An Integrated Strategy to Treat and Control Acid Mine Drainage from Waste Rock and Underground Workings at the Former Franklin Mine in Nova Scotia, Canada: Field Performance Monitoring

Abstract

Acid mine drainage (AMD), which is primarily caused by the exposure of sulfidic minerals to oxygen and water during mining operations, remains a significant contributor to environmental pollution. Numerous technologies have been developed to prevent/control and treat AMD, including the isolation of waste from the atmosphere and treatment systems for AMD-impacted water. Many field studies on mine site reclamation have involved an individual AMD source and/or technology, with a limited number of studies looking at reclamation programs integrating multiple approaches to manage AMD stemming from both surface and underground sources. The former Franklin mine site in Nova Scotia, Canada, was impacted by the deposition of waste rock across the site and the discharge of mine water from underground workings, with the adjacent Sullivan’s Pond serving as the main environmental receptor. Site reclamation was completed in 2010 and involved the following: (1) excavation of the dispersed waste rock (117,000 m²) and backfilling with clean soil; (2) consolidation of the excavated waste rock into a covered, compact waste rock pile (WRP) (25,000 m²); and (3) construction of a passive treatment system for the discharging underground mine water. An extensive field sampling program was conducted between 2011 and 2018 to monitor a range of meteorological, cover material, waste rock, groundwater, and surface water quality parameters. The results confirm that the multi-layer, geomembrane-lined WRP cover system is an extremely effective barrier to air and water influx, thereby minimizing the rate of AMD generation and seepage into groundwater and eliminating all contaminated surface water runoff. A small AMD groundwater plume emanates from the base of the WRP, with 50% captured by the underground mine workings over the long term and 50% slowly migrating towards Sullivan’s Pond. Excavation of the former waste disposal area eliminated the AMD source from the previously dispersed waste, with only clean surface water runoff and a diminishing legacy groundwater plume remaining. Finally, the passive treatment system, which contains a series of treatment technologies such as a limestone leach bed and settling pond, successfully treats all mine water loading (~50 kg/day) discharging from the underground workings and surface runoff. Its additional treatment capacity (up to ~150 kg/day) ensures it will be able to manage any potential drop in treatment efficiency and/or increased AMD loading from long-term WRP seepage. This comprehensive study of mine site reclamation and AMD management at an abandoned mining site can be of great reference value for environmental management and policymakers in the mining sector.

1. Introduction

Mining is one of the oldest and largest industries worldwide and to this day remains an important economic sector and job creator in many countries, including Canada, Australia, and the United States. For example, mining has been conducted on an industrial scale in Canada since the 1700s and continues to be a critically important part of its economy, contributing 5% to the total gross domestic product in 2021 [1]. Despite the significant contributions provided to those countries that once were, or still are, extracting valuable minerals such as coal and metals, mining operations can cause significant damage to the environment via toxic acid mine drainage (AMD) [2].

AMD is formed when sulfidic minerals, such as pyrite, galena, and pyrrhotite are exposed to the atmosphere during mining activities. The interaction of these minerals with oxygen and water results in a range of complex chemical reactions and transformations that produce sulfuric acid [3,4,5]. As this acid moves through the rock, it leaches metals and other compounds, resulting in a highly toxic leachate entering the environment [6,7]. AMD can significantly reduce the pH and enhance the contents of heavy metals (such as iron, lead, and zinc) in nearby soils and water [8,9]. AMD can infiltrate aquifers and generate effluents that converge into surface water bodies and pollute freshwater sources where it has a detrimental effect on aquatic life and ecosystems [10,11,12]. Further, AMD poses a significant risk to plant growth and human health owing to the release of heavy metals from AMD-impacted soils [13,14]. AMD is a serious environmental problem encountered by mining and mineral processing industries worldwide and poses ongoing challenges for environmental remediation and sustainable resource extraction [15,16,17].

Some of the most common sources for AMD generation and release at former mining sites include abandoned underground workings and disposed mine waste rock. After mining operations cease, the water being pumped out of the underground workings is stopped and the rebounding water is in prolonged contact with the exposed rock lining the workings, generating AMD that is then released to the environment via old shafts and tunnels [18,19,20]. Mining activities also leave behind mineralized waste rock that is typically deposited in large storage piles on the ground surface [21,22]. Oxygen and water infiltration into the waste rock can generate and release AMD to the environment via seepage into the underlying groundwater aquifer and runoff into surface water bodies [14,23].

The environmental and economic impacts associated with AMD have motivated the strategic development of AMD prevention strategies, which preclude the formation of AMD at the source [24], and remediation solutions, which treat AMD-impacted soil and waters [25]. Detailed descriptions of various AMD prevention and remediation approaches are provided by Johnson and Hallberg [26], Akcil and Koldas [27], and Kefeni et al. [28]. The most common AMD preventative strategies include overlaying the waste with engineered cover systems to limit the exposure of sulfidic waste minerals to air and water [29,30] and installing drainage controls to divert water around the waste material [31]. Engineered cover systems have been implemented at waste rock piles (WRPs) in various forms, ranging from a simple soil cover to a complex multi-layer system containing various materials such as geomembrane liners, geotextiles, and earthen material [32]. A single-layer soil cover is most effective in arid and semi-arid environments where it can act as a moisture store-and-release system to minimize water ingress [33]; however, in humid climates, it is ineffective at limiting water ingress and is best used to eliminate AMD-contaminated surface runoff [30]. While more costly and complex to install, multi-layer cover systems provide the flexibility to meet a wide range of site-specific requirements and environmental conditions by integrating differing materials; for example, geomembrane liners provide an effective barrier to air and water, geosynthetics are used to reinforce soils, protect liners, and/or enhance drainage, while earthen materials provide a vegetative canopy [32,34,35].

Treatment methods for AMD use a range of chemical, biological, and mechanical processes for the neutralization, precipitation, and/or filtration of mine water [28,36]. Active systems generally neutralize the mine water with an alkaline reagent such as lime or caustic soda, allowing its metals to precipitate as flocculates that are then collected in ponds [18,37]. These systems are reliable and effective, but their power, maintenance, and cost requirements can make them impractical for remote and abandoned mining sites. In contrast, passive systems do not require continuous operational input or routine maintenance activities. These systems sequentially remove acidity and/or metals via gravity and naturally occurring chemical and biological processes and implement technologies such as open limestone channels, limestone/slag leach beds, anoxic limestone drains, and constructed wetlands [38,39]. Passive systems are popular for treating mine water that is not heavily contaminated as they can be nearly self-sustainable over the long term. Detailed overviews on passive treatment technologies for mine water are provided by Wolkersdorfer [40] and Skousen et al. [41].

While many studies have implemented individual remedial technologies (e.g., engineered cover systems, mine water treatment) at mine sites, very few studies have integrated two or more technologies. Most, if not all, of these studies have combined different treatment technologies, such as sulfate-reducing bacteria followed by ultrafiltration, to enhance the treatment of mine water [38,42,43]. Some mining sites that contain multiple sources of AMD contamination, including deposited mine waste material and overflowing underground workings, require both prevention and treatment technologies for successful mine site closure; however, there are few, if any, studies available that have comprehensively examined the effectiveness of an integrated mine site reclamation program at a former mining site over many years.

The objective of this study was to evaluate the performance of a multi-faceted reclamation program at the Franklin mine site within the Sydney Coalfield in Nova Scotia, Canada. AMD was being generated and released across the site from two sources: (i) mine waste material from two adjacent mines deposited on the ground surface across the site, and (ii) flooded underground workings. Reclamation included the excavation, consolidation, and covering of the waste rock into a WRP, and the construction of a passive treatment system for managing mine water discharging from the workings and surface runoff across the site. Field monitoring was performed between 2011 and 2018 and confirmed the effectiveness of reclamation to provide immediate improvements and long-term stability to environmental receptors, including the quality of water entering and then leaving a large neighboring pond. This study provides a successful real-world example of how a former mining site consisting of multiple AMD sources can be reclaimed using complementary control and treatment measures to effectively walk away from the site.

2. Materials and Methods

2.1. Site Description

2.1.1. Pre-Reclamation

The Franklin site, which is located in Bras d’Or, Nova Scotia, Canada (Figure 1), was one of the many sites left behind by the historical mining activities that occurred within the Sydney Coalfield. The site included two adjacent former mine operations associated with the Edwards coal seam, namely, Franklin mine and No. 5 Prospect mine, which date back as far as 1938 and as recently as 1957. Initial environmental site assessments performed in 2005/2006 characterized the site by the presence of acidic mine discharges, seeps, and drainage.

A large area (117,000 m²) of the site, hereafter referred to as the Franklin Site and Waste Area (FSWA), was covered with disposed coal mine waste rock/fines (orange dashed outline in Figure 1a), with an extensive test pit program indicating a waste thickness ranging from 0.2 to 4.6 m, with an estimated 91,300 m³ of waste rock. Further, an underground access tunnel to the No. 5 Prospect mine workings, which are located 15–30 m below ground surface, was flooded and continually discharging AMD-contaminated water to the surface. This discharging mine water, together with the AMD-contaminated water generated from the exposed waste, flowed into surface drainage channels and eventually into Sullivan’s Pond, which ultimately outflows into the Atlantic Ocean. In 2006, surface water sampling at select locations across the site confirmed the high levels of AMD contamination characterized by low acidity and high concentrations of toxic heavy metals (e.g., iron [Fe²⁺], manganese [Mn²⁺], aluminum [Al³⁺]). The installation and monitoring of groundwater wells confirmed that near-surface groundwater was also affected by the large quantities of waste material spread across the site.

The Franklin site is located within a humid continental climate (Cfb climate in the Köppen classification), experiencing a mean annual precipitation and potential evaporation of 1500 mm and 450 mm, respectively. The site geology consists of stony, sandy silt to silty sand glacial till (1–20 m thickness) over sedimentary bedrock consisting of shale, mudstone, siltstone, and coal seams dating back to the Carboniferous period. The strata have gentle to open folds and generally dip 4–15 degrees northeast towards the Atlantic Ocean [44]. Historical site investigations indicated that the near-surface till layer exhibits a hydraulic conductivity and porosity of 5 × 10⁻⁷ m/s and 0.3, respectively.

2.1.2. Reclamation Efforts

(i).

Consolidated and capped waste rock pile

The Franklin site was remediated as part of the mine site closure and reclamation program implemented by Public Services and Procurement Canada (PSPC) throughout the Sydney Coalfield. All of the mine waste that was distributed across the FSWA, along with waste from three nearby sites (Colonial No. 1, Colonial No. 4, and Atlantic Mines), was consolidated into a single WRP. The WRP was placed at the western margin of the site (black dotted outline in Figure 1a) as it allowed for WRP construction on virgin ground that was not underlain by an AMD groundwater plume and was above the high groundwater level mark that was typical of the rest of the site.

The final WRP contains 187,984 m³ of waste rock placed on top of approximately 3–4 m of native till overlaying bedrock and spread over an area of 25,000 m². The waste rock had a maximum thickness of 13 m to a small plateau on top, with side slopes of 4:1 (see Figure 1b). In 2010, the Franklin WRP was reclaimed with a multi-layer cover system. The cover system implements a geotextile fabric on top of the exposed waste rock, which is then overlain with a 60-mil high-density polyethylene (HDPE) liner. A geocomposite drainage net (‘geonet’) was placed over the HDPE liner and consists of an internal 6.35 mm thick layer of intersecting HDPE resin strands that are heat bonded on both sides with nonwoven polypropylene geotextile fabric to prevent fine soil particles from entering and clogging the geonet. It was manufactured to maintain drainage performance over time, even under high compressive loads such as heavy machinery traffic that traverses the WRP, especially during cover installation. A 0.6 m thick layer of till was placed on top of the geonet and then hydroseeded to provide a sustainable vegetative layer. Site photographs during cover system placement are presented in Figure A1 in Appendix A.

(ii).

Clean-up of Franklin Site and Waste Area

The FSWA was cleared of all waste rock material and then followed by confirmatory sampling to ensure all waste rock contamination was removed. The surface was then overlain with 0.5–1 m of local till material and hydroseeded. Site photographs are presented in Figure A2 (Appendix A) to show the conditions at the site before and after reclamation. This clean-up is expected to eliminate contaminated surface water runoff from this area, along with the infiltration of contaminated water into the underlying groundwater aquifer.

(iii).

Passive treatment system

A passive treatment system was constructed to treat the underground mine water discharging from the No. 5 Prospect water drain tunnel prior to discharge to Sullivan’s Pond (Figure 1c). Passive treatment systems were designed to promote pH changes that allow metal precipitation and collection, thus improving the quality of the water that is discharged [41]. At the Franklin site, the passive treatment system contains the following: (i) limestone bed to neutralize the low pH and precipitate metals in the mine water discharging from the No. 5 Prospect tunnel, (ii) settling pond to remove these metals, (iii) open limestone channel to assist oxidation and precipitation of ferrous iron, and (iv) constructed wetland for polishing via aerobic reactions to improve the quality of the effluent prior to discharge to Sullivan’s Pond.

All runoff and interflow from the WRP are intercepted by a perimeter ditch and flow into a sedimentation pond before being directed to a slag leach bed. Further, even though it is expected to now be uncontaminated, surface water runoff from the remediated FSWA is collected by drainage ditches and diverted to the same slag leach bed (see Figure 1). The slag leach bed adds alkalinity to the runoff waters before they merge with and further alkalize the treated No. 5 Prospect mine water within the settling pond.

2.2. Reclamation Objectives

The main compliance criterion is to demonstrate a net positive change in the site conditions to protect and improve the water quality entering and then leaving Sullivan’s Pond. The main components of interest are as follows: (i) the performance of the WRP cover system, (ii) basal seepage and the ‘new’ AMD groundwater plume stemming from the WRP, (iii) passive treatment of the surface water (e.g., discharging mine water from the No. 5 Prospect tunnel, surface runoff from WRP and FSWA), and (iv) historical AMD contamination associated with pre-closure waste rock deposits across the FSWA. The Canadian Council of Ministers of the Environment (CCME) Water Quality Guidelines for Freshwater Aquatic Life (FAL) and Nova Scotia Environment (NSE) Tier 1 Environmental Quality Standards (EQSs) for Fresh Surface Water are also included when analyzing the field monitoring data.

2.3. Field Monitoring

A field monitoring program was implemented to assess the effectiveness of reclamation activities in making improvements to the pre-remediation site conditions. A range of monitoring instrumentation was installed at the WRP in 2011 to monitor key parameters in the atmosphere, cover system, and waste rock and to allow cover performance to be assessed, while a number of monitoring wells (MWs) and surface water sampling locations were placed at select locations across the site. A summary of the various monitoring instruments is presented in

Table 1. Field monitoring instrumentation at the Franklin site.

Section	Component	Sample Locations	Monitoring Period
Waste rock pile	Cover system   Basal seepage	Weather station, soil monitoring stations, weir 10CM-2G	January 2012–December 2018   January 2012–December 2018
Surface water	No. 5 Prospect tunnel Leach bed WRP sedimentation pond Slag leach pad Limestone channel Sullivan’s Pond inflow Sullivan’s Pond outflow	08SW-13 10SW-07 10SW-01 11SW-01, 11SW-02 11SW-03 08SW-06 08SW-07	January 2011–December 2018 January 2011–April 2014 January 2011–April 2014 January 2011–April 2014 January 2012–December 2018 January 2011–April 2014 January 2011–April 2014
Groundwater	WRP plume   Former FSWA	10MW-01, 10MW-02, 10MW-03, 10MW-04 09MW-03, 08MW-04, 08MW-05, 08MW-07, 08MW-10, 08MW-12	January 2011–April 2014   January 2012–April 2014

.

2.3.1. Water Influx to WRP

The water balance method was used to estimate the net percolation (NP) into the waste rock, which is found as the residual of the following equation:

$$NP=PPT-R-AET-\Delta WS-\Delta SS$$

(1)

where PPT is the precipitation, R is the surface runoff, AET is the actual evapotranspiration, ΔWS is the changes in water storage, and ΔSS is the changes in snow storage.

Meteorological monitoring at the WRP consisted of measuring rainfall, air temperature, relative humidity, wind speed and direction, barometric pressure, net radiation, and snowpack depth. Rainfall was measured with a Hydrological Services Model CS700 tipping bucket gauge (Campbell Scientific, Edmonton, AB, Canada, ±0.2 mm). Air temperature and relative humidity were measured with a Vaisala Model HMP45C probe (Vaisala, Vantaa, Finland, ±1% RH; ±0.5 °C), while a 05106-10 wind monitor (R.M. YOUNG Company, Traverse City, MI, USA) was used to measure wind speed and direction (speed: ±0.3 m/s; direction: ±3°). Net radiation was measured with a Kipp & Zonen NR-LITE2 net radiometer (OTT HydroMet Corp., Delft, The Netherlands), while an R.M. Young model 61302V sensor (R.M. YOUNG Company) was used to measure barometric pressure (±30 Pa). Snow depth was recorded with a SR50AT sonic ranging sensor (Campbell Scientific, ±1 cm). A CSI CR800 datalogger (Campbell Scientific), powered by a solar panel/rechargeable battery source, was used to control the meteorological sensors. Each parameter was automatically recorded every three hours. Total PPT was then calculated using a combination of both rainfall and snow weight equivalent data. ΔSS was based on the changes in snow stored atop the cover.

Four soil monitoring stations (SMSs) equipped with data acquisition systems were installed to continuously measure volumetric water content (VWC), along with matric suction (negative porewater pressure) and soil temperature, at multiple depths within the cover system and shallow waste rock. CSI Model CS616-L time domain reflectometry sensors measured VWC (±0.1%), while CSI Model 229-L thermal conductivity sensors measured temperature and matric suction (±0.5 °C; ±1 kPa). These sensors were installed along single depth profiles at each station, with sensor depths within the growth medium (0.05, 0.1, 0.2, 0.3, 0.4, and 0.5 m), drainage layer (0.6 m), bedding sand (0.7 m), and waste rock (1.8 and 2.6 m). ΔWS was based on changes in VWC within the cover material.

A zero-height V-notch weir structure was installed within the perimeter ditch at the base of the WRP. A CSI SR50AT sonic ranger was used to continuously record the water level (stage) behind the weir plate and was used with the weir equation to determine R. AET was measured from an Eddy Covariance system installed at the site and supplemented with empirical estimates of potential evaporation via the Penman equation [45]. Further details on the water balance generated at the Franklin site are provided by Hersey and Power [34].

2.3.2. Oxygen Influx to WRP

Pore-gas concentrations were measured from sampling ports installed at three depths at each SMS, corresponding to the till growth medium (0.4 m), just below the HDPE (0.7 m), and shallow waste rock (2.4 m). A Nova Model 309FGK portable gas analyzer (±0.1%) was connected to the sampling line for each port to manually measure oxygen (O₂) and carbon dioxide (CO₂) concentrations every month. O₂ flux into the waste rock can occur via (i) molecular diffusion through the geomembrane liner, (ii) advection though the exhaust vent, and (iii) dissolved O₂ in the infiltrating water. Fick’s law [46] was used to calculate the diffusive flux (Q_diff), which is driven by the concentration gradient across the HDPE liner and the diffusion coefficient:

$$Q_{diff}=D_m{\displaystyle \frac{dC}{dx}}{A}_c$$

(2)

where dC is the difference in the O₂ concentration above and below the liner [%], dx is the thickness of the HDPE liner [m], D_m is the diffusion coefficient specific to O₂ and HDPE [m²/s], and A_c is the surface area of the cover system [m²].

Subsurface gas flow through a vadose zone well is driven by the difference in pressure between the atmosphere and internal waste rock, perforated length of the well, and the air permeability of the waste rock. Two Setra Model 264 differential pressure transducers were used at each SMS to measure the difference in pressure between the atmosphere and internal waste rock. The following analytical solution by Rossabi and Falta [47] was used to estimate advective flow (Q_adv) through the exhaust vent on the Franklin WRP:

$$Q_{adv}=2\pi b{\displaystyle \frac{k_r}{{\mu }_g}}\Delta P_z{\displaystyle \frac{2}{\ln \left(2.25\tau \right)}}$$

(3)

where b is the well screen length [m³], k_r is the air permeability [m²], μ_g is the air viscosity [kg/m·s], ΔP_z is the differential pressure between the atmosphere and waste rock [Pa], and τ = k_rP_avgt/ϕS_gμ_gr_w², where S_g is the gas saturation [m³/m³] and r_w is the radius of the vent [m]. The dissolved oxygen flux (Q_diss) was determined from dissolved oxygen concentrations and NP.

The total O₂ flux can be converted to an acidity load on the basis of sulfuric acid (H₂SO₄) equivalent (mol/m²/yr). It is assumed that (i) all sulfide is present as pyrite (FeS₂); (ii) O₂ is available for pyrite oxidation; and (iii) no O₂ is consumed by the oxidation of organic material in the cover system or carbonaceous material in the waste rock, thus creating the maximum amount of acidity per mole of O₂. The overall summary reaction of FeS₂ weathering to form AMD is shown in Equation (4), where the oxidation of 1 mole of FeS₂ generates 2 moles of H₂SO₄ per 3.75 moles of oxygen (O₂) present [48]:

$$Fe{S}_2+7/2H_{2}O+15/4O_2\to Fe{\left(OH\right)}_3+2H_{2}S{O}_4$$

(4)

The most important acid-producing minerals in this respect are iron-containing sulfides, especially FeS₂, while the carbonates, especially calcite (CaCO₃), are the main rapid neutralizers.

2.3.3. Waste Rock Acidity

Acidity in the waste rock exists in two forms: stored acidity or potential acidity. Stored acidity already exists within the pile and is readily available for transport to the receiving environment. Potential acidity requires the oxidation of the sulfide minerals to create additional stored acidity available for transport from the pile. Acid base accounting (ABA) tests, which are well established and widely accepted to characterize waste rock acidity [49], were performed on waste rock samples that were collected during the consolidation of the WRP.

2.3.4. Basal Seepage from WRP

Figure 2 presents a conceptual model of the Franklin mine site (cross-section A-A’ in Figure 1a). Basal seepage into the till layer underlying the waste rock will mix with upgradient groundwater, creating an AMD contaminant plume that will follow the groundwater gradient.

A continuous multi-level tubing well (10CM-2G) was installed at the center of the WRP to a depth of 13.7 m (15.6 m a.s.l. elevation) and screened in the till layer underlying the waste rock material. Waste rock samples were analyzed using standardized laboratory procedures to measure the VWC, porosity, and particle size distribution of the extracted waste rock [50]. 10CM-2G allowed measurements of the water level within/below the WRP and the collection of porewater samples from the base of the WRP for subsequent geochemical analyses. The key AMD indicator parameters are acidity, SO₄²⁻, pH, alkalinity, and the dissolved concentrations of Fe²⁺, Mn²⁺, and Al³⁺.

2.3.5. Groundwater Quality

MWs were placed throughout the site to assess the evolution of groundwater quality underlying the WRP and FSWA, as shown in Figure 1a. Four MWs were installed around the perimeter of the WRP to monitor the flow direction and quality of the groundwater plume emanating from the WRP. Well 10MW-01 was located to the northwest of the WRP, with the other three MWs (10MW-02, 10MW-03, and 10MW-04) located along the eastern side of the WRP. Seven MWs (09MW-03, 08MW-04, 08MW-05, 08MW-07, 08MW-08, 08MW-10, and 08MW-12) were also installed to monitor the historical near-surface groundwater within the FSWA. These MWs can also be used to investigate if the groundwater plume emanating from the WRP is impacting further downgradient locations within the FSWA, which is conceptualized in Figure 2.

Groundwater levels and samples were collected monthly at each MW to assess the groundwater flow regime at the site and the evolution of groundwater quality over time. The groundwater samples were analyzed for the same key AMD indicator parameters as 10CM-2G. Further, any historical data collected prior to reclamation were used to compare the quality of groundwater before and after reclamation. ArcGIS Pro was used to create site contour maps of groundwater elevations and groundwater quality, which were based on key AMD indicators.

2.3.6. Surface Water Quality

A number of surface water sampling locations were used to provide a holistic image of the evolution of surface water quality across the site and determine the effectiveness of the reclamation. The sequence of surface water flow and associated sampling locations is illustrated in the conceptual cross-section in Figure 2. Surface water was sampled at (i) discharge from the No. 5 Prospect, (ii) runoff from the WRP and FSWA, (iii) discharge into Sullivan’s Pond, and (iv) outflow from Sullivan’s Pond (see Figure 1).

The quality and flow rate of mine water discharging from the No. 5 Prospect tunnel were assessed at 08SW-13, with this water quality sampled again at 10SW-07 following the addition of alkalinity to the limestone leach bed to increase the pH. The surface water runoff from the WRP was collected via perimeter ditches and directed to a sedimentation pond to allow the settlement of any metals, with the outflow water quality examined at 10SW-01 (see Figure 2). The surface water runoff from the former FSWA was collected in a drainage ditch, with water quality sampled at 11SW-01. Surface waters from the WRP and FSWA were directed into a slag leach bed, and the effluent was sampled at 11SW-02.

The surface waters from both the limestone leach bed (treated No. 5 Prospect discharge) and the slag bed (treated surface runoff) converge prior to a settling pond to allow metals to settle, before being directed through the open limestone channel, which aims to assist the oxidation and precipitation of ferrous iron. The outflow from this channel is sampled at 11SW-03 before it discharges into a polishing pond and then into Sullivan’s Pond. The surface water quality at the inflow and outflow of Sullivan’s Pond was sampled at 08SW-06 and 08SW-07, respectively.

Surface water samples were collected monthly from January 2011 to December 2018 (see Table 1) and analyzed for key AMD indicators. Any available historical data prior to reclamation was used to compare the pre- and post-reclamation surface water quality.

3. Results and Discussion

3.1. WRP Cover Performance

3.1.1. Water Ingress

Figure 3a presents the cumulative flux of each water balance component for each year between 2012 and 2018. The average annual flux values for PPT (1566 mm), R (1058 mm), AET (462 mm), ΔWS (10 mm), ΔSS (8 mm), and NP (28 mm) correspond to approximately 67.6%, 29.5%, 0.6%, 0.5%, and 1.8% of total PPT, respectively. The annual NP into the waste rock ranges from 24 to 37 mm, with the primary mechanism being flow through defects in the liner [34]. Since NP into exposed waste rock has been found to be as high as 34% PPT [51], it is evident that the cover system is highly effective at limiting water influx to the waste rock (1.7–2.1% PPT).

3.1.2. Oxygen Ingress

Figure 3b presents the pore-gas concentrations measured at three depths at the four SMSs: 0.4 m (cover material), 0.7 m (below HDPE liner), and 2.4 m (shallow waste rock). It is noted that over time, the pore-gas measurements became unreliable and erratic due to possible leakages or blockages in the tubing (e.g., atmospheric conditions within the waste rock or zero flow to the pore-gas analyzer), with the last set of reliable measurements on 15 February 2014. Nevertheless, the measurements that are plotted in Figure 3b are reliable and can still provide a valuable assessment of pore-gas behavior within the cover system and waste rock.

O₂ concentrations within the cover material (blue circles) range from 17–21%, with the corresponding CO₂ concentrations (blue squares) between 0 and 5%. Below the HDPE liner, O₂ concentrations (red circles) are reduced to 2–6%, while the corresponding CO₂ concentrations (red squares) increase to 15–19%. This indicates that O₂ ingress through the cover system is limited throughout the available monitoring period and that the cover acts as an effective barrier. While no reliable data are available after February 2014, it is expected that the cover system continues to act as an effective O₂ barrier, with annual site visits and erosion/vegetation surveys confirming that no alterations (e.g., slope failure, creation of liner defects) occurred.

Using Equations (2) and (3), O₂ flux via diffusion, advection, and dissolution during the monitoring period is estimated to be 3.2 mol/m²/yr, 0.6 mol/m²/yr, and 0.004 mol/m²/yr, respectively, providing an annual total O₂ influx of 3.8 mol/m².

O₂ concentrations (green circles) at 2.4 m depth within the waste rock are higher, ranging from 14 to 21%, with the corresponding CO₂ concentrations being lower, with a range of 0 to 5%. Higher O₂ levels within the waste rock exist from pre-cover conditions, especially since the WRP is ‘new’ and received waste material that was previously distributed across the site and open to the atmosphere. Additional pore-gas concentrations measured at deeper locations within the waste rock (5.2, 6.7, 8.2, 9.7 m) at 10CM-2G are also plotted in Figure 3b, with O₂ concentrations (black +) varying from 0.3 to 19.8% and CO₂ concentrations (red ×) varying from 0.1 and 20.5%. This demonstrates the heterogeneous distribution of O₂ within the waste rock and the complexity of AMD generation and release within WRPs. It also suggests that O₂ is entering/contaminating the sampling ports.

3.1.3. Landform Stability

Site inspections conducted periodically at the WRP, including erosion surveys each spring, vegetation surveys each summer, and snow surveys each winter, confirmed a stable landform and healthy vegetative canopy (see Figure A2 in Appendix A). The cover system and landform are expected to remain geomorphically and geotechnically stable given the relatively small landform surface, divergent surface flow contours, low slope gradient that gradually transitions to the plateau, and mature vegetation. Further, the drainage net is highly effective at limiting the head of water above the HDPE liner and ensuring that the buildup of positive pore-water pressures is negligible.

3.2. Groundwater

3.2.1. Groundwater Flow Regime

Figure 4a presents a contour map of the groundwater elevations measured in January 2012 at each MW screened in the till layer. As shown, the highest groundwater elevations are located at the WRP and gradually decrease in an east-to-northeastern direction towards the lowest elevations at the mouth of Sullivan’s Pond. The higher groundwater elevation at 10CM-2G, which is due to groundwater mounding within the waste rock, means groundwater flows outwards from the WRP, with an estimated 50% (west and north of WRP) directed towards the No. 5 Prospect and 50% (south and east) towards downgradient groundwater. Based on the hydraulic gradients and corresponding soil properties, the groundwater flow rate is estimated at approximately 2 m per year below the WRP to the northeast and slows to a rate of 0.5 m per year beyond the northeast corner of the WRP. The transition point is estimated to be approximately 30–40 m from the eastern margin of the WRP.

Figure 4b,c plots the evolution of the monthly groundwater elevations between January 2012 and April 2014 at MWs within the WRP and FSWA areas, respectively. The fluctuations evident at all MWs indicate that groundwater elevations are directly influenced by PPT/recharge, with elevation increases following storm events (with some lag time) and elevation decreases during drier months. For example, the large monthly rainfall (298 mm) in October 2011 is followed by increasing groundwater elevations at all four MWs surrounding the WRP, albeit with a lag time of approximately two months. Similarly, groundwater elevations at most MWs rise in response to the 291 mm of rainfall in September 2012, particularly those nearest to the WRP (10MW-01 to 10MW-04; 09MW-03; 08MW-07). While these periodic fluctuations occur throughout the monitoring period, the general groundwater flow regime indicated in Figure 4a is maintained.

3.2.2. AMD Seepage from WRP

Basal Seepage Rate

During drilling of well 10CM-2G, the waste rock was observed to be highly saturated. Laboratory analysis of waste rock samples indicated a VWC of 0.31, porosity of 0.33, and waste rock as fine textured with approximately 40% fines (i.e., <0.075 mm) and 45% sand. Following placement of the cover system, the water ‘trapped’ within the waste rock drains downwards over many years. Figure 5a shows the groundwater level measured from 10CM-2G from 2012 to 2018, declining from 21.43 to 18.35 m a.s.l. at an approximate rate of 450 mm/yr. Since the waste rock ceases at the top of the underlying till at 16.5 m a.s.l., the drain-down is expected to be complete (i.e., no more groundwater mounding) by 2022, providing a total time to drain-down of 12 years.

While drain-down is occurring over the short term, the contribution of NP to basal seepage will likely be small in comparison; however, as drain-down is completed and the waste rock reaches a long-term steady state water content, the basal seepage rate is expected to decrease to a rate equivalent to the NP through the cover system (2012–2018 average: 28 mm/yr). It is noted that toe seepage was not observed around the WRP perimeter since the HDPE liner was keyed into the perimeter drainage ditch; however, it is possible that seepage water is discharging into the sedimentation pond adjacent to the WRP.

Basal Seepage Quality

Figure 5b plots the evolution of water quality at the base of the WRP in terms of key AMD indicator parameters, namely, acidity, sulfate (SO₄²⁻), pH, alkalinity, Fe²⁺, Mn²⁺, and Al³⁺. As shown, all parameters fluctuate over time, with no discernable improvement or decline in water quality. This is to be expected as the waste rock has undergone considerable weathering and oxidation prior to cover installation, inferring high AMD generation, with the drain-down transporting this AMD porewater to the underlying groundwater aquifer. The concentrations for each parameter are lower during the wet winter/spring months when PPT is higher (e.g., March–April in 2014). The higher groundwater recharge, levels, and flow during these months result in elevated mixing and dilution of the AMD-impacted basal seepage by upgradient groundwater, thereby reducing AMD parameter concentrations.

It is noted that the water sampled at 10CM-2G is collected immediately below the base of the waste rock and may not yet have mixed with upgradient groundwater and/or been neutralized by natural alkalinity. This means that the water quality shown in Figure 5b represents the most conservative scenario of AMD basal seepage into groundwater. Further, the migrating plume will evolve because of various complex processes, including dispersion, aquifer recharge, retardation, neutralization, and redox reactions, which will likely reduce its impacts on the receiving environment.

Acidity Loading from WRP

The long-term seepage rate from the base of the WRP is 700 m³/yr (28 mm/yr over the WRP footprint). Assuming the acidity concentration in the seepage remains at the average concentration observed during the monitoring period (i.e., 1530 mg/L), the long-term acidity loading from the WRP is estimated to be 1.1 t/yr, equivalent to ~3 kg/d. This is a conservative estimate as it is anticipated that reducing bacteria will contribute to alkalinity, and it is known that processes including dispersion, aquifer recharge, retardation, neutralization, and redox reactions would contribute to improved water quality in the plume. Note that if the waste rock was not covered, NP into the waste rock would be upwards of 500 mm/yr, which would provide a loading of ~90 t/yr (245 kg/d).

The waste rock characteristics determined from the ABA tests are summarized in

Table A1. Summary of acid base accounting results for all waste rock.

Waste Rock Source		Franklin/ No. 5 Prospect		Colonial (Colonial 1 + 4)		Atlantic		Franklin WRP Weighted Mean
Proportion of Waste on WRP (%)		69		27		4		100
WR Parameters	Units	Mean	Median	Mean	Median	Mean	Median	Mean
Field paste pH	-	2.57	2.88	3.41	4.07	2.40	2.46	2.67
Field EC	μS/cm	1028	619	305	181	2540	2575	897
Lab paste pH	-	3.43	3.41	3.49	3.62	3.29	3.45	3.44
ANP	kg CaCO3/t	3.10	−1.30	4.13	−0.85	−5.40	−6.10	3.02
AGP	kg CaCO3/t	21.50	12.00	8.15	8.30	11.10	10.60	17.48
NNP	kg CaCO3/t	−18.40	−13.30	−4.02	−9.15	−16.50	15.20	−14.46
ANP/AGP	ratio	0.14	−0.11	0.51	−0.10	−0.49	−0.58	0.17
Total Sulfur	%	1.45	1.17	0.93	0.86	1.34	1.25	1.31
Sulfide-sulfur	%	0.84	0.48	0.33	0.36	0.36	0.34	0.68
Sulfide-sulfur as a % of Total Sulfur	%	53.00	34.00	28.00	34.00	27.00	27.00	45.00
Sulfate-sulfur	%	0.67	0.43	0.38	0.30	0.78	0.80	0.60
Stored Acidity		20.94	13.44	11.88	9.38	24.38	25.00	18.63
Carbon	%	24.00	27.00	26.58	28.50	12.80	12.30	24.23
Carbon trioxide	%	2.20	0.15	6.38	0.26	0.10	0.10	3.23
Carbonate ANP	kg CaCO3/t	36.60	2.50	106.32	4.35	1.50	1.40	53.85

WR = waste rock; EC = electrical conductivity; ANP = acid neutralization potential; AGP = acid generation potential; NNP = net neutralization potential.

in Appendix A. Paste pH for the waste ranges from 2.40 to 3.41, indicating that it is oxidized and contains a reasonable stored acidity component present as water-soluble acid salts. Potential acidity can be represented by the sulfide-sulfur data and the corresponding acid generation potential (17.48 kg CaCO₃/t), while stored acidity can be reflected in the sulfate-sulfur (18.63 kg CaCO₃/t). As there are approximately 319,000 t of waste rock within the WRP, this corresponds to 5576 and 5942 t of potential acidity and stored acidity, respectively.

Using Equation (4), the diffusion, advection, and dissolved oxygen fluxes estimated in Section 3.1.2 are converted to acidity loads of 5.3, 0.99, and 0.01 t/yr, respectively. This means that while 1.1 t of stored acidity will be transported from the waste rock each year, 6.3 t of the 5576 t of potential acidity will be simultaneously oxidized and added to the total stored acidity. Therefore, it will take 800–900 years to convert all potential acidity to stored acidity, followed by another 9000–10,000 for all waste rock acidity to be depleted. Although the alkalinity contained within the waste rock would neutralize some of the acidity, it is ignored to obtain a conservative estimate of acidity depletion.

3.2.3. AMD Groundwater Plume

Figure 6 plots the evolution of groundwater quality, in terms of acidity, SO₄²⁻, and pH, at the Franklin site between 2011 and 2014. The corresponding plots of the Fe²⁺, Mn²⁺, Al³⁺, and alkalinity concentrations are presented in Figure A3 in Appendix A.

Figure 6a–c presents the groundwater quality from the four MWs surrounding the WRP (10MW-01 to 10MW-04) to assess the extent of the AMD plume emanating from the WRP. Assuming the plume began migrating immediately after the construction of the WRP in late 2010, the AMD plume is anticipated to have traveled approximately 7 m by April 2014. Based on the plume’s flow path, it is estimated that the ‘new’ WRP plume, and 50% of acidity loading from the WRP, will migrate northwards and reach the No. 5 Prospect tunnel within ~20 years, while the remaining 50% of the plume will slowly migrate east-to-northeastern through the former FSWA and reach Sullivan’s Pond within ~800 years.

At the start of the monitoring period, the acidity concentrations at 10MW-01 (dark gray squares), 10MW-02 (green triangles), 10MW-03 (red lozenges), and 10MW-04 (purple circles) are 202 mg/L, 492 mg/L, 1 mg/L, and 2 mg/L, respectively. It is suspected that the higher concentrations at 10MW-01 and 10MW-03 are due to their proximity to the No. 5 Prospect mine water tunnel that may be hydraulically exchanging with the surrounding aquifer. Over time, the acidity at 10MW-01 and 10MW-03 declines steadily to 11 mg/L and 30 mg/L, respectively. The hydraulic gradients and flow regime in this area were altered following consolidation of the WRP, and more background groundwater is now flowing towards the tunnel. The acidity at 10MW-02 fluctuates between its lowest values in the winter (~5 mg/L) and its highest values in the summer (~200 mg/L), while 10MW-04 remains relatively constant at ~2 mg/L.

The SO₄²⁻ concentrations at 10MW-01, 10MW-02, and 10MW-03 are similar at ~10 mg/L and remain relatively constant at this concentration over time. In contrast, SO₄²⁻ at 10MW-04 starts higher at 110 mg/L and decreases gradually to 80 mg/L over time. 10MW-04 is further downgradient of the WRP and in proximity to the FSWA, which has historical AMD contamination, and SO₄²⁻ is the most conservative AMD tracer. The pH at all MWs indicates neutral conditions, ranging from 7.5 to 8.5 throughout the monitoring period.

The dissolved Fe²⁺, Mn²⁺, and Al³⁺ concentrations are shown in Figure A3 (Appendix A) with 10MW-01, 10MW-02, and 10MW-03 all steadily decreasing over time, while 10MW-04 remains relatively constant. The alkalinity values at 10MW-01, 10MW-03, and 10MW-04 are similar at ~150 mg/L, while 10MW-02 has an alkalinity of ~100 mg/L, with all MWs remaining relatively constant at these values over time. Overall, the values and evolution confirm that AMD seepage from the WRP, which has much higher acidity (~1000 mg/L) and SO₄²⁻ (~1500 mg/L) during the same period, is not yet impacting these locations and likely still contained within the WRP footprint.

Figure 6d–f shows the groundwater quality from the six MWs downgradient of the WRP within the FSWA. The two wells closest to the WRP—08MW-05 (red circles) and 09MW-03 (blue lozenges)—exhibit relatively low and constant acidity (<10 mg/L), similar to that of the nearby 10MW-04. The other four wells were located more centrally in the FSWA and all started with higher concentrations before declining over time with some seasonal variations. The SO₄²⁻ concentrations also indicate that 08MW-05 and 09MW-03 are relatively low over time, although again, similar to SO₄²⁻ at 10MW-04, the other four MWs exhibit higher concentrations. The pH ranges from 7 to 8.5. Figure A3 indicates that the Fe²⁺, Mn²⁺, and Al³⁺ concentrations within the FSWA are again higher than the metals within the WRP area and show much higher variations over time. The alkalinity ranges from 150 to 350 mg/L, with the upgradient 08MW-04 (light blue ×) and 08MW-05 (red circles) having the highest alkalinity. The higher AMD contamination stems from the ‘stored’ AMD in the groundwater that came from the FSWA operations, with this contamination slowly flowing to new locations with groundwater flow direction, hence why 08MW-07 (green circles) and 08MW-12 (orange squares) are increasing and decreasing over time as they are the farthest downgradient of the flow regime (i.e., mouth of Sullivan’s Pond).

Figure 7 presents site contour maps of the groundwater acidity plume prior to reclamation (October 2008) and the last ‘full’ monitoring date (April 2014). Monitoring data at the five MWs installed in 2008 were used to generate the map, with the other MWs that would be installed within the unused west portion of the site (where the WRP would be deposited) assumed to exhibit background water quality. As shown in Figure 7a, the groundwater acidity plume exists within the FSWA, with the highest concentrations at the center and north of the area via the groundwater flow direction. Figure 7b indicates that in April 2014 (after reclamation), the groundwater acidity plume has dissipated within the FSWA, with the highest acidity concentrations at its center (08MW-07). The largest acidity concentrations now emanate from the WRP as expected, with the plume flowing the general groundwater flow direction towards No. 5 Prospect mine tunnel.

Figure 7 confirms that the excavation of the near-surface waste material was successful in reducing the extent of AMD contamination that was 117,000 m² to a concentrated footprint within the WRP of 25,000 m².

3.2.4. Groundwater Discharge to No. 5 Prospect Tunnel

The No. 5 Prospect water drain tunnel will act as a sink and capture approximately 50% of the acidity load from the WRP (i.e., 1.5 kg/d). In addition to groundwater percolating through the tunnel walls, the underground workings of No. 5 Prospect mine provide water flow to the water drain tunnel, along with the nearby Colonial No. 4 mine, which also operated on the Edwards coal seam and is therefore hydraulically connected.

The remaining plume will flow downgradient in an east-to-northeastern direction and discharge to the polishing pond near the mouth of Sullivan’s Pond, mixing with surface water exiting the passive treatment system.

3.3. Surface Water

All seepage water from the WRP percolates into groundwater via the base of the WRP, so the covered WRP should not be a source for surface water contamination. Similarly, contaminated surface water runoff from the FSWA should be eliminated due to its excavation and covering. Therefore, the main AMD loading to surface water is expected to be discharged by the No. 5 Prospect tunnel.

3.3.1. WRP and FSWA

Figure 8a–c presents the evolution of surface water quality running off from the WRP (10SW-01) and FSWA (11SW-01) and following its initial treatment within the slag bed (11SW-02). The acidity and SO₄²⁻ at 10SW-01 (blue lozenges) fluctuate over time with a gradual decline. It is suspected that a portion of the groundwater AMD is discharging to the sedimentation pond, though it is improving over time as the pond matures.

The surface water quality of runoff water from the FSWA is indicated by 11SW-01 (green triangles), with the low concentrations and improved quality over time indicating that the reclamation activities were successful in excavating the waste material and covering with clean inert material to eliminate contaminated runoff waters. It is evident that the combined waters from 10SW-01 and 11SW-01 that have entered the slag bed have been treated effectively (11SW-02; red circles). The dissolved concentrations of Fe²⁺, Mn²⁺, and Al³⁺ shown in Figure A4 (Appendix A), along with alkalinity, also confirm the relatively good quality of the surface water, with a stable pH between 7 and 8.

3.3.2. No. 5 Prospect Drain Tunnel

Figure 8d–f shows that the quality of the discharge from the No. 5 Prospect tunnel (08SW-13; orange circles) is currently a mixture of seepage water from the underground mine workings and groundwater, which may include basal seepage in the future. The discharge from the No. 5 Prospect tunnel (08SW-13) exhibits high acidity and SO₄²⁻ levels.

Figure A5 plots the discharge from the tunnel along with the corresponding monthly rainfall. The discharge ranges from 0.5 to 33 L/s, with flow tending to increase with rainfall. It does have a longer lag time of 2–3 months compared to the 1-month lag time of the groundwater response in nearby monitoring wells (e.g., 10MW-03; Figure 4). This longer lag time suggests that rainfall recharges the groundwater, which then contributes to the underground workings and increases the tunnel discharge rate. The contribution of flow from the underground workings would represent the ‘baseline’ flow, while groundwater percolating through the tunnel walls in response to rainfall would contribute to fluctuations around the baseline flow observed as the shorter response. This also explains the similarity between the acidity of groundwater at 10MW-03 (Figure 6a; red lozenges) and the tunnel discharge at 08SW-13 (Figure 8d; orange circles), where recharge is flushing sulfate-based minerals from the tunnel.

To assess the influence of increasing flow on acidity concentrations, the acidity load is plotted against the corresponding flow rate in Figure 9a. The linear trend (R² = 0.91) indicates that the concentration is generally constant and that an increase in load is caused by an increase in flow, or greater flushing of stored acidity. Generally, an acidity load of 100–150 kg/day is the maximum recommended load for a passive treatment system [52]. Since the discharge from the No. 5 Prospect water drain tunnel is usually less than 50 kg/day, the passive treatment system at the site is adequate. The limestone leach bed was the first treatment step for the mine water discharging from the No. 5 Prospect tunnel and can consistently lower the acidity and Fe²⁺, Mn²⁺, and Al²⁺ concentrations, though it has little effect on SO₄²⁻. The slag leach bed had little-to-no effect on the quality of surface water runoff from the WRP and FSWA, though it is noted that the quality of the runoff water improved over time regardless.

While AMD seepage from the base of the WRP did not reach the tunnel during the monitoring period, it is anticipated that its effects will be negligible as the mine water within the tunnel is already impacted by significant AMD and the flow is significantly higher than basal seepage.

3.3.3. Discharge to Sullivan’s Pond

Figure 8d–f shows that the surface water quality is much improved by the end of the treatment system (11SW-03; dark gray squares). The outflow water quality confirms the effectiveness of the passive treatment system to treat the discharge from the No. 5 Prospect tunnel, with reductions in both the acidity and SO₄²⁻ concentrations. Further, while the surface water runoff from the WRP and FSWA should remain uncontaminated, the treatment system has the capacity to treat runoff if it became contaminated in the future (e.g., major failure of the cover system leading to runoff over exposed waste rock).

Figure 9b presents the acidity loading at the end of the passive treatment vs. the corresponding flow rate, with the linear trend (R² = 0.88) indicating a relatively constant concentration and that an increase in load is caused by an increase in flow. The outflow loading from the treatment system is generally less than 20 kg/d, which is an acceptable load into a large surface water body such as Sullivan’s Pond. Additionally, when the groundwater plume eventually reaches and possibly discharges into Sullivan’s Pond, its loading is negligible (1.5 kg/d) in comparison to the outflow of the treatment system. The surface water quality at the mouth of Sullivan’s Pond (08SW-06; blue triangles) exhibits similar quality to that at 11SW-03 (dark gray squares), though it is sometimes higher in concentration, which suggests that the legacy groundwater plume is discharging to the polishing pond to match the expected groundwater flow. Nevertheless, the outflow from Sullivan’s Pond (08SW-07; green lozenges) maintains low concentrations to meet aquatic levels.

Figure 10 presents a site map representing surface water quality, in terms of acidity, both before (October 2008) and after reclamation (April 2014). The acidity concentrations are represented as isolated proportional bubbles rather than an interpolated contour map since the SW sampling locations are not interconnected like the groundwater. The red bubbles and label represent pre-reclamation acidity concentrations, while the green bubbles represent post-reclamation conditions. Improved water quality occurs at every sampling location, especially throughout the central area of the FSWA.

4. Conclusions

This study assessed the effectiveness of a multi-faceted reclamation approach to mitigate the impacts of historical AMD at the former Franklin mine site in the Sydney Coalfield in Nova Scotia, Canada. Prior to reclamation, the site was impacted by the surface disposal of exposed waste rock across the Franklin Site and Waste Area (FSWA) (117,000 m²) and mine water discharge from the No. 5 Prospect underground mine workings. The NP through exposed waste rock is 34% PPT, which can mean an NP upwards of 500 mm/yr in this region. Combined with a large surface area, this indicates a potential seepage rate of 58,500 m³/yr into groundwater. The dispersed waste rock was excavated from the FSWA and consolidated into a small WRP footprint (25,000 m²) that was then overlain with a highly effective engineered cover system that greatly reduced NP to only 2% PPT. This significantly reduced waste rock seepage into groundwater to a long-term rate of 700 m³/yr (~98% reduction). This minimized the impacts of AMD on groundwater quality with a groundwater plume now only emanating from the WRP and slowly transporting a small acidity load to the No. 5 Prospect tunnel and Sullivan’s Pond. This minimal seepage rate also means that AMD depletion from the WRP will take thousands of years.

The backfilling of the excavated FSWA with clean soil, along with the covering of the excavated waste rock within the WRP, eliminated all contaminated surface water runoff. The remaining source of AMD to surface water was the discharge of mine water from the No. 5 Prospect tunnel, which provides an acidity load less than 50 kg/d that is then treated by a passive treatment system that was designed to handle up to 150 kg/d. The treatment system contains a series of technologies: limestone leach bed, slag leach bed, settling pond, open limestone channel, and an aerobic wetland. The field monitoring program included sampling locations before and after each technology (e.g., 08SW-13 and 10SW-07 located at either end of the limestone leach bed), thereby allowing them to be evaluated independently and as a collective. Surface water runoff exhibited some AMD impacts in the early months following site reclamation, before improving steadily over time. This suggests that the new WRP sedimentation pond needed time to mature and/or some legacy AMD-impacted groundwater was discharging to the FSWA drainage ditches. Therefore, the mine water discharging from the No. 5 Prospect tunnel was the baseline to evaluate treatment performance.

The limestone/slag leach beds and settling pond significantly lowered acidity (>40 to <5 mg/L) and increased pH (<4 to >7.5), with net alkaline water entering the aerobic wetland, which was necessary to ensure its long-term efficiency. The concentrations of Fe²⁺, Mn²⁺, and Al³⁺ were all above CCME FAL and NS EQS guidelines following discharge from the No. 5 Prospect tunnel but were successfully lowered to acceptable levels. It is noted that while the mine water concentrations were generally constant when discharging from the tunnel, fluctuations were prominent during the treatment process. Finally, the aerobic wetland increased the acidity and metal concentrations, though its performance does gradually improve over time as the wetland matures.

The performance of passive treatment systems over time can diminish due to a number of factors. Examples include reduced efficacy due to seasonal periods of increased drainage and metal concentrations that lead to mineral precipitation, surface passivation, and flow bypass [53]. The limestone can become armored with metal precipitates, which has been found to greatly reduce its effectiveness [54]. The settling pond may be physically filling up as freeboard is lost over time, particularly with net alkaline mine water leading to iron oxide accumulation. This suggests that infrequent maintenance may be necessary, such as periodic (every two to three years) and rehabilitative (once every decade). This passive treatment system does have excess capacity and has the ability to manage fluctuations in performance for some time.

The results presented in this study confirm the effectiveness of the multi-strategy reclamation program that should provide long-term sustainability with little operational input or cost. The excavation of waste rock dispersed across the site and subsequent backfilling with clean soil eliminated contaminated surface water runoff and the AMD source for shallow groundwater. While all the waste rock remained on the site, it was consolidated into a much smaller footprint that was then completely isolated from the atmosphere via the engineered cover system. The minimal seepage from the WRP provides a very low AMD loading to groundwater, with 50% of that being intercepted by the No. 5 Prospect tunnel and then subsequently being treated with the discharging mine water. If no cover was placed over the WRP, then the AMD loading to groundwater would have been above water quality guidelines across the site, and the inadvertent loading to the No. 5 Prospect tunnel would have been problematic and exceeded the treatment capacity of the passive system.

This research provides a comprehensive study of a multi-strategy mine remediation program, using diverse and extensive field monitoring data to demonstrate remedial effectiveness in a real-world application. The findings can help to understand and improve mine site reclamation and AMD management at abandoned former mining sites.