their geotechnical behaviour. Nevertheless, a conceptual framework now exists by means of which rational hydrogeological analyses in mined areas can be reliably under- taken.
References
Contact
Paul L. Younger
School of Civil Engineering and Geosciences
University of Newcastle
Drummond Building
UK - NE1 7RU Newcastle upon Tyne
Tel.: + 44 191 222 7942, Fax.: + 44 191 222 6669
Email: p.l.younger@ncl.ac.uk