Mine Water Rebound in the Former Hard Coal Mining Area of the Ruhr District – Hydrogeochemical Monitoring for a Sustainable, Ecological and Economical Feasible Post-Mining Approach

Mine water rebound including mine water management and water treatment are key perpetual tasks for the operating mine company RAG Aktiengesellschaft, Essen/Germany, and associated foundation RAG-Stiftung, Essen/Germany, since hard coal mining in the Ruhr District ceased at the end of 2018. The scientific support of the mine water rebound process includes a robust risk management and associated hydrogeochemical monitoring to verify containment and conformance of mine water during the entire rebound process. Hydrogeochemical monitoring of mine waters becomes increasingly important in the view of a limited access to the mine workings. The aim is to design a monitoring guideline which is fit for purpose, i. e. sustainable, ecological and economically feasible, and supports the optimization of the rebound process, i. e. identifying a mine water level, which is self-regulating. Here, the authors report on a hydrogeochemical data repository on mine water and formation waters of the overburden. The goal is to identify natural tracers which qualify as toolset to prevent and mitigate loss of containment, i. e. detect infiltrating mine water into surrounding aquifers. Extensive studies during the last 60 years have investigated the hydrochemistry of mine waters in the Ruhr District. However, a tracer concept for Upper Carboniferous hosted mine water and the understanding of the entire hydrogeological system in the Ruhr area remains largely a gap of knowledge. The integration of bromide vs. chloride concentration, molar Na/Cl ratios and water isotopes, δ²H and δ¹⁸O, was able to distinguish saline Upper Carboniferous mine waters from the regional saline Cenomanian-Turonian (CT) aquifer and determine the origin of salinity. Further, bromide to chloride ratios was evaluated to trace formation water of the Emscher Formation, which resembles mine waters in terms of elevated sulfate and iron concentrations. Hence, first assessment of trace elements within mine and formation waters resulted in candidates for mine water characterization and identifying its solute candidates with formation waters during rebound.

Introduction

Mine water rebound in the former Upper Carboniferous hard coal mining region of the Ruhr Valley is a long-term billion Euro project operated by the former coal mining company RAG Aktiengesellschaft, Essen/Germany. Financing of so-called perpetual tasks to account for mine water management purposes, i. e. mainly pumping costs, polder measures and water treatment/remediation purposes, is secured by funds of the foundation RAG Stiftung, Essen/Germany (Figure 1).

German industrial hard coal mining districts comprise the Saar and Ruhr Districts including the geologically isolated Ibbenbüren colliery. The research presented here focuses on the Ruhr District which is the most extensively exploited mining area situated in urban areas with approximately five million inhabitants which are impacted by mining and the rebound process. The approximately 150-year history of industrial hard coal mining in North-Rhine Westphalia ceased at the end of 2018 due to the end of coal subsidies. Previously, the coal mining at the Saar region ended in 2012 due to unforseen induced seismicity events. Therefore, German hard coal mining is officially on transition to the post mining stage since 2019 (1). Phasing out coal until 2038 the latest is part of a long-term strategy to fight global climate change and hence, research associated with the long-term projects of mine water rebound in German hard coal mining areas can act as a role model for such necessary task in the coming years. Guidelines obtained from the associated research can be transferred to rebound processes on various mine sites elsewhere in the future. Leveraging experiences from European mine water management and rebound processes at different stages in Poland, the United Kingdom (UK), France and Spain including its mine water inventory (2) has been taken into account in order to apply the lessons learned to the German hard coal mining districts (3). One major aspect for the rebound strategy of RAG is that the mine water rebound is operated in a controlled manner compared to former rather uncontrolled natural mine water rise in the UK or France.

Mine water quality

Gombert et al. (2) summarizes average, minimum and maximum concentrations of aqueous species for European hard coal regions and identified sulfate and iron as key contaminants besides minor impact by As, Cr and Zn. Chloride (Cl) has only been reported with high concentrations in mine waters of Germany, Poland and the UK. However, metal mobilization in non-flooded, unsaturated zones need to be taken into account for accurate forecasting of future mine water discharges. Such initial outflow of mine water after initially infiltrating the unsaturated zone including uptake of soluble species has been termed “first flush” effect (4). The effect is proposed to decay exponentially after the initial breakthrough during mine discharge. Hence, continuous monitoring of mine water hydrochemistry, before, during and after the rebound process and in combination with a baseline concept are essential in order to verify such effects and forecast their imprint on the overall mine discharge rate and quality in dependency of a defined mine water level (5). Especially, the imprint of in-situ sulfide oxidation including a more accurate in-situ sampling of mine waters in underground shafts in dependency of a defined mine water level are key aspects to optimize the forecast of hydrochemistry of mine waters during rebound. Mine waters associated with hard coal mining generally are rich in sulfate and iron species as products from sulfide oxidation processes, called acid rock drainage (6). Often deeply seated mine waters exhibit high electrical conductivity (EC) exceeding 100 mS/cm or are even saline brines, rich in chloride, due to prolonged water-rock interaction with its host rock or dissolution of adjacent evaporites.

Numerous studies including the recent reviews on European hard coal mine waters by Gomberti et al. (2) and a more global perspective of mine waters by Wolkersdorfer und Mugova (7) reviewed the influence of mine water management and discharges to ground and surface waters. Depending on the extent of sulfide oxidation and buffering capacity by mainly dissolved bicarbonate in mine waters, Wolkersdorfer and Mugova (7) reported discrete pH values from mine workings ranging from pH 3.0 to pH 7.5 with respect to hard coal and lignite mining examples. Gomberti et al. (2) reported on pH values from 6.0 and higher with an average on pH 6.5 based on approximately 2,500 samples from European hard coal mine discharges. pH values are important in terms of transition metal mobilization potential. A potential contamination risk exists at acidic pH levels <5.6, because toxic transition metals, i. e. Pb, Zn, Al and Cd, cations increase in solubility. Such parameters mentioned above were identified to be the main characteristics of hard coal associated mine waters.

The proposed mine water concept of RAG in the Ruhr District area foresees a controlled, stepwise rise of current mine water levels to an initial set depth of 600 m below sea level (mbsl) (8). The reason is to keep mine water provinces still hydraulically disconnected from each other and to obey the major prerequisite to protect drinking water reserves, i. e. Upper Cretaceous Haltern Formation sandstones, and economic usage of deeper aquifers.

The thirteen mine dewatering stations currently installed will be transformed to six main operating dewatering stations including backup stations, which will discharge approximately 70 Mm³/a of mine water into the receiving water courses of Ruhr, Lippe and Rhine each year, keeping the river Emscher free of mine water drainage in the future (Figure 2). Therefore, the future dewatering of the former central mining area makes use of higher dilution of the drainage by the river Rhine compared to more local receiving water courses of Ruhr, Lippe and Emscher.

Mine water levels

The mine water rebound is currently set to a final 150 m vertical distance from the potentiometric surface of the mine water towards any drinking reservoir in the region. However, a sustainable, economic and ecological feasible mine water rebound level has not been determined accurately. This task is part of the ongoing research activities and aims of the Forum Bergbau und Wasser (FBW), a foundation consisting of experts in hydrogeology. The foundation assists the mine water rebound process in all former hard coal mining areas of the RAG since 2017. The research presented here has been conducted as part of an internal research project of the FBW.

The current situation of mine water levels in the former Ruhr District mining concession area, which is approximately 4,400 km² large, is very heterogeneous as mentioned previously. Eastern mine water provinces of collieries Königsborn and Westfalen, which are geologically and hydraulically isolated by large sealing faults and not part of the central water provinces, exhibit mine water levels close to the ground surface with hydraulic potentials already equilibrated. Instead, central or northern mine water provinces of the most recently closed collieries exhibit current mine water levels situated as deep as 1,000 mbsl (Figure 3).

Depending on permits for the water provinces, such mine water levels are maintained by pumping of large submersible motor pumps located at the dewatering stations. At currently three collieries in the Ruhr District, i. e. Heinrich as dewatering station at the river Ruhr, Auguste Victoria (closure end of 2015) and Prosper Haniel (closure end of 2018), multi-sensor mine water probes designed by the Research Center of Post Mining (FZN) of the TH Georg Agricola University (THGA), Bochum/Germany, have been installed. Such monitoring stations are part of discrete monitoring to receive hydraulic heads, hydrochemical data (pH, EC, ORP) as well as dissolved mine gas data from deeper levels during the rebound process (9, 10).

Risk management

In order to manage risk of contamination to the hydrosphere and accurately forecast mine water rebound water levels, conformance and containment of mine water needs to be frequently monitored during the active mine water rebound phase. In 2019 legal authorities including the Mining Authority and the Department of Mining and Energy of the District Council of Arnsberg in North-Rhine Westphalia have initiated an integrative monitoring concept including three major topics, which are identified to be relevant during mine water rebound: 1. water, 2. mine degassing and 3. ground movement (inundation/upheaval). Induced seismicity during mine water rebound as a potential consequence is included in the issues of the integrative monitoring concept and currently researched on by the EU funded project FloodRisk. The research presented here is focusing on hydrodynamics and hydrogeochemistry being complementary to the “water” topic of the integrated monitoring initiative.

As part of the activities within the foundation FBW, a risk based framework for measurement, monitoring and verification (MMV) for an effective mine water monitoring plan with emphasis on hydrogeochemistry of mine water has been set up (11). The bow-tie method has been selected by the FBW as method to identify monitoring tasks (Figure 4).

It relates back to the work of Reason (12) on human error investigation and marks a visual representation by highlighting the “unwanted event”. It has been used in the past to enhance process safety of long-term projects. The method builds a solid foundation on risk management by identifying and listing safeguards, which act in terms of prevention or mitigation of consequences of a so-called top event. The more safeguards for specific threats are applicable, the better the MMV can act in case of an unwanted accident. In total five elements are described in the bow-tie scheme with the top event as central topic of the tie, threats and consequences together with the associated safeguards on each side of it. In relation to mine water rebound processes, Heitfeld et al. (13) have used such an approach in the past together with the local authorities in the Netherlands to manage risks related to upheaval threats during rebound in the Limburg mining area. So far, a risk management analysis is an iterative process and needs to be adjusted each time, an “event” occurred. The tracer tool set act as preventive safeguard in a sense that it reveals incremental changes in the hydrochemistry of the sampled water, which gives an alert/alarm when the baseline threshold value is exceeded. It can also act as corrective safeguard in its role as “geochemical eye” to reduce the impact of a top event, if the spatial sample density is appropriate.

This paper describes the hydrogechemical monitoring as one important pillar of an integrated MMV plan. The scientific approach consists of quality-controlled literature data combined with a continuous record of own hydrogeochemical data acquisition since 2016 and comprises more than 750 individual data points. Sources of literature data presented here are from Wedewardt (14), Michel et al. (15) and Puchelt (16). A more focused and detailed origin and tracking of the various water bodies within the entire mine workings and the overburden is necessary due to the limited underground access. Former shafts were plugged and sealed off as part of the regulations in the operational plan for mine closure. Mine water rebound hydraulics are closely monitored and past rebound paths as well as recent forecasts are closely linked with the MMV. During mine water rebound entering the overburden host rock, interaction with regional aquifers and formation waters of overlying Upper Cretaceous sediments should be kept minor. Mine waters can be masked or mixed during rebound. The worst assumptions would be to encounter false positive “alarms” during monitoring from waters, which closely resemble mine waters but are true formation waters with comparable dissolved constituents.

General mine water hydrochemistry

Upper Carboniferous hosted mine waters of the former Ruhr mining district, hence groundwater, which came into contact with underground mine workings, are characterized by high salinity and high levels of sulfate and iron species resulting from (di)sulfide oxidation processes within the mine. One of the major tasks in characterizing deep, saline fluid inventories is to identify the origin of salinity. Brine-type, Na-Cl dominated basinal waters can have different origins (17). Two major processes leading to brines are

(a) evaporative pathways of formerly trapped or infiltrated seawater; and
(b) subsurface dissolution of salt minerals, preferentially halite.

Initial seawater composition, i. e. sensu strictu connate waters, can be subsequently modified after entrapment by diagenesis, water-rock interaction, microbial interactions and mixing with groundwater in the subsurface. The approach reports on first data in order to distinguish fluids affected by such interactions. In terms of higher chlorine contents, Zechstein based brines and mixing of Upper Cretaceous brine-type waters infiltrating Upper Carboniferous hosted mine waters in the past need to be identified and tracked. The Upper Cretaceous Emscher formation forms a major hydraulic barrier in the region. The formation hosts organic-rich marly claystones rich in sulfates and iron minerals originated from sulfide oxidation processes similar to the ones associated with the coal measures. The tracer toolset aims to still link clearly elevated concentrations of dissolved sulfate and iron to either mine water or distinguish a mixing of formation water of Emscher Fm. with mine water. We report on a first example in distinguishing the regional saline groundwater aquifer of Upper Cretaceous Cenomanian to Turonian age superimposed on the hard coal-bearing Upper Carboniferous host rocks from the mine waters using bromide to chloride ratios, molar ratios of sodium versus chloride and water isotopes, i. e. δ²H and δ¹⁸O.

Regional geology of the Ruhr District

The former coalfield of the Ruhr Valley is situated in folded and weakly metamorphosed Paleozoic basement rock of Upper Carboniferous age referred to as the Rhenish Massif (18). Upper Carboniferous siliciclastics intercalated with coal measures (Namurian to Westphalian A-C) forms the host rock of the former mining area. The folded Upper Carboniferous strata is gently dipping towards the North. It is unconformably overlain by Mesozoic sediments of the Muensterland Basin, which are dominated by Upper Cretaceous marls and carbonates (19) (Figure 5).

These two geological provinces contain both Na-Cl type basinal groundwaters with electrical conductivities up to 200.000 µS-/-cm. In the Muensterland Basin, the brine infiltrates Cenomanian-Turonian (CT) strata and builds a regional aquifer extending from the northern edge of the basin, towards the southern border to the Paleozoic basement known as the Haarstrang ridge (21). Numerous spas and former saltworks are located in the region and making use of this saline aquifer since mediaeval times. Up to 1,500 m thick, marly sediments of the Emscher Formation, Middle Coniacian to Upper Santonian, were deposited on top of the CT carbonate sequence. The Emscher Formation consists of homogeneous grey to dark grey coloured sediments of finely laminated to thick layered and in parts organic rich (Type II) marls. Hydrogeologically, the Emscher Formation is a thick aquitard whereas its uppermost part includes a local fractured aquifer in the Ruhr District. Enclosed by aquitards, the CT aquifer is confined. Evaporite deposits in the area are represented by Permian (Zechstein) and Triassic (Upper Bunter, Roet Fm.) salt sequences. They were deposited on the northern rim of the Muensterland Basin and along the western margin down to the Lower Rhine region (Figure 6).

The Upper Santonian Haltern Formation (Haltern Sands) including the Recklinghausen Formation overlying the Emscher Formation represents an important drinking water reservoir for the region. It is anticipated that during the main rebound phase mine water will infiltrate the overburden and potentially mix with CT saline aquifer, and subsequently, will rise within the pore network and fault pathways of the Emscher Formation. As prerequisite, Haltern Formation strata as drinking water reservoir must be kept free from mine waters.

Mine water hydrochemistry study

In total over 750 separate data points have been collected comprising own hydrogeochemical analysis during the period of 2017 to 2021 and literature data. Upper Carboniferous hosted mine waters, CT aquifer samples and Coniacian-Santonian Emscher Formation waters have been focused on in the hydrochemical data assessment (22). Data points cover the entire mining area of the Ruhr district and extends to the Muensterland basin where spas have been sampled (Figure 7).

Hydrogeochemical analyses of major cations and anions incl. trace elements were conducted by certified lab analysis according to DIN ISO 11885 and DIN 10304-1 protocols. Oxygen and hydrogen stable isotope measurements were analysed by an optical CRDS analyser, Picarro 2130i, with standard deviations of 0,06 ‰ for 18O and 0.48 ‰ for δ²H.

Mine waters at the points of discharge are characterized by iron hydroxides and oxyhydroxides precipitates as shown at the discharge from the mine dewatering station Carolinenglück or in the case of the discharge at the mine dewatering station of Robert Müser have been treated with peroxides to form natural elemental sulphur giving the water a milky appearance (Figure 8).

Stable water isotopes, i. e. δ²H and δ¹⁸O, distinguish CT saline aquifer waters from Upper Caroni-ferous mine waters (Figure 9).

The trend shown here for the CT aquifer reveals former meteoric water percolating through the overburden rocks as the origin of water for such aquifer. Instead, Upper Carboniferous mine waters follow that trend partially. Samples from deeper intervals in the mine workings exhibit a trend towards more positive δ¹⁸O values being the consequence of intense water-rock interaction with the host rock.

Origin of salinity has been deduced from plotting molar concentrations of sodium and chloride species (Figure 10). A trend of the ratio of both species reveals a slope of 0.86 as trend of evaporating seawater versus a slope of 1 for halite dissolution in case of the Upper Creteacous CT samples.

Finally, bromide to chloride concentrations are plotted in a log-log plot in order to test its ability to act as natural tracer for mine waters. In figure 11 mine waters can be characterized by chloride to bromide ratios of 1,450 < x < 450. This intermediate range of ratios is flanked by lower values represented by formation waters of Emscher Fm. and higher values by the CT brines of the regional aquifer.

First data assessment reveals the usefulness of detailed screening of hydrochemistry. Trace elements like bromide have been identified to contribute significantly to a robust monitoring plan and risk management purposes. When integrated with more natural hydrochemical tracers, the mine water tracking can potentially be more precise giving more insights on representative hydrochemistry for water provinces in the Ruhr District. Hence, it allows identifying mixing with adjacent water provinces to enhance mine water management. In the future, the aim is to enlarge the hydrochemical analyses and to identify more tracers suitable for risk management evaluation.