Assessing Mining-Related Water Impacts: A Case Study-Based Systematic Review Supporting a More Comprehensive Approach

Abstract

Mining-related impacts on water are often more persistent, costly, and extensive than suggested by standard reporting practice which frequently omits the indirect impacts on third parties associated with mining-induced alterations to water flow and quality and rarely covers the associated economic knock-on effects (costs). Based on a literature review, this paper aims to identify and categorise such underreported impacts using 15 international case studies from nine countries covering surface and underground operations and a wide range of commodities. For each of the seven proposed water impact categories such as discharge of mine effluents, mine waste-related pollution, de- and rewatering, hydrological alterations, etc., corresponding case studies and associated key impact parameters are compiled. Linking mining-induced pollution to water quantity and costs is proposed as a potential means to curb the externalisation of costs. In addition to the literature data, original calculations are provided on how the creation of vast “lakescapes” in former lignite districts exacerbates water stress in two of Germany’s driest regions. The main objective is to strengthen the future licencing decisions of authorities and better anticipate mining-induced water impacts at a time when rising demand for energy transition metals intensifies the competition for limited water resources in many already water-stressed mining regions.

1. Introduction

Mining is as essential for any modern society as agriculture is for feeding humankind. Mineral extraction has long been a key driver of economic and social development by providing the raw materials needed for infrastructure, industry, and everyday resources that sustain and improve human life [1,2]. Mining activities also generate incomes and foster business opportunities across primary, secondary, and tertiary sectors, supporting prosperity at local, regional, and national levels. Mining is expected to grow with the global energy transition, as demand rises for the critical minerals used in lithium batteries, electric vehicles, solar panels, and wind and hydropower technologies [1,2]. Between 1970 and 2022, raw metal extraction nearly quadrupled, increasing from 2.7 to 9.4 gigatons. However, the mining industry is also recognised as a major water user that impacts water resources directly (e.g., via discharging polluted effluents) as well as indirectly (e.g., via generating large volumes of contaminated mine waste, whose seepage can in turn pollute surface and groundwater) [1,3,4,5].

In this context, the often-invisible impacts of mining on water resources frequently remain underestimated, poorly characterised, and underreported in routine environmental documentation and corporate sustainability reporting, despite ongoing efforts to improve water reporting in the industry [6,7,8,9,10,11,12,13]. This is partly because current corporate and regulatory water reporting practices in the mining sector tend to focus on operational indicators, such as water withdrawals, recirculation, and discharges, without adequately accounting for the cumulative, indirect, or long-term effects [9,14]. In many cases, diffuse contamination, aquifer degradation, or the persistence of liabilities after mine closure are not included in environmental reports or corporate accounting [5,10,13], and are not adequately considered, if at all, when granting mining licences.

Furthermore, water-related effects are often dissociated from their social, ecological, and economic costs, making it difficult to understand the full scope of the problem. Health risks from chronic metal exposure and the loss of ecosystem services are rarely addressed in routine water reports or in dedicated sustainability disclosures [13,15]. Persistent impacts on groundwater, including reduced availability and prolonged metal contamination in Mexico, have been documented even years after mining operations have ceased [16]. Independent assessments have also identified critical gaps in current reporting systems, particularly concerning impacts beyond mine boundaries, post-closure effects, systemic risks, and unaccounted for externalities that affect communities and governments [9,10,11]. Likewise, agricultural losses due to the use of mining-polluted water, which may also cause costs to industrial users, lack of access to water for domestic or recreational uses, and the degradation of aquatic habitats are rarely considered part of the mining-related water impact within prevailing reporting practices (e.g., environmental compliance documentation and corporate sustainability reporting) [13,17]. Despite improvement efforts by the industry, this disconnection has led, and continues to lead, to planning decisions that underestimate the real extent of water use, shifting associated costs onto governments, local communities, and future generations [6,9,10,11,12,18,19].

In addition, water impacts through mining do not occur under static hydroclimatic conditions. Predicted changes in climate extremes and the hydrological variability associated with global warming may intensify some of the impacts discussed by reducing effluent dilution during prolonged droughts, or increasing contaminant dispersion during extreme rain events [20,21].

Having reviewed diverse mine types and regions across the globe, this paper identifies and categorises mining-related water impacts using a primarily qualitative approach. Analysed impacts include water pollution by acid mine drainage, mine waste and contaminated effluent discharges, dewatering-induced land degradation, hydrological alterations of surface and groundwater systems, changes to regional water balances, and long-term impacts associated with mine flooding (rewatering). While the synthesis is predominantly descriptive, quantitative data reported in the literature are extracted where available for selected key parameters, such as the river length affected by mines, pH ranges, contaminant concentrations, discharged effluent volumes, or area sizes impacted by groundwater drawdown. These parameters allow for limited comparison across case studies but do not constitute a consistent quantitative framework applicable across different impact categories.

The present review focuses on identifying and structuring water-related impacts that are not sufficiently captured by current reporting. A comprehensive aggregation of direct and indirect impacts into a unified quantitative mine water footprint is not attempted here and would require additional data beyond the scope of this study. Nevertheless, by compiling and categorising underreported or even ignored impact pathways, this work aims to raise awareness among regulators and decision-makers of potentially unforeseen water implications when granting mining licences, and to provide a basis for the future development of fully quantitative assessments.

2. Methodology

2.1. Literature Review and Definitions of Terms

A structured literature review was conducted and reported in accordance with the PRISMA 2020 guidelines [22] to identify documented mining-related impacts on surface water and groundwater resources. The completed PRISMA checklist is provided as Supplementary Table S1. Given the qualitative and case study-based nature of this work, the review was designed as a scoping review and was not prospectively registered. Searches were performed using Scopus, Web of Science, and Google Scholar, covering peer-reviewed journal articles, conference proceedings, book chapters, and selected grey literature, including governmental and company reports as well as post-disaster assessments.

The search strategy combined mining-related terms with water impact-related terms using Boolean operators. Core search strings included combinations such as “mining” AND “water impacts”, “acid mine drainage”, “tailings”, “dewatering”, “mine flooding”, “effluent discharge”, and “hydrological alteration”. Where database functionality allowed, searches were applied to titles, abstracts, and keywords. A comprehensive list of keywords, grouped by thematic category, is provided in Supplementary Table S2. The literature search covered publications from the early 20th century onwards and was last updated in November 2025. Publications in English were prioritised, while relevant documents in German and Spanish were also included due to the historical importance of covered mining activities.

After removal of duplicate records across databases, all remaining records were screened at title and abstract level. Records were excluded if they did not address mining-related impacts on water quantity or quality, focused exclusively on operational water efficiency without environmental relevance, or lacked a clear link between mining activities and hydrological or water quality effects. Full-text articles were subsequently assessed for eligibility, as summarised in the PRISMA 2020 flow diagram (Figure 1). While the key references indicated in Figure 1 can be regarded as characteristic of each selected case study, the selection process and subsequent characterisation of the 15 mining examples was naturally based on a much larger number of relevant references.

Identified case studies were included in the qualitative synthesis if they met the following criteria: (I) documented impacts on surface and/or groundwater quantity or quality; (II) impacts directly attributable to mining activities, including active, abandoned, or legacy sites; (III) availability of descriptive information on impact mechanisms and direct or indirect effects; and (IV) representation of diverse impact pathways across different geographical, climatic, and geological settings.

Several case studies draw on multiple publications describing the same mining system or hydrological setting. For the purpose of the qualitative synthesis and the PRISMA 2020 flow diagram, such systems were treated as single analytical case studies. Following this process, 15 case studies from 9 countries were retained for qualitative synthesis, as illustrated in Figure 1. A list of the selected case studies and their key reference is provided in Supplementary Table S3.

In the context of this paper, “direct impacts” refer to all consequences of all mining-related activities that alter the quality and/or flow of receiving water bodies on surface and underground without impacting on other environmental compartments first (i.e., the discharge of polluted mine effluents via pipes or canals into a stream). Direct impacts are confined to the affected water body.

In contrast, “indirect impacts” generally involve other environmental components/entities that are affected by the consequences of direct water impacts, i.e., the formation of sinkholes in karst areas after dewatering is an indirect impact of drawing down the groundwater table (direct impact). Similarly, the contamination of fluvial sediments is an indirect impact (consequence) following from stream pollution through the discharge of polluted mine effluents affected through dissolved contaminants becoming immobilised as solid phase to precipitation or sorption and accumulate in the sediment. Indirect impacts may also include more than one additionally impacted entity. For example, if a mine pollutes a stream (direct impact) whose water is then used for irrigation it results in (a) contamination of the soil in which the plants grow and (b) contaminating crops grown in the soil. Adverse health effects of people or animals consuming the contaminated crops constitute a third tier of indirect water impacts of mines.

Where indirect impacts result in quantifiable economic losses, the term “knock-on effect” is used to denote that these impacts may cause costs for third parties. Since some of these costs have been compensated by mines (e.g., gold mines in the Far West Rand buying up sinkhole-degraded agricultural land at pre-impact value) the term “externalised cost” could not be generally used.

2.2. Calculating Climatic Water Balances (CWBs) for Pit Lakes

In addition to reviewing secondary data, primary data were generated relating to quantifying changes in natural climatic water balances of lignite mining regions in Central and Eastern Germany. Based on reported data for the last climatic normal period (or longer), we calculated evaporation rates and annual climatic water balances using data for 15 m-deep reference lakes. As the deepest reference lake category for which long-term statistics are available, it also covers lakes deeper than 15 m (like most pit lakes) whose thermal properties are deemed to differ only marginally, if at all. For interim periods where lakes are not yet fully filled-up, data for 6 m-deep reference lakes were also used, displaying a slightly larger magnitude of seasonal temperature fluctuations (and thus evaporation rates) than deeper lakes with their greater thermal inertia.

For the Central German lignite region, the average annual precipitation figures (1971–2000) of stations falling into the mining area ranged from 441 mm/a (Salzmünde) to 606 mm/a (Leipzig-Holzhausen) mm/s (estimated regional average 500 mm/a) and for the Lusatian Region it ranged from 560 mm/a (Cottbus) to 640 mm/a (Hoyerswerda) with an estimated regional average of 600 mm/a [23,24,25,26].

Since rainfall over lakes is generally less than over rougher land surface, the rainfall figures for lakes in the two mining regions were adjusted downward by 18% based on the long-term difference between precipitation over German reference lakes (644 mm/a) and the average across the whole of Germany (789 mm/a).

For calculating the climatic water balance (CWB) of lakes, the mean annual evaporation rate (MAE) from lakes is subtracted from the mean annual precipitation rate (MAP) corrected for lakes (CWB [mm/a] = MAP [mm/a] − MAE [mm/a]). The corresponding volumes for CWB, MAP, and MAE are calculated by multiplying the respective rates with the size of the relevant area (CWB [l/a] = area size [m²] × CWB [l/m² × a]). Results are subsequently converted into million cubic metres per annum [Mm³/a] (Table 1).

Table 1. Long-term average evaporation rates over different surfaces in Germany (1893–2014; data: [23,24,25,26]).

Surface Type/ Land Use	Actual Evaporation [mm/a]	Precipitation [mm/a]	Climatic Water Balance [mm/a]
Reference lakes in Germany (2/6/15 m av. depth) [27] (a)	749/740/730	644 (c)	−105/−96/−86 *
Pit lakes (av. depth 15 m) (Central Germany/Lusatia)	>730 (b) (718) (a) [28]	410/492 (c)	<−320/<−238 *
Unvegetated farmland (av. Germany/CGCF/LCF)	360	789/500/600	+429/+140/+240 **
Grassland (over sand) (av. Germany/CGCF/LCF)	435	789/500/600	+354/+65/+165 **
Sealed surfaces (towns, etc.) (av. Germany/CGCF/LCF)	86	789/500/600	+703/+414/+514 **

^(a) Water loss is highest over shallow lakes (2 m depth) as they warm up quicker during summer than deeper ones increasing evaporation rates. Deeper lakes with larger volumes of water to be heated generally display a greater thermal inertia with lower water surface temperatures and thus lesser evaporation. The effect is limited to a maximum depth of 16 m below which no further reduction in evaporation is observed. With average pit depths of 20–140 m, the evaporation values for 16 m-deep lakes are used for the pit lakes; ^(b) Evaporation rates from lakes in Germany generally increase from north (Schleswig-Holstein) to south reaching their maxima in southern Brandenburg and Saxony [23,24,25,26]. In the absence of site-specific data, the greater-than symbol (>) is used to denote this fact; ^(c) local precipitation corrected for lower values over open water (−18%); * = values with a minus denote a negative CWB; ** = values with a plus a positive one.

Table 1. Long-term average evaporation rates over different surfaces in Germany (1893–2014; data: [23,24,25,26]).

Surface Type/ Land Use	Actual Evaporation [mm/a]	Precipitation [mm/a]	Climatic Water Balance [mm/a]
Reference lakes in Germany (2/6/15 m av. depth) [27] (a)	749/740/730	644 (c)	−105/−96/−86 *
Pit lakes (av. depth 15 m) (Central Germany/Lusatia)	>730 (b) (718) (a) [28]	410/492 (c)	<−320/<−238 *
Unvegetated farmland (av. Germany/CGCF/LCF)	360	789/500/600	+429/+140/+240 **
Grassland (over sand) (av. Germany/CGCF/LCF)	435	789/500/600	+354/+65/+165 **
Sealed surfaces (towns, etc.) (av. Germany/CGCF/LCF)	86	789/500/600	+703/+414/+514 **

^(a) Water loss is highest over shallow lakes (2 m depth) as they warm up quicker during summer than deeper ones increasing evaporation rates. Deeper lakes with larger volumes of water to be heated generally display a greater thermal inertia with lower water surface temperatures and thus lesser evaporation. The effect is limited to a maximum depth of 16 m below which no further reduction in evaporation is observed. With average pit depths of 20–140 m, the evaporation values for 16 m-deep lakes are used for the pit lakes; ^(b) Evaporation rates from lakes in Germany generally increase from north (Schleswig-Holstein) to south reaching their maxima in southern Brandenburg and Saxony [23,24,25,26]. In the absence of site-specific data, the greater-than symbol (>) is used to denote this fact; ^(c) local precipitation corrected for lower values over open water (−18%); * = values with a minus denote a negative CWB; ** = values with a plus a positive one.

2.3. Defining Categories of Mining-Related Water Impacts

In this paper, seven main impact types are proposed, based on underlying mechanisms and observable effects: (I) acid mine drainage (AMD) cross-cutting through several other categories; (II) mine waste-related impacts (e.g., seepage, dam failure, slurry discharge); (III) dewatering-induced effects (e.g., aquifer drawdown, land subsidence); (IV) post-closure rewatering processes (e.g., mine flooding and subsequent pump-and-treat of emanating mine water); (V) point-source effluent discharges (e.g., contamination from process water and water pumped from underground); and (VI) alterations of the landscape hydrology (e.g., stream diversions, building of dams, cross catchment water transfer, changes in runoff coefficients, etc.). Each case was assigned to one or more of these categories according to the primary documented impact. The seventh category relating to consumptive water use is not explored in great detail, as this dimension is commonly well covered by standard reporting systems. The emphasis is therefore placed on commonly underreported categories occurring off-site and after mine closure often including indirect and cumulative impacts across the entire mine life cycle.

Where multiple impact pathways were reported for a given case study, a primary category was assigned based on the dominant hydrological mechanism described in the literature, while co-occurring impacts were also documented as qualitative description. The selected case studies are not intended to be exhaustive, but exemplary. Many sites exhibit overlapping and interrelated impacts, reflecting the interconnected nature of mining impacts on water resources. Special emphasis was placed on showcasing representative cases highlighting the chronic, cumulative, and often irreversible character of these impacts as well as potential knock-on effects and externalised costs for third parties, including the natural environment.

3. Mining Impacts on Water: Selected Case Studies

To enable comparison across different climatic, geological, and hydrological contexts, the wide range of mining impacts on water were organised into distinct categories. Although each site presents local particularities, the cases analysed show that certain types of impacts recur systematically across diverse mining contexts, regardless of the mineral extracted, the extraction technology used, or the phase of the mine life cycle.

For example, in the Far West Rand goldfield (South Africa), extensive groundwater dewatering was treated as the primary impact mechanism to highlight this impact type. Co-occurring impacts, including sinkhole formation, contamination of karst aquifers by tailings seepage, and long-term post-closure water quality degradation, were documented as secondary but interconnected effects within the qualitative case description.

To provide an overview of these recurring processes, Figure 2 presents a schematic classification of mining-related water impacts across the mine life cycle. The diagram highlights both phase-specific mechanisms (e.g., dewatering during operations or rewatering after closure) and cross-cutting processes such as AMD, which connect multiple categories and timeframes. This framework serves as a conceptual map for the following subsections, where each impact category is examined in detail and illustrated with representative case studies.

Each subsection of this part presents one impact category, which is described in general terms and illustrated with one or more representative case studies. The categories are not mutually exclusive but display a degree of overlap (e.g., AMD which relates to pollution by mine waste as well as flooded mines). For each case, information is provided on the geographic context, mine type, affected water bodies, key contaminants, duration of impact, distinction between direct and indirect effects, and associated socio-economic effects when available.

3.1. Acid Mine Drainage

Acid mine drainage (AMD) is one of the most common and persistent forms of water pollution caused by mining. It occurs when sulphide-bearing materials, especially iron sulphides, are exposed to oxygen and water, triggering a series of oxidation reactions that acidify water leading to low pH values, high electric conductivity (EC) and elevated concentrations of toxic trace metals, metalloids, natural radionuclides, and sulphate (SO₄²⁻). Although AMD can occur naturally, mining activities greatly accelerate its rate of formation by exposing large volumes of rocks to the atmosphere and greatly expanding the reactive surface through crushing, grinding, and milling of the sulphide-containing ore [29]. Rates of AMD formation depend on local geology, the grade and type of sulphide minerals, and the availability of oxygen and water. Biological factors, such as the presence of microbes, can also accelerate the process [30].

AMD can severely degrade surface and groundwater quality, affect ecosystems, and pose health risks to humans and animals. The global variability in AMD generation makes it difficult to predict, assess, and remediate, contributing to an estimated worldwide liability of nearly USD 100 billion [31].

3.1.1. Centuries-Old Polymetallic Mines at Cerro Rico (Bolivia)

Cerro Rico, located in the highlands of southern Bolivia, lies at the headwaters of the Pilcomayo River, a transboundary watercourse shared by Bolivia, Paraguay, and Argentina. Mining and ore processing have historically been intensive in the region, particularly for silver, tin, lead, and zinc [32]. The site features an extensive network of underground workings and numerous active, informal, and abandoned mine entries [32,33]. Centuries of underground mining, lack of wastewater treatment, and ongoing local operations have led to continuous discharges of AMD, heavily contaminating surface waters and sediments throughout the Pilcomayo River system. These impacts extend well beyond the local area, affecting the entire basin [33,34].

The resulting increase in toxic metal levels affected streams are shown in Table 2 and Supplementary Tables S4–S6 [35,36,37,38,39,40,41,42,43]. The use of mining-contaminated water for irrigation has led to heavy metal accumulation in agricultural soils and crops exposing local populations through drinking polluted water and consuming contaminated food [34,44] (Table 2). Livestock and wildlife are also affected through bioaccumulation, impacting food security, local economies, and natural ecosystems, respectively [34]. While contaminant levels in streams generally decrease with distance from the mine, fluvial sediments may accumulate toxic metals well away from the actual source of pollution, creating long-term secondary sources of contamination [45].

In addition to the long-term water contamination, the physical stability of the Cerro Rico Mountain itself has become critically compromised after nearly five centuries of intensive underground extraction. Recent assessments report increasing cave-ins and widespread structural weakening of the upper slopes, driven by renewed informal mining and the removal of disseminated ore bodies. These collapses directly endanger local Quechua communities living and working on the mountain and further complicate environmental management efforts [46].

Table 2. Overview of AMD-related mining impacts on water resources. Abbreviations: ~: approximately; a: years; BCE: before common era.

Studied Mine(s)	Period	Affected Water Bodies and Extent	Environmental and Ecological Impacts	Socio-Economic Consequences	References
Polymetallic mines at Cerro Rico (Bolivia)	1545–present (~480 a)	Pilcomayo River (main course + tributaries); ~200 km surface, ~560 km sediments	Acidification; metal contamination (As, Pb, Zn, Cd); biodiversity loss; metal bioaccumulation in crops and livestock	Reduced agricultural productivity; livestock and fishery losses; health risks (neurotoxicity, renal and cardiovascular effects)	[32,33,34,44,45,47,48]
Au and U mines in the Witwatersrand (South Africa)	1886–present (~139 a)	Klip, Rietspruit, and Wonderfonteinspruit rivers; wetlands and karst aquifers (>100 km)	Surface and groundwater contamination; U and metal accumulation; loss of aquatic species	Costly water pre-treatment; reduced land usability; long-term liability and exposure risks	[49,50,51,52]
Polymetallic mines, Rio Tinto (Spain)	Since ca. 3000 BCE; industrial mining since 1873 (~152 a)	Río Tinto and Odiel rivers → Huelva estuary (~125 km total)	Persistent acidification; transfer of metals to soils and sediments; loss of fish and aquatic macroinvertebrates	Restricted agriculture and grazing; high remediation and treatment costs; enduring environmental liability	[53,54,55,56,57,58]

Note: “→” indicates that both rivers flows into the Huelva estuary.

Table 2. Overview of AMD-related mining impacts on water resources. Abbreviations: ~: approximately; a: years; BCE: before common era.

Studied Mine(s)	Period	Affected Water Bodies and Extent	Environmental and Ecological Impacts	Socio-Economic Consequences	References
Polymetallic mines at Cerro Rico (Bolivia)	1545–present (~480 a)	Pilcomayo River (main course + tributaries); ~200 km surface, ~560 km sediments	Acidification; metal contamination (As, Pb, Zn, Cd); biodiversity loss; metal bioaccumulation in crops and livestock	Reduced agricultural productivity; livestock and fishery losses; health risks (neurotoxicity, renal and cardiovascular effects)	[32,33,34,44,45,47,48]
Au and U mines in the Witwatersrand (South Africa)	1886–present (~139 a)	Klip, Rietspruit, and Wonderfonteinspruit rivers; wetlands and karst aquifers (>100 km)	Surface and groundwater contamination; U and metal accumulation; loss of aquatic species	Costly water pre-treatment; reduced land usability; long-term liability and exposure risks	[49,50,51,52]
Polymetallic mines, Rio Tinto (Spain)	Since ca. 3000 BCE; industrial mining since 1873 (~152 a)	Río Tinto and Odiel rivers → Huelva estuary (~125 km total)	Persistent acidification; transfer of metals to soils and sediments; loss of fish and aquatic macroinvertebrates	Restricted agriculture and grazing; high remediation and treatment costs; enduring environmental liability	[53,54,55,56,57,58]

Note: “→” indicates that both rivers flows into the Huelva estuary.

3.1.2. Gold and Uranium Mines in the Witwatersrand Basin (South Africa)

The Witwatersrand Basin, South Africa, is one of the world’s most important gold mining regions, with first mining activities dating back to 1886 [50]. In well over a century, industrial-scale deep-level mining created extensive underground networks of tunnels and voluminous tailings deposits on the surface now located within the densely populated towns and cities that grew around the mines [50,59,60,61].

After the mines closed and the associated pumping of ingress from underground mine workings ceased, the interconnected mine voids known as the Western, Central, and Eastern Basins gradually filled up with acidic water, eventually decanting via old shafts, boreholes, or fractures and polluting receiving streams, wetlands, and lakes with iron, zinc, copper, nickel, uranium, and other compounds [39,40,41,42,43] (Table 2 and Supplementary Tables S4–S6).

While the flooding of mine voids and the subsequent decant are important contributors, seepage from tailings deposits covering large proportions of small headwater catchment areas is another significant AMD source highlighting that surface legacies, not only underground flooding, are central to persistent post-closure contamination [49,50,52]. In addition, metals such as uranium, nickel, and zinc accumulate in river sediments that turn into potential secondary sources of pollution persisting for decades [49,62,63,64]. Bioaccumulation in aquatic species and human populations has been demonstrated; U concentrations in the hair of residents, especially children, are elevated near tailings sites, suggesting metabolic uptake of the radiotoxic heavy metal [64,65].

Following recommendations of a Governmental Task Team in 2011, pump-and-treat systems have been constructed in all three flooded basins at an “immediate short-term” costs of ZAR 10 billion (approximately EUR 500 million) to keep the mine water below what is termed environmental critical levels (ECLs) through ongoing pumping water out from the underlying mine voids, neutralising it through liming, and eventually discharging the neutralised mine water into nearby streams still containing the original salt load and much of the metal contents [66]. Owing to the exhausted storage capacity of adjacent tailings dams, the generated residual sludge is reinjected into the underground mine voids from which the water is pumped in the first place, increasing both the volumes and contaminant load of the pumped mine water. With no end in sight and running costs of several hundreds of millions of Rand per year, this places a significant burden on the South African taxpayer.

3.1.3. Millennia-Old Polymetallic Mines in the Iberian Pyrite Belt: Rio Tinto (Spain)

The Tinto and Odiel rivers, situated in southwestern Spain within the Iberian Pyrite Belt, belong to one of the world’s oldest mining regions, where mineral extraction has continued for over 5000 years, including Romans extensively mining silver to fund their imperial expansion [55]. Industrial-scale mining began in the 19th century, mainly through open-pit operations such as Corta Atalaya, although earlier phases relied on underground workings [53,54]. These operations targeted massive sulphide deposits rich in copper, zinc, lead, and iron [53,54].

Large volumes of waste rock and tailings were left exposed to the elements, and their prolonged oxidation produced extreme levels of AMD. In the 2017/18 hydrological year, the receiving river Río Tinto (“Red River”—named after the characteristically reddish colour from dissolved iron) carried 5000 t of iron, 2600 t of aluminium, and hundreds of tonnes of zinc, copper, and manganese displaying typical low pH values over its entire approximately 100 km-long course [54,56]. The 150 km-long Río Odiel carried approximately 2800 t of iron and 4600 t of aluminium per year injected by a multitude of mines, together with substantial loads of manganese and nickel [67] (Table 2 and Supplementary Tables S4–S6).

Both rivers convey this acidic, metal-rich water well beyond the mining districts into the Huelva estuary, where deposition and bioaccumulation contaminate sediments, soils, and aquatic biota. Metal(loid) levels in floodplain soils and forage frequently exceed agronomic and livestock safety thresholds [53,54,67,68]. The recent promulgation of the Critical Raw Materials Act (CRMA) by the European Union has revived mining interest in reactivating some of the above mentioned mines in the Iberian Pyrite Belt [69].

3.2. Mine Waste-Related Water Impacts

Mine residues, particularly waste rock dumps and tailings dams, are among the most persistent and widespread sources of mining-related water contamination. Global tailings production currently reaches 5–7 billion tonnes annually [70]. Tailings consist of finely ground ore from which target commodities (e.g., gold, copper) are extracted by physico-chemical processes. Unlike coarser wastes such as rock or slags, they are composed of sand-, silt-, and clay-sized particles and often retain residual reagents, unrecovered metals, and sulphides [71,72]. Tailings are typically transported and deposited as a water–tailings mix (slurry) to impoundments or tailings dams for long-term disposal [70,72]. Because of their fine-grained and waterborne nature, tailings (both subaquatic and dry deposits) retain significant volumes of porewater that are difficult to remove. In dry deposits such as the “slimes dams” of South African gold mines, this porewater sustains an elevated piezometric surface driving continuous seepage into subjacent aquifers and nearby streams [73]. The commonly used term “tailings storage facility” (TSF) is misleading, as deposition is generally final rather than temporary as the term “storage” suggests.

Due to their composition and reactivity, particularly when sulphides are present, tailings have a high potential to generate AMD [72]. Poor management of tailings dams can cause seepage, erosion, and the associated contamination of surface and groundwater as well as catastrophic collapse, with devastating, long-lasting ecological and human impacts [70,72,74].

Beyond seepage, tailings also affect surface waters through erosion, wind-blown dust, and polluted surface runoff during intense rainfall. Readily dissolvable efflorescent salt crusts forming on tailings during dry-weather periods add to contamination potential. In extreme cases, catastrophic dam failures can bury entire valleys and settlements under toxic sludge, while pipeline ruptures or deliberate discharges into rivers, caves, and lakes (like practised in the past in Canada, Papua New Guinea, and South Africa) have caused severe water, sediment, and land degradation.

3.2.1. Tailings Dam Failure at Los Frailes Sulphide Mine (Spain)

Los Frailes mine, located at Aznalcóllar in the Iberian Pyrite Belt, one of Europe’s main metallogenic provinces, has been exploited since Roman times, but modern operations intensified in 1987 with surface extraction [75,76]. The ore body consists of complex polymetallic sulphides dominated by pyrite, with significant amounts of sphalerite, galena, and chalcopyrite, as well as minor arsenopyrite [75,76,77]. Historically, these deposits were already exploited in Roman times mainly for copper and silver. In modern times, the focus was on the extraction of zinc, lead, and copper, whose concentrates were exported to smelters in Spain and abroad. Silver remained an important by-product.

In 1998, Los Frailes tailings dam failed, releasing some 4 million m³ of acidic water and ca. 2 million m³ of toxic sludge into the Agrio River. The plume rapidly travelled down to the Guadiamar River and partly impacted areas of Doñana’s National Parks, killing all the fish and shellfish in the affected watercourses [75]. Subsequently, dozens of tonnes of fish were removed. Analyses wells in the affected areas revealed trace metal concentrations above drinking water limits [75,78] (Table S7).

The subsequent large-scale remediation included sediment removal and a remediation project, which improved water and soil quality (Table S7). Currently, efforts are underway to reopen the site for underground mining and the old pit to be used for tailings disposal [76,77,79].

3.2.2. Tailings Dam Failures at Fundão and Brumadinho Iron Mines (Brazil)

The Fundão and Brumadinho tailings dams, both located in the state of Minas Gerais (Brazil), were associated with large-scale mining operations dedicated to iron extraction [80,81]. Tailings at both sites were stored in upstream-constructed dams built with iron ore waste and slimes, deposited hydraulically and compacted over time. These structures relied on the formation of stable beaches to prevent water from contacting the dam walls [82,83].

In 2015, Brazil experienced the largest environmental disaster in its history when the Fundão dam near Mariana collapsed, releasing millions of cubic metres of mining waste into the Doce River basin and contaminating approximately a large extension of waterways and confirmed fatalities [74,80,84] (Table S7). Just three years later, in 2019, another catastrophic failure occurred at Vale’s dam in Brumadinho, located at the Córrego do Feijão mine. This event is considered one of the worst tailings dam disasters in global mining history. Millions of cubic metres of tailings were released, burying mining infrastructure and surrounding areas, and rapidly contaminating the Paraopeba River. The disaster led to the deaths of numerous people [84] (Table S7). The contamination severely affected aquatic and riparian ecosystems and disrupted the lives of populations relying on these rivers for drinking water, fishing, and agriculture [80,84]. Additionally, the tourism sector was negatively affected in the impacted regions [84]. Ecologically, the tailings slurry smothered or buried entire aquatic communities (Table S7). Protected areas and Krenak indigenous lands were coated with tailings [74,85], and metal-laden fine particles were detected over 220 km offshore, where they suppressed coral growth and tripled skeletal iron loads on the Abrolhos Bank [86].

The socio-economic consequences were equally dramatic. The collapse of Fundão completely destroyed the historic village of Bento Rodrigues, displacing its entire population. The direct and indirect damages have been valued at over USD 20 billion, while chronic exposure to contaminated water and food continues to pose a threat to public health [84,85,87,88].

3.2.3. Uranium Tailings Disposal into Lakes at the Athabasca Basin (Canada)

U mining in Canada has left a notable environmental legacy, particularly in the northern regions where early operations lacked effective regulation. During the second half of the 20th century, Elliot Lake (Ontario) and Uranium City (Saskatchewan) became the main production centres, dominated by underground mines [89,90,91,92,93,94].

At Gunnar (1955–1964), Beaverlodge (1952–1982), and Lorado (1954–1960) in northern Saskatchewan, mine and mill waste was directly discharged into lakes and natural topographic depressions. Although this subaqueous disposal masked the waste, it failed to provide long-term containment, and its consequences became evident decades later [89,91,94,95]. At Gunnar, millions of tonnes of acidic tailings entered Mudford Lake, later overflowing into Langley Bay of Lake Athabasca [96,97]. At Beaverlodge, waste discharged into Fookes and Marie Lakes eventually contaminated Beaverlodge Lake [98]. Lorado generated less waste but, lacking a large lake, deposited tailings beside Nero Lake, from where runoff dispersed contaminants [95].

Environmental consequences were substantial, as reflected by U concentrations in Beaverlodge Lake reaching 125 µg/L in 1985, far above drinking water limits. Models predicted a decline to 10 µg/L by 2118, more than a century later [99]. In Langley Bay, radionuclide-rich sludge accumulated on the lakebed, invertebrate populations declined, and bioaccumulation in fish led to fishing restrictions. Nero Lake experienced acidification and widespread water quality degradation. Lack of radiological protection caused elevated lung cancer mortality among miners exposed to radon and dust [90,97,99,100,101]. Local communities also lost access to safe water for domestic and agricultural use, while tailings deposits still pose health risks.

Following extensive remediation, Gunnar is expected to reach final closure in 2025, Lorado was rehabilitated in 2016, and Beaverlodge, also closed in 2025, remains under monitoring [95,102,103,104] (Table S7).

3.3. Dewatering-Induced Impacts

Underground as well as open-pit mines frequently need to lower the local groundwater table to ensure safe and economic operations. By pumping out more water entering into the mine (“ingress”) than can be naturally replenished by rainfall, the water table in the overlying aquifer is gradually lowered. This draw-down typically extends well beyond the actual mining site and may affect several hundreds to thousands of square kilometres of the surroundings.

During active mining, large volumes of groundwater are pumped out to keep underground mine workings dry, in a process known as dewatering, which is typically implemented through a network of underground sumps, dams, and pumping stations [105]. Sustained dewatering over extended periods depletes aquifers, reduces surface water flows, and can induce land subsidence due to the progressive lowering of the water table as the erosion base [106,107]. Effects include the drying up of boreholes, springs, smaller rivers, and wetlands, while increasing flow in streams receiving the pumped-out water, where it may lead to the deepening and widening of stream channels through higher fluvial erosion. As mine water is often of lesser quality than the original groundwater, this frequently causes problems to farmers and other downstream users.

3.3.1. Dewatering-Induced Ground Instability and Long-Term Pollution in Karst Aquifers of the Far West Rand Goldfield (South Africa)

The Far West Rand region, located within South Africa’s Witwatersrand gold belt, represents a prime example for illustrating the hydrogeological and geotechnical consequences of extensive dewatering by deep mining operations [108,109,110,111].

Large-scale pumping systems were installed to extract millions of litres of water per day, completely draining several overlying dolomitic compartments, and lowered the groundwater table by up to 1000 metres in places [108,110,111,112] (Table 3). This draw-down of the groundwater table triggered an unexpected acceleration of ground instability in karstic terrains, manifested as sinkholes and subsidence, causing fatalities and damage to roads, railway lines, houses, and other infrastructure [108,109,110,111]. Over six decades after dewatering commenced, sinkholes are still actively forming. In 2016, the municipality of Merafong declared a state of disaster in Khutsong, Welverdiend, and Carletonville after the sudden appearance of at least ten open sinkholes into which water and sanitation networks collapsed, leaving thousands of residents without basic services [113]. This region hosts one of the highest concentrations of sinkholes induced by human activity in South Africa, with more than 1200 recorded events and at least 38 confirmed fatalities [114]. In areas such as Khutsong, the intensification of ground collapse has led to a social and structural crisis. Numerous residents have abandoned their homes out of fear of being swallowed by sinkholes, and at least one documented case reports a person falling into one while walking down the street [115,116]. In certain compartments of the Far West Rand, attempts were made to fill sinkholes with tailings as a cost-effective means to reduce the amount of surface water recharging the dewatered karst aquifer and entering the underlying mine void during and after rain events. However, in many cases, tailings in nearly filled sinkholes collapsed into underlying receptacles, injecting large amounts of toxic tailings directly into the karst aquifer. Sinkholes forming underneath tailings dams due to the large outflow of seepage had a similar effect. As tailings leach U and other toxic metals, they pose a long-term threat to water quality after mine closure, when the karst aquifer will be rewatered again [107] (Table 3).

Table 3. Overview of cases studies relating to dewatering of aquifers and rewatering of mine voids. Abbreviations: ~: approximately; →: flows into; a: years; c.: century; M: million; m³/d: cubic metres per day; ML/d: megalitres per day; Mm³: million cubic metres; n.a.: not available; est.: estimated.

Dewatering-Induced Impacts
Studied Mine	Period	Void Volume	Extent of Dewatering	Pumping Volumes	Water Bodies Affected	Impacts/Costs	References
Far West Rand (South Africa)	1930s–present	Total mine void ~600 Mm3	Up to 1000 m drawdown of GW table	>200 ML/d	Dolomitic karst aquifer (Venters-post, Bank, Oberholzer, W. Gemsbokfontein)	>1200 sinkholes (≥38 fatalities); infrastructure collapse; tailings infill; persistent contamination after rewatering	[107,108,109,110,111,112,114,117]
Ruhr coal mines (Germany)	1800s–2018	Underground voids	~4450 km2 affected by dewatering cone	~70 Mm3/a (~192 ML/d) in perpetuity	Rivers Emscher and Ruhr; regional aquifers	≤25 m subsidence; reversal of direction of stream flow; inundation; 5.3 M residents reliant on perpetual pumping at costs of ~EUR 300 M/a).	[118,119,120,121,122]
Post-Closure Mine Rewatering Impacts
Studied Mine	Period	Void Volume/Area	Rewatering	Treatment Volumes	Water Bodies Affected	Impacts/Costs	Reference
Wheal Jane (Cornwall, UK)	17th c.–1991 (>300 a)	n.a.	1991–1992 (flooding and decant)	1991 decant ~5000 m3/d (5 ML/d) 1992 spill ~50 ML/d	Carnon River → Restronguet Creek → Falmouth Bay	Stream and estuary contamination; long-term cost of pump-and-treat: ~GBP 2 M/a	[123,124,125,126,127,128,129]
Western Basin (West Rand, Krugersdorp/Mogale, SA)	1887– 1998	~45 Mm3	1998–2002 (uncontrolled flooding, decant for next ten years)	Est. ~30 ML/d (pumped and treated)	Tweelopiespruit → Crocodile River	Damage of aquatic habitats and downstream game reserve; ecosystem; threat of UNESCO world heritage site; perpetual pump-and-treat required	[130,131,132,133,134]
Central Basin (Johannesburg, Central Rand SA)	1886– 2008	n.a.	Feb. 2008–2013 (uncontrolled flooding, decant avoided)	Est. ~50 ML/d * (pumped and treated)	Klip River → Vaal system	Flooding/corrosion threat to underground urban infrastructure; threat of flooding of touristic mine at Gold Reef City; sterilisation of gold reserves at old mines; salinisation of Vaal River (used for irrigation); backfilling of water treatment sludge into dewatered mine void; perpetual pump-and-treat required
Eastern Basin (East Rand, SA)	1890s–2010	n.a.	2011–2014 (uncontrolled flooding, decant avoided)	Est. ~80 ML/d * (pumped and treated)	Blesbokspruit → Suikerbosrand → Vaal River	Downstream Marievale wetland lost Ramsar status; salinisation of Vaal River used for irrigation; cost-intensive perpetual pump-and-treat required
U mine Schlema-Alberoda (Saxony, Germany)	1946–1990	~36.5 Mm3; 22 km2	1991–1997 (controlled flooding)	4.3–9.5 Mm3/a treated (~12–26 ML/d)	Zwickauer Mulde River	Long-term pump-and-treat required and costly disposal of treatment sludge	[135,136,137,138,139,140,141,142,143,144,145]
U mine Königstein (Saxony, Germany)	1964–1990 (in situ leaching after 1984)	~9 Mm3; 6 km2	1991–2001 (pump-and-treat) 2001–2013 (controlled stepwise flooding), 2013–2024: (steady level flushing)	2001–2017 ~2.9 Mm3/a (~8 ML/d, mean est.) ** pumped + treated	Elbe River	~22 Mm3 of acid-contaminated sandstone ongoing flushing of mine void and water treatment and disposal of water treatment sludge, continued injection of NaOH to neutralise void water
U mines at Ronneburg (Thuringia, Germany)	1950s–1990 (U underground and open-pit mining)	~26.7 Mm3 (19.4 Mm3 flooded)	1990s–2000s (controlled flooding)	Acidic water treatment from dumps and underground galleries; variable flows	Local tributaries → Weiße Elster system	Continued pump-and-treat and sludge disposal required
U coal mine Gittersee (Saxony, Germany)	1964–1990 (U, in situ leaching after 1984	~2.3 Mm3 (1.74 Mm3 flooded)	1990–2016 (remediation)	Decant via gravitational flow in adit to Elbe River	Elbe River	Periodic removal of sludge deposits from decants adits required in perpetuity

* Pumping rates are frequently overreported since hourly rates are extrapolated to ML/d even though pumps run only for 19 h per day. The rate for the Central Basin has been adjusted accordingly based on the pumping rate measured at ERPM (55 ML/d; due to a less deep level of current pumping and the associated decrease in ingress, a slight reduction is assumed to 50 ML/d). ** Mean annual pumping volume estimated by the author from the cumulative discharge reported in Jenk and Paul (2018) [137].

Table 3. Overview of cases studies relating to dewatering of aquifers and rewatering of mine voids. Abbreviations: ~: approximately; →: flows into; a: years; c.: century; M: million; m³/d: cubic metres per day; ML/d: megalitres per day; Mm³: million cubic metres; n.a.: not available; est.: estimated.

Dewatering-Induced Impacts
Studied Mine	Period	Void Volume	Extent of Dewatering	Pumping Volumes	Water Bodies Affected	Impacts/Costs	References
Far West Rand (South Africa)	1930s–present	Total mine void ~600 Mm3	Up to 1000 m drawdown of GW table	>200 ML/d	Dolomitic karst aquifer (Venters-post, Bank, Oberholzer, W. Gemsbokfontein)	>1200 sinkholes (≥38 fatalities); infrastructure collapse; tailings infill; persistent contamination after rewatering	[107,108,109,110,111,112,114,117]
Ruhr coal mines (Germany)	1800s–2018	Underground voids	~4450 km2 affected by dewatering cone	~70 Mm3/a (~192 ML/d) in perpetuity	Rivers Emscher and Ruhr; regional aquifers	≤25 m subsidence; reversal of direction of stream flow; inundation; 5.3 M residents reliant on perpetual pumping at costs of ~EUR 300 M/a).	[118,119,120,121,122]
Post-Closure Mine Rewatering Impacts
Studied Mine	Period	Void Volume/Area	Rewatering	Treatment Volumes	Water Bodies Affected	Impacts/Costs	Reference
Wheal Jane (Cornwall, UK)	17th c.–1991 (>300 a)	n.a.	1991–1992 (flooding and decant)	1991 decant ~5000 m3/d (5 ML/d) 1992 spill ~50 ML/d	Carnon River → Restronguet Creek → Falmouth Bay	Stream and estuary contamination; long-term cost of pump-and-treat: ~GBP 2 M/a	[123,124,125,126,127,128,129]
Western Basin (West Rand, Krugersdorp/Mogale, SA)	1887– 1998	~45 Mm3	1998–2002 (uncontrolled flooding, decant for next ten years)	Est. ~30 ML/d (pumped and treated)	Tweelopiespruit → Crocodile River	Damage of aquatic habitats and downstream game reserve; ecosystem; threat of UNESCO world heritage site; perpetual pump-and-treat required	[130,131,132,133,134]
Central Basin (Johannesburg, Central Rand SA)	1886– 2008	n.a.	Feb. 2008–2013 (uncontrolled flooding, decant avoided)	Est. ~50 ML/d * (pumped and treated)	Klip River → Vaal system	Flooding/corrosion threat to underground urban infrastructure; threat of flooding of touristic mine at Gold Reef City; sterilisation of gold reserves at old mines; salinisation of Vaal River (used for irrigation); backfilling of water treatment sludge into dewatered mine void; perpetual pump-and-treat required
Eastern Basin (East Rand, SA)	1890s–2010	n.a.	2011–2014 (uncontrolled flooding, decant avoided)	Est. ~80 ML/d * (pumped and treated)	Blesbokspruit → Suikerbosrand → Vaal River	Downstream Marievale wetland lost Ramsar status; salinisation of Vaal River used for irrigation; cost-intensive perpetual pump-and-treat required
U mine Schlema-Alberoda (Saxony, Germany)	1946–1990	~36.5 Mm3; 22 km2	1991–1997 (controlled flooding)	4.3–9.5 Mm3/a treated (~12–26 ML/d)	Zwickauer Mulde River	Long-term pump-and-treat required and costly disposal of treatment sludge	[135,136,137,138,139,140,141,142,143,144,145]
U mine Königstein (Saxony, Germany)	1964–1990 (in situ leaching after 1984)	~9 Mm3; 6 km2	1991–2001 (pump-and-treat) 2001–2013 (controlled stepwise flooding), 2013–2024: (steady level flushing)	2001–2017 ~2.9 Mm3/a (~8 ML/d, mean est.) ** pumped + treated	Elbe River	~22 Mm3 of acid-contaminated sandstone ongoing flushing of mine void and water treatment and disposal of water treatment sludge, continued injection of NaOH to neutralise void water
U mines at Ronneburg (Thuringia, Germany)	1950s–1990 (U underground and open-pit mining)	~26.7 Mm3 (19.4 Mm3 flooded)	1990s–2000s (controlled flooding)	Acidic water treatment from dumps and underground galleries; variable flows	Local tributaries → Weiße Elster system	Continued pump-and-treat and sludge disposal required
U coal mine Gittersee (Saxony, Germany)	1964–1990 (U, in situ leaching after 1984	~2.3 Mm3 (1.74 Mm3 flooded)	1990–2016 (remediation)	Decant via gravitational flow in adit to Elbe River	Elbe River	Periodic removal of sludge deposits from decants adits required in perpetuity

* Pumping rates are frequently overreported since hourly rates are extrapolated to ML/d even though pumps run only for 19 h per day. The rate for the Central Basin has been adjusted accordingly based on the pumping rate measured at ERPM (55 ML/d; due to a less deep level of current pumping and the associated decrease in ingress, a slight reduction is assumed to 50 ML/d). ** Mean annual pumping volume estimated by the author from the cumulative discharge reported in Jenk and Paul (2018) [137].

3.3.2. Perpetual Dewatering at Former Hard Coal Mines in the Ruhr Area (Germany)

In Western Germany, the Ruhr area stands as one of the country’s most densely populated and historically industrialised regions, with a current population of over 5 million people [146,147]. During the 19th and 20th centuries, it emerged as a major centre for underground hard coal mining in Europe via extensive, deep-level operations of up to nearly 1000 m below surface that lasted for over 150 years [118,147] (Table 3). The extraction of large amounts of coal from shallow depths together with large-scale dewatering resulted in widespread land subsidence. Vertical displacements exceeding 25 m in several urbanised areas not only damaged existing infrastructure like houses, underground reticulation systems, and roads but also reversed the flow direction of rivers that no longer drained towards the main river and created inundated depressions, commonly referred to as polders, that lie below the natural elevation of river systems and depend on artificial drainage to prevent flooding. If underground pumping would be stopped and groundwater allowed to recover to natural levels, much of the densely populated Ruhr area (some 4450 km²) would be flooded [118,119,120,121,122,148,149,150,151].

According to RAG, around 70 Mm³/a (~192 ML/d) of mine water needs to be pumped out in perpetuity at a current cost of some EUR 300 million per year, which is set to rise in line with increasing electricity tariffs [14,119,150,151] (Table 3).

3.4. Post-Closure Impacts of Flooded Mine Voids (Rewatering)

The progressive flooding of mine voids after mine closure (rewatering) marks a key hydrogeological transition. Once pumping ceases, water levels begin to rise either passively through naturally infiltrating groundwater or actively through the injection of surface water [152,153]. Depending on the local climatic water balance and volume of the mine void, complete flooding may take years to decades. Where mines are simply abandoned, uncontrolled flooding eventually leads to contaminated mine water overflowing via shafts, boreholes, adits, or natural fractures into nearby streams and shallow aquifers [152,153].

The inflow of water into underground mine workings or open pits on the surface frequently triggers long-term geochemical reactions degrading water quality [154,155].

Rewatering also carries geotechnical risks. Flooded voids and increased groundwater levels can reactivate ground instability, compromising the stability of basement structures especially in low-lying areas [156,157,158]. In open-pit settings, pit lakes often form, typically acidic and metal-laden, that usually require active neutralisation through liming and long-term monitoring to reach acceptable ecological conditions [156,157,158,159].

3.4.1. AMD from Centuries-Old Wheal Jane Tin Mine (United Kingdom)

Wheal Jane, a tin mine in Cornwall, was intensively exploited from the seventeenth century and, during the twentieth century, became interconnected with adjacent workings, creating an extensive system of galleries several hundred metres deep [123]. The mine was characterised by exceptionally high groundwater inflows, requiring continuous pumping of up to 60,000 m³ of water per day (60 ML/d). This water was strongly acidic (pH of 2.8) and had elevated concentrations of iron, zinc, cadmium, and copper. Although partially treated, discharges had already degraded the Carnon River and the downstream Restronguet Creek estuary prior to closure [127,160,161].

Following mine abandonment in 1991, rising groundwater levels led to overflow via historic adits, as predicted by hydrogeological models [123,129] (Table 3). In 1992, the collapse of a blockage in the Nangiles adit triggered a sudden release of approximately 50 million litres of AMD. The resulting plume, dominated by Fe oxides and with high concentrations of cadmium, copper, zinc, and arsenic, rapidly spread through the Carnon River to Restronguet Creek and Falmouth Bay, causing severe ecological damage [124,126,127].

Emergency measures included re-pumping and the temporary storage of water in the Clemows Valley tailings dam, but these proved ineffective given the extreme contamination [124]. In 1994, a pilot programme tested passive systems such as lagoons and biogeochemical processes, but they were found unable to cope with flows exceeding 300 L/s [128]. Consequently, in 2000, a permanent active treatment plant was commissioned to precipitate metals and control acidity by liming. Each year, approximately 5.6 million m³ (15.3 ML/d) of mine water are treated, removing more than 429 tonnes of iron together with other metals [124,125].

3.4.2. AMD from Uncontrolled Flooded Gold Mines in the Western, Central, and Eastern Basins (South Africa)

After more than a century of deep-level gold mining in the Western Basin, dewatering stopped in 1998 when Randfontein Estate Gold Mine (REGM), as the last active mine in the basin, ceased operations. As a consequence, the underground system of hydraulically interconnected mine voids originally belonging to different owners began to refill with infiltrating groundwater that, in 2002, eventually overflowed via a borehole (and later an incline shaft) into the adjacent Tweelopiespruit [130,131,132] (Table 3). The decanting mine water consisted of undiluted AMD with extreme low pH (<2), high EC values, and elevated levels of sulphate, iron, manganese, and uranium, leading to fish kills and the bioaccumulation of metals in sediments and crops. Precipitating iron hydroxide (“yellow boy”) soon covered the entire bed of the Tweeloopiespruit as the receiving stream, extending up to and beyond the Krugersdorp Game Reserve, where several lions were reported dead under circumstances possibly linked to contaminated water [108,110,111]. These impacts also posed risks to local communities through irrigation and groundwater use as well as to tourism as the downstream UNESCO World Heritage site at the dolomitic Sterkfontein Caves (Cradle of Humankind) was reported in the media to be threatened as well [130,132,133] (Table 3).

In the Central Basin, beneath Greater Johannesburg, pumping stopped unplanned following an explosion of the last remaining pumping chamber at ERPM gold mine setting off the uncontrolled flooding of the mine void. Projections at the time indicated that AMD in the flooded underground mine void system would corrode deep-reaching concrete basements of high-rise buildings in the Central Business District (CBD) and eventually decant into densely populated urban areas, polluting shallow aquifers as well as streams in the process [133]. In response, to the recommendation of an appointed Government Task Team, a pumping and treatment plant was installed and operated by the Trans-Caledonian Tunnel Authority (TCTA) of the Department for Water Affairs and Forestry to keep the mine water table below the “environmental critical level” (ECL) [130]. An independent study commissioned by two major banks with headquarters in the CBD found that no risks exist for compromising the stability of high-rise buildings or for potential outflows into built-up areas. It also pointed out that streams or aquifers in the area would not deteriorate in quality as all potentially affected streams already displayed pollution levels for the past few decades that were even greater than those observed in the decanting AMD. This pollution was mainly caused by runoff and seepage from tailings dams covering large proportions of their relatively small headwater catchments. The report also found that decant-related ground instability in dolomitic areas proposed by the Government task team can be excluded as a potential risk (Table 3).

In the Eastern Basin, the Grootvlei mine was the last to maintain active dewatering until early 2011. Pumping stopped following the change in ownership of the mine and its subsequent (illegal) stripping by the new owners, who again had made no provision for a controlled flooding of the associated mine void [130]. In its final years, part of the pumped water was discharged untreated into the Blesbokspruit, contributing to the degradation of the Marivale wetland and loss of its Ramsar status. A TCTA-run pump-and-treat station was later commissioned here as well, operating in much the same way as in the other two basins. Ever since, contaminants continue to reach the Vaal River system, compromising the quality of stream water used for irrigation and drinking water supply [132] (Table 3).

3.4.3. Controlled Flooding of Uranium Mines in Southeast Germany

After World War II, East Germany became one of the world’s leading uranium producers under the Soviet company Soviet Joint Stock Company (SAG) Wismut, later operated as a Soviet–German joint state enterprise (SDAG Wismut). Mining was concentrated in southern Saxony and eastern Thuringia, across geologically diverse deposits that left a legacy of environmental contamination and land degradation [135,136,137,141,162,163,164]. Affected mining sites include Schlema-Alberoda, Königstein, Ronneburg, and Gittersee (Table 3). Operating from 1946 to 1991, Wismut produced well over 200,000 tonnes of uranium, making it the single largest uranium producer worldwide, ranking fourth in total output behind entire countries such as the Soviet Union, the United States, and Canada [143,165]. Since 1991, when SDAG Wismut was transformed into Wismut GmbH, a dedicated remediation company wholly owned by the German Federal Government, approximately EUR 7 billion has been spent to date on addressing the legacy of more than 40 years of Cold War uranium mining [138,139,140,143].

Water-relevant remediation activities include the controlled flooding of underground mine voids and the operation of associated pump-and-treat systems (e.g., Schlema-Alberoda and Königstein), the dewatering and capping of tailings ponds, as well as the reshaping and covering of numerous waste rock dumps [135,141,162,163]. Drainage systems installed at remediated tailings ponds and rock dumps ensure that most seepage and runoff are now captured and treated before being discharged into public streams. At nearly all sites, long-term water management (pump-and-treat) systems remain necessary as long as contaminant levels exceed regulatory thresholds [137,138]. In addition to the energy demand for pumping and treatment, this also entails costs for chemicals used in water purification and for managing the resulting sludge, which are placed in engineered containment cells within remediated mine waste facilities (Table 3).

The Schlema-Alberoda mine represents one of the most technically complex remediation projects of Wismut, involving controlled and stepwise underground flooding, treatment of contaminated mine and seepage water, and the long-term stabilisation of waste rock dumps [141,142].

In contrast to most other Wismut sites requiring continuous pumping and treating of water from flooded mine voids at the Gittersee mine, excess water from the flooded void requires neither pumping nor active treatment but flows gravitationally via two adits into the Elbe River. Although no energy is required for either pumping or treatment, ongoing monitoring and maintenance of the adits still result in substantial long-term costs. These include the periodic removal of precipitated iron hydroxide sludge accumulating at the bottom of the two adits as well as its disposal. Containing elevated concentrations of natural radionuclides and toxic metals, the sludge is partly dewatered and disposed of in specially designed containment cells at the nearby Königstein mine [139,143].

3.5. Impacts from Discharged Mine Effluents

Among the multiple pathways through which mining activities affect water resources, discharges of effluents represent a well-documented yet often underestimated mechanism of contamination. These can be broadly divided into controlled point-source discharges and (usually uncontrolled) diffuse outflows. Controlled discharges are regulated and monitored releases of effluents through engineered outlets such as canals and pipes, whereas uncontrolled diffuse outflows occur through seepage, stormwater runoff, or leaking infrastructure, typically without prior treatment [1,166].

While point discharges are commonly subjected to regulatory monitoring, discharges from older or abandoned mining sites often occur with little or no oversight [167].

The chemical composition of discharged effluents varies depending on ore type, processing technologies, and local geology, but often includes elevated concentrations of heavy metals (e.g., iron, copper, zinc, and lead), sulphates (from pyrite weathering), high acidity (ditto), and in some cases, naturally occurring radionuclides such as U or Ra [168,169]. Point discharges can cause severe impacts, particularly when they contaminate drinking water sources, agricultural lands, or sensitive aquatic ecosystems [169,170] not only locally but up to hundreds of kilometres downstream. Over time, significant contaminant reservoirs can build up in fluvial sediments especially in slow-moving water bodies such as dams, wetlands, or lakes. Their environmental and socio-economic consequences can be disproportionate, especially in cases of acute failures or in streams with limited dilution and attenuation capacity.

3.5.1. Ok Tedi Copper–Gold Mine (Papua New Guinea)

The Ok Tedi mine, an open-pit copper and gold operation in Papua New Guinea’s Western Province, exemplifies one of the most extreme cases of intentional, long-term point-source discharge of mining effluents into a natural water system. Initially approved under the condition that a tailings dam would be constructed, the project changed course after a 1984 earthquake-triggered landslide destroyed the tailings dam under construction. The company abandoned any further attempts of tailings containment and proceeded to discharge their tailings and process water directly into the Ok Tedi River [171,172] (Table S8).

Over more than three decades, tens of millions of tonnes were transported downstream into the Fly River system. The sediment load altered the river morphology and raised the floodplain, while seasonal floods deposited contaminated sediments over lowland agricultural areas leading to elevated concentrations of copper, zinc, and lead. This had cascading effects on crops and riparian forests [173], posing risks to aquatic ecosystems [174]. Consequently, indigenous riverine communities suffered the collapse of fisheries, crop failure, and limited access to clean water, compromising food security and public health [172]. Although the mine brought roads, health services, and employment, long-term benefits for local populations were limited, and many communities experienced net losses in autonomy and sustainability [175,176,177,178].

3.5.2. Nickel–Zinc Sulphide Mines at Talvivaara (Finland)

The Talvivaara deposit in Sotkamo (east-central Finland), now operated as the Terrafame Sotkamo mine, is a large low-grade nickel sulphide ore body that also contains zinc, copper, and cobalt [179,180,181]. Discovered in 1977, it remained undeveloped for decades due to technical and economic limitations. In the early 2000s, Talvivaara Mining Company acquired the rights from Outokumpu and, in 2008, began production using bio-heap-leaching, an uncommon approach in Europe for this type of deposit [179,180,181,182]. Acidic irrigation of crushed ore promotes the microbial dissolution of metals, producing solutions from which nickel–cobalt sulphide is precipitated and later refined into industrial-grade nickel–cobalt sulphate. The value chain has recently expanded to battery-precursor production and uranium recovery as a by-product [180,182].

The project was considered strategic for Finland due to the domestic production of critical metals and regional economic benefits. However, environmental management proved challenging, and several leaks occurred before a major accident in 2012 [179]. The failure originated in the gypsum pond, designed to contain semi-solid sludge from water treatment but increasingly used to also store large volumes of acidic process waters during wet periods. Elevated liquid levels and subsequent increases in hydrostatic pressure, combined with a defective basal sealing and internal erosion, caused the pond to fail [183].

Millions of cubic metres of contaminated water were released, with hundreds of thousands entering the environment (Table S8). The spill was highly acidic and contained elevated concentrations of sulphate, metals, and radionuclides such as U [181]. Estimated loads included ~150 tonnes of iron, 150 tonnes of manganese, 2 tonnes of nickel, 1 tonne of zinc, 70 kg of uranium, 60 kg of cobalt, and 2 kg cadmium [184]. The discharge spread through the Kalliojärvi and Kivijärvi Basins and downstream river systems, reaching concentrations capable of causing acute toxicity to aquatic organisms [183,184,185,186,187].

3.6. Hydrological Alterations

Mining not only leads to water pollution and reduced water availability for local communities but frequently also alters natural water balances as well as flow regimes in rivers and groundwater [188,189].

Generally, these transformations are not incidental but the result of strategic planning to meet the water needs of mines. Specifically, the construction of dams, dewatering of aquifers, inter-basin water transfers, and channel realignment are commonly implemented to secure water supply or facilitate the discharge and dilution of mine effluents [188]. Such interventions and their consequences often alter key processes such as infiltration, aquifer recharge, runoff, and hydrological connectivity between ecosystems [189,190]. In many instances this includes unforeseen consequences like the massive occurrence of dewatering-related sinkholes in the karst area of the Witwatersrand Basin that irreversibly changed the natural runoff–recharge ratio of an entire catchment resulting in reduced stream flow and changed spring yields. Many of these infrastructures outlast the mining operations and eventually are embedded into regional water management schemes [188,189,190].

3.6.1. Dewatering-Induced Alterations of Natural Water Balances: Far West Rand Goldfield (South Africa)

As described earlier, the large-scale dewatering of the dolomitic compartments in the Far West Rand had profound and lasting effects on the hydrology of the Wonderfonteinspruit catchment. The permanent loss of major karst springs eliminated the perennial baseflow that sustained local agriculture and livestock, dried-up high-yield boreholes, and reduced streamflow across the basin [62,191]. These changes propagated far beyond the mining area, reshaping catchment-scale water availability.

Dewatering also triggered the formation of more than 1200 sinkholes, which intercepted surface runoff and, in places, opened directly within the former streambed, diverting much of the remaining stream flow to the underlying karst aquifer [108,109]. This, in turn, increased inflow into the mine workings beneath the karst aquifer, undermining the initial dewatering efforts.

To reduce this unintended recharge, mining companies diverted the stream flow of the Wonderfonteinspruit into a 30 km-long pipeline of 1 m-diameter running across the three dewatered compartments. While effective in preventing losses to the subjacent karst aquifer, the diversion desiccated the original riverbed, destroyed wetlands, and exposed contaminated sediments that oxidised and released metals such as U and heavy metals [107,131,192]. Heavy rainfall repeatedly exceeded the pipeline’s capacity, causing floodwaters to re-enter the old channel and flushing out liberated contaminants from the oxidised sediments. Operators attempted to limit the resultant increase in ingress by filling sinkholes that had formed in the streambed with freshly produced tailings for weeks on end only to find that the tailings had finally collapsed into the underlying karst receptacles, creating potential sources of secondary water pollution once active dewatering ceases and groundwater levels rebound [107,131].

Mining also pierced through the dykes that used to separate the dolomitic compartments from each other, effectively merging them into a single “mega-compartment” and irreversibly altering regional groundwater flow and springs yields after mine closure [107,193].

3.6.2. Creating Water-Negative Post-Mining “Lakescapes” in Lignite Regions of Central and Eastern Germany

The onset of industrialisation transformed the landscapes of Central and Eastern Germany through the large-scale extraction of lignite, particularly in the Central German coal field (Halle–Leipzig–Bitterfeld region) and the Lusatian coal field (southern Brandenburg and northern Saxony). Both regions are located in two of Germany’s driest zones, with mean annual precipitation ranging between 441 and 606 mm in the Central German field (known as the “Mitteldeutsche Trockengebiet”) and 560 and 640 mm in Lusatia, both well below the national average [23].

Mining expanded rapidly during the 20th century, reaching 145 Mt/a in Central Germany by 1963 and 200 Mt/a in Lusatia in 1988, with cumulative outputs of 8.7 billion t and 8.4 billion t, respectively [194,195]. After German reunification in 1990, production in both regions declined sharply, and as part of the national decarbonisation plan, a complete lignite mining phase-out was set for 2038 [196].

Since 1992, more than EUR 12 billion has been invested in remediating closed lignite mines in Lusatia alone, covering roughly 1000 km², with the state-owned remediation company LMBV projecting an additional EUR 4.8 billion over the next 25 years [197].

Open-pit mining required continuous dewatering of the pits from the early 1900s onward. As pits deepened, groundwater inflow intensified, forcing companies to install pumping wells around each mine. The extracted water was discharged into nearby rivers such as the Pleiße, Mulde, Spree, and Elster, where it increased streamflow, sometimes doubling pre-mining rates. At the same time, regional groundwater levels were lowered on a grand scale affecting a total surface area of some 3000 km² [27,194].

Aquatic ecosystems such as the UNESCO biosphere reserve Spreewald, and even waterworks in Berlin (the German capital) grew dependent on this artificially elevated streamflow. With the cessation of mining and the associated stop of active dewatering, flow in affected rivers dropped drastically, prompting proposals such as intercepting water from the nearby Elbe River to sustain the Spree [195]. Increased flow velocities and the associated channel erosion during active lignite mining had further deepened stream beds, which during low flow now drew more baseflow from their alluvial aquifers, exacerbating the regional groundwater depletion. The extent of mining, dewatering, and related water impacts in the Central German and Lusatian coal fields are summarised in Table S9.

Once mining ceased, the vast open pits could not be backfilled, as the excavated lignite was burnt for energy generation leaving substantially lower volumes of ash behind. Instead, the open lignite pits were allowed to gradually fill up with naturally infiltrating groundwater and meteoric water, creating hundreds of pit lakes. As natural flooding is slow in these semi-arid regions and prolonged exposure to weathering is prone to destabilise the newly contoured slopes of the pits, flooding was accelerated by diverting water from nearby rivers into the pits particularly during flood events such as the one in 2013 in the Saale and Mulde catchments (Central German coal field) [194].

The large-scale formation of pit lakes and the associated rebound of regional groundwater levels have created new environmental challenges. The oxidation of pyrite-bearing sands exposed during dewatering now produces acid mine drainage (AMD) when re-submerged by rising groundwater. With such groundwater feeding into rivers and newly formed lakes, the pH values of some lakes dropped to 2–4 accompanied by rising sulphate and iron concentrations, with the latter causing a distinct brown discolouration of streams (“Braune Spree”; “Braune Pleiße”) [194]. To counteract acidification, lime is applied as a neutralising agent via specifically designed ships criss-crossing the lakes and is added in several water treatment plants neutralising water before it is released into rivers. Iron hydroxide sludge accumulating in stream beds requires regular dredging to prevent river channels from being clogged. The removed sludge is partly dewatered and deposited in dedicated disposal sites at additional costs. In order to reduce the latter, experiments are currently conducted to test the potential of such sludge to act as a soil improver [194,197].

Once all pit lakes are fully flooded, the regional water balance will be significantly altered. By replacing vegetated land with an open water surface, mining has drastically increased evaporation losses in the region. Over lakes, where evaporation is unrestricted by protective cover layers, potential and actual evaporation are equal [23,24,25,26]. On average, German lakes lose 730–750 mm/a of water to the atmosphere, while receiving only ~640 mm/a of rainfall resulting in an annual deficit of 13–16% of received rainfall, which rises to nearly 100% during spring and summer [23,24,25,26] (Table 1; Figure 3).

In the two lignite coal fields, approximately 24% of the mined land, about 327 km², will ultimately turn into open water. Our calculations indicate that this transformation of land use has rendered the Central German coal field fully water-negative by shifting the annual water balance from a pre-mining surplus (+32 Mm³/a) to an overall deficit (−44 Mm³/a) [27,194]. The Lusatian field fares slightly better, as the deficit calculated in this paper (−84 Mm³/a) still leaves the region water-positive despite losing nearly 60% of its pre-mining water balance [28]. Combined, the two regions are set to lose about 73% of the freshwater available before mining started (Table 4). Without continuous inflow from rivers draining catchments upstream of the Central German coal field, water levels in the newly formed pit lakes would gradually recede and draw in groundwater, thereby further lowering the regional water table. This, in turn, would eventually cut off baseflow to wetlands, small streams, and riparian ecosystems, and cause the wilting of shallow-rooted vegetation. While the new “lakescapes” with their beaches, marinas, and waterways project a newly created abundance of water to the general public, they are in fact, somewhat counterintuitively, contributing to drying out two of Germany’s driest regions even further and increasing their dependency on rivers importing water from elsewhere (Table 4).

Table 4. Climatic water balances in the Central German and Lusatian coal fields before and after the complete flooding of open pits (MAP—mean annual precipitation; MAE—mean annual evaporation; CWB—climatic water balance).

Parameter	Central German Coal Field	Lusatian Coal Field	Combined (a)
Total mine area [194] (% of combined area)	485 km2 36%	875 km2 64%	1360 km2 100%
Projected pit lake area [27] (% of total mine area)	119 km2 (24%)	208 km2 (24%)	327 km2 (24%)
Remaining land surface	366 km2	667 km2	1033 km2
Regional MAP [24,25] MAE grass on sand	~500 mm/a 435 mm/a	~600 mm/a 435 mm/a	~565 (b) mm/a 435 mm/a
Pre-mining CWB (c) (CWB [mm/a] × lake surface area)	65 mm/a (32 Mm3/a)	165 mm/a (144 Mm3/a)	129 mm/a (b) (176 Mm3/a)
MAP over pit lakes (a) (=82% of regional MAP)	408 mm/a (=500 × 0.82)	492 mm/a (=600 × 0.82)	463 (b) mm/a
MAE from pit lakes (15 m-deep) [24,25]	>730 mm/a	>730 mm/a	>730 mm/a
CWB of pit lakes (15 m-deep) (a)	−322 mm/a	−238 mm/a	−267 (b) mm/a
	−36 Mm3/a	−50 Mm3/a *	−86 Mm3/a
Lost groundwater recharge (c)	–8 Mm3/a (=65 mm/a × 119 km2)	−34 Mm3/a (=165 mm/a × 208 km2)	−42 Mm3/a (sum)
Mining-induced water loss (% of pre-mining balance)	−44 Mm3/a (−138%)	−84 Mm3/a (−58%)	−128 Mm3/a (−73%)

^(a) Own calculations, all values rounded to account for underlying uncertainties; ^(b) area-weighted average; ^(c) for simplicity reasons, pre-mining land use is assumed to consist exclusively of grassland over sand, somewhat overestimating the actual water loss as evaporation from unvegetated farmland and sealed areas in villages and towns as other pre-mining land uses are lower. * This estimate aligns well with the up to 60 Mm³/a needed by the River Spree according to the [198].

Table 4. Climatic water balances in the Central German and Lusatian coal fields before and after the complete flooding of open pits (MAP—mean annual precipitation; MAE—mean annual evaporation; CWB—climatic water balance).

Parameter	Central German Coal Field	Lusatian Coal Field	Combined (a)
Total mine area [194] (% of combined area)	485 km2 36%	875 km2 64%	1360 km2 100%
Projected pit lake area [27] (% of total mine area)	119 km2 (24%)	208 km2 (24%)	327 km2 (24%)
Remaining land surface	366 km2	667 km2	1033 km2
Regional MAP [24,25] MAE grass on sand	~500 mm/a 435 mm/a	~600 mm/a 435 mm/a	~565 (b) mm/a 435 mm/a
Pre-mining CWB (c) (CWB [mm/a] × lake surface area)	65 mm/a (32 Mm3/a)	165 mm/a (144 Mm3/a)	129 mm/a (b) (176 Mm3/a)
MAP over pit lakes (a) (=82% of regional MAP)	408 mm/a (=500 × 0.82)	492 mm/a (=600 × 0.82)	463 (b) mm/a
MAE from pit lakes (15 m-deep) [24,25]	>730 mm/a	>730 mm/a	>730 mm/a
CWB of pit lakes (15 m-deep) (a)	−322 mm/a	−238 mm/a	−267 (b) mm/a
	−36 Mm3/a	−50 Mm3/a *	−86 Mm3/a
Lost groundwater recharge (c)	–8 Mm3/a (=65 mm/a × 119 km2)	−34 Mm3/a (=165 mm/a × 208 km2)	−42 Mm3/a (sum)
Mining-induced water loss (% of pre-mining balance)	−44 Mm3/a (−138%)	−84 Mm3/a (−58%)	−128 Mm3/a (−73%)

^(a) Own calculations, all values rounded to account for underlying uncertainties; ^(b) area-weighted average; ^(c) for simplicity reasons, pre-mining land use is assumed to consist exclusively of grassland over sand, somewhat overestimating the actual water loss as evaporation from unvegetated farmland and sealed areas in villages and towns as other pre-mining land uses are lower. * This estimate aligns well with the up to 60 Mm³/a needed by the River Spree according to the [198].

Future climatic changes due to the burning of coal and other fossil fuels are predicted to result in rising temperatures that, in turn, would increase water losses through evaporation worsening water scarcity in the dry lignite mining regions, and everywhere else where sufficient streamflow is needed to cope with discharged mine effluents. Long-term meteorological data of the German weather service (DWD) from 1893 to 2024 suggest that average annual evaporation from open water across Germany (6 m-deep reference lake) indeed increased by 55 mm/a (7.5%). Fortunately, this is more than off-set by a simultaneous 8.9% increase in rainfall (65 mm/a) over much of the same period (1881–2023) resulting in a net increase in available water across the entire country of 10 mm/a. This increase in precipitation is most pronounced in Eastern Germany where between 1991 and 2020, figures rose by 2.5–15% compared to the preceding climatic normal period (1961–1990) [26]. While this is not enough to compensate for the mining-induced evaporation losses in the two lignite regions, it is a mitigating factor.

4. Discussion

4.1. Type and Consequences of Impacts

The case studies reviewed indicate that mining-related impacts on water are often far more severe, longer-lasting, costly, and extensive than originally anticipated [10]. The impacts of mines frequently extend well beyond the actual mining sites, affecting users and aquatic ecosystems over hundreds of kilometres downstream and up to thousands of square kilometres around the mine [10,54,108,110,111]. Some effects diminish over time, but others, are irreversible and will persist in perpetuity [131,194]. Because many legacy sites require ongoing long-term post-closure water management at significant cost at times when they no longer generate revenue, private companies frequently default on their obligations, effectively transferring liabilities to government and civil society [199]. Mining, in many instances, reduces water availability and degrades quality to levels unsuitable for human or ecological use without being held responsible for the associated knock-on effects and costs imposed on third parties (externalised costs) [17]. While some mines have expanded local water supply, e.g., by installing desalination plants or by piping water from elsewhere, such examples remain rare and still need to prove how sustainable they are once mining ceases [6].

Mining-related water impacts observed in the reviewed case studies can broadly be grouped into two recurrent patterns: high-frequency, low-magnitude impacts, such as the continuous outflow of relatively low volumes of tailings seepage, and low-frequency, high-magnitude impacts, such as catastrophic failures of tailings storage facilities suddenly releasing large volumes of tailings into the environment. Some dewatering-related processes, such as progressive sinkhole formation, occupy an intermediate position by continuously unfolding over decades with limited damage while occasionally resulting in catastrophic consequences.

4.2. Patterns and Extent of Mining-Related Water Pollution

Across different pollution pathways (i.e., effluent discharges, mine waste-related contamination, and acid mine drainage), a consistent pattern emerges in which the primary metal mined represents the dominant contaminant in impacted waters. While this holds for the majority of metal mines analysed [3,14], it does not apply to gold mining where co-occurring elements (like uranium in the case of South Africa) are the main contaminants of concern. In coal mining, this often is mercury [52,133]. Despite the widespread occurrence of AMD, many impacted streams retain circum-neutral pH values, highlighting that severe contamination is not confined to acid conditions [49,50,51].

Measured maximum contaminant levels in impacted streams were about five times higher than the reported average concentrations, ranging from 1.5 to 33 times. The higher ratios (15–33) were only observed for two streams, while all others displayed maxima between 1.5 and 4 times the respective average concentrations.

To enable the comparison of different contaminants, enrichment factors (EFs) were used to normalise reported concentrations against global natural background levels [9,12,200]. EF values for the various case studies indicate that mean concentrations of primary contaminants in mining-affected streams exceed natural background levels by about two orders of magnitude (~240 times on average) rising to four orders of magnitude for the highest reported maximum (EF = 27,000; U in the Klip River near Johannesburg). For accompanying (secondary) contaminants, enrichment factors are generally significantly lower. At an average EF of 0.3, the analysed contaminants did not even reach their respective background levels, while the highest recorded maximum exceeded them by 5700 times.

4.3. Translating Water Quality Impacts into Water Quantities

While such water quality metrics allow intercomparisons between pollution levels at different mines, they do not facilitate comparisons with water quantity impacts. In order to achieve this, it is proposed to express water quality degradation in terms of the hypothetical volume of clean water required for diluting the mining contaminants back to pre-impact levels. For example, if a mine increases contaminant concentrations in a receiving stream say from 1 to 10 mg/L, it would need ten times the streamflow volume at the point of impact of clean water to dilute contaminant levels in the stream back to pre-impact conditions (for a stream with a flowrate of say 10 ML/d, this would translate into 100 ML/d of clean water that the mine would have to add). Applying the average EF for primary contaminants mentioned above, mines would need to augment flow in impacted water courses by 240 times. While this is hardly feasible in practice, it would help to visualise the often-invisible water quality changes across all types of contaminants. By using the EF for the main (primary) contaminant, all other accompanying pollutants (at lower concentrations in the receiving stream) are automatically addressed as well.

4.4. Monetisation of Water Quality Impacts: Internalising Externalised Cost

Since such volumes are rarely available to any mine, this dilution concept remains hypothetical. To make it applicable in practice, the calculated dilution volumes could be translated into monetary costs (i.e., pollution penalty fees). Using local water tariffs would reflect regional differences in water value. While the resulting fees may appear prohibitive at first, they are grounded in physico-chemical mass-balance considerations and are intended to fully internalise the costs of mining-related pollution. As such, they may be less arbitrary than many current penalty schemes, which often rely on less transparent criteria. Higher penalty fees could incentivise mines to invest in improved water treatment as a lower-cost alternative. At present, such investments are often not considered economically justified because many pollution-related costs are borne by third parties.

Using costs as a generic impact metric would also facilitate the inclusion of monetised indirect and knock-on effects, such as agricultural losses from using polluted water (crop failure, livestock losses), health impacts on exposed populations (increased disease burden, medical expenses, lost income, and costs of bottled water or household filtration), and ecological damage.

4.5. Synopsis

Preventing unaccounted for or underreported water impacts in future requires pre-mining assessments that move beyond conventional water reporting and incorporate processes commonly overlooked in current reporting systems [9,11,12]. While Environmental Impact Assessments (EIAs) are now widely required for mining projects, many of the impacts discussed here originate from operations approved before EIA legislation existed, or from assessments whose scope focused on operational phases and local compliance rather than long-term, off-site, or post-closure hydrological impacts [201]. The examples explored in this paper include decade-long or even perpetual pumping and treatment of mine water, stream diversions, drying up of springs and wetlands, contamination of irrigation water and municipal drinking water systems, drying out of entire regions, degradation of agricultural land and settlements through dewatering-induced ground instability, century-long discharge of excess water ecosystems and cities which grew dependent on that suddenly stop, and pollution of streams, floodplains, and fluvial sediments that through accumulating radioactive and toxic metals turn into long-term threats to water supply systems with associated risks for public health and food security.

Beyond formal regulatory frameworks, the effectiveness of water-related impact assessment and management in mining is strongly influenced by governance quality and institutional capacity. In several mining regions, limited technical resources, weak enforcement, fragmented responsibilities, or deficiencies in transparency and accountability can undermine monitoring, compliance, and long-term water management, even where regulations formally exist. Such governance constraints help explain why the cumulative, off-site, and post-closure water impacts documented in this review often persist despite established legal requirements. While a detailed governance analysis lies beyond the scope of this study, these factors represent an important contextual driver shaping the real-world performance of mining-related water management.

While some of the selected studies represent extreme cases in terms of magnitude, they all illustrate that water reporting should be significantly expanded to better reflect the full scale and variety of mining impacts. Externalised costs remain largely absent from formal assessments and seldom appear in standard evaluations, nor are the accumulated impacts of several mines on single water bodies captured adequately [32,33,118,119,146,202]. Further shortcomings relate to a lack of monitoring of withdrawals by source (local surface water or groundwater, imported or purchased water, water encountered underground, recycled water) as well as low levels of enforcement of now often adequately updated legislation due to capacity constraints or maladministration. Future reporting frameworks should incorporate basin-scale indicators, post-closure phases, and means to monetise externalities such as ecosystem service losses and social cost caused by pollution or depleting local water resources [11,199,203]. Examples are plentiful ranging from pollution-related agricultural losses and the resultant food insecurity, damage of industrial users who no longer can use water for cooling or other processes, need for additional treatment for water service providers, and citizens resorting to buying bottled water instead to actual health impacts on people and the associated suffering and financial burden (

Table 5. Representative knock-on effects and externalised costs from the case studies analysed in this paper.

Case Study (Direct Water Impacts)	Knock-On Effects	Associated Externalised Costs
Polymetallic mines at Cerro Rico, Bolivia (transboundary AMD pollution)	Long-distance metal transport (~200 km waterways; ~560 km sediments); bioaccumulation in crops, livestock, and wildlife; soil contamination; long-term health risks; structural instability of the mountain affecting local communities	Reduced agricultural productivity; food security losses; healthcare costs; loss of income for farmers; long-term sediment remediation costs borne by governments; cross-border environmental burden shared by downstream countries
Witwatersrand Au/U mines, South Africa (decanting AMD; tailings seepage)	Persistent AMD discharge; contamination of rivers, wetlands, and groundwater; bioaccumulation of U and metals in food chain, human hair; loss of Ramsar status; fish kills; threat to UNESCO world heritage site and game reserves	Perpetual pump-and-treat (~R400–600 M/a); public expenditure of ~ZAR10B for “short-term” solutions; medical costs linked to chronic exposure; loss of land usability; municipal water treatment costs; long-term ecological rehabilitation borne by the state
Polymetallic sulphide mines at Rio Tinto, Spain (stream pollution by AMD)	Persistent AMD affecting ~125 km of rivers; metal transfer to estuary and soils; loss of fish and macroinvertebrates; degradation of riparian vegetation; contamination affecting grazing livestock	Reduced agricultural and grazing potential; long-term treatment and dredging costs; economic losses for fisheries and local communities; public remediation of historic liabilities
Far West Rand, South Africa (extensive dewatering of karst aquifers)	>1200 sinkholes causing fatalities, damage to roads, railways, houses, water infrastructure; drying up of karst springs, drying out of wetlands; altered water balances; groundwater contamination by tailings injected into the karst aquifer	Loss of life, infrastructure repairs; buy-out and relocation of households; agricultural losses from irrigation shortfalls and polluted water; bottled water replacement; loss of ecosystem services; rising pumping costs if sinkholes increase ingress
Ruhr coal field, Germany (deep coal mines; perpetual pumping)	Land subsidence (up to 25 m vertical displacement); reversed drainage; creation of polders reliant on artificial pumping; inundation risk for urban areas	Perpetual pumping of ~70 Mm3/a (~EUR 300 million per year); infrastructure maintenance; long-term public liability borne by RAG (initially supported by public levees on coal-generated electricity, “Kohlepfennig”); increased costs for water management in urban areas (polders)
Wheal Jane Sn mine, UK (decanting AMD, stream pollution)	Sudden AMD release (50 ML/d); contamination from river to estuary and bay; destruction of aquatic habitat; metal-rich plume	Long-term active treatment (GBP 2 million per year); emergency pumping costs; sludge handling and disposal; public funding for environmental remediation
Ok Tedi Cu/Au mine, Papua New Guinea (tailings discharged into river)	Chronic intentional discharge of tailings; raised floodplain; destruction of riparian forests; loss of fisheries; toxic sediment deposition	Collapse of subsistence agriculture; loss of fisheries; community displacement; long-term health risks; government and community-led remediation with limited operator contribution
Ni/Zn sulphide mines at Talvivaara, Finland (collapse of tailings dam wall)	Failure of tailings pond; release of millions of m3 of acidic/metalliferous water; lake and river contamination; toxic effects on biota; radionuclide mobilisation	Emergency containment by company and state; long-term monitoring funded publicly; ecological losses; water treatment infrastructure; public capital injection after operator bankruptcy
Wismut U mines, Germany (controlled flooding of mine voids; tailings seepage)	Acidic and metal-rich mine waters; long-term contaminant loads (U, Ra, As); temporary increase in contaminant load during early flooding; seepage from tailings and waste rock; sludge disposal in containment cells; water hardness limits industrial downstream use	Remediation > EUR 7 billion to date, perpetual pump-and-treat at several sites; chemical inputs and sludge disposal costs; long-term monitoring; provision of alternative water supply to affected residents
Central German and Lusatian lignite mines (altered climatic water balance through pit lakes)	Stop of dewatering and drying out of streams and wetlands; flow reduction in major rivers; creation of partly acidic pit lakes; large evaporation losses from pit lakes exacerbate existing regional water scarcity	Remediation-induced regional water deficits; costly river diversions to maintain flows; liming of acidic lakes; dredging of iron hydroxide sludge from streambeds; threat to downstream UNESCO biosphere reserve, loos of agricultural land, damage to aquatic habitats

).

5. Conclusions

Using selected case studies from around the globe, this review illustrates how mining activities impact water resources at magnitudes and over time scales far exceeding what is typically considered by current permitting and reporting systems. Nearly all examined cases exhibit damage persisting over decades, centuries, or even millennia affecting whole rivers, entire catchments, and up to several thousands of square kilometres around dewatering mines. This relates to acid mine drainage, aquifer drawdown, toxic effluent release, and hydrological alteration at the local and regional scales. While mining itself at any given site is generally finite, its water legacy at many sites may not be. In fact, the most severe burdens arise often only after closure, yet post-mining phases remain frequently absent from current reporting systems or are only inadequately covered, frequently omitting processes that drive longer-term risks. These include multi-decade obligations to pump and treat water, diffuse contamination from weathered mine wastes, and permanent hydrological deficits. Given the globally increasing competition for water for food production, industry, and domestic needs, together with the projected expansion of mining to meet the soaring demand for minerals associated with the energy transition, a proactive approach to assessing and reporting mining-related water impacts is increasingly needed.

By identifying and categorising direct and indirect water impacts that are often omitted or insufficiently covered in standard water reports, this paper argues for a more comprehensive representation of mining-related water implications in future assessments.