Water quality of the constructed wetland system treating acid mine drainage – Luikonlahti case study
Marja Liisa Räisänen, Geological Survey of Finland, P.O.Box 1237, Finland marja-liisa.raisanen(ai)gtk.fi
The closed Luikonlahti copper mine is located in the Kaavi municipality, 76 km east of Kuopio in eastern Finland. The decision to open the mine was made in 1965, but production did not start until 1968. The mining of ore stopped in 1983. In the autumn of 1979 talc ore processing started as a parallel circuit stage with sulphide ores. Talc ores were transported from Polvijärvi talc mines, first from the Sola talc mine, and later from the Horsmanaho talc-Ni mine. Talc processing continued at the Luikonlahti site until 2007. Since 2012, the mill has been used for processing copper ore transported from the Kylylahti mine, Polvijärvi.
Geology, ore mineralogy and chemistry
Three separate Cu-Co-Zn-Ni sulphide deposits were mined from the Luikonlahti serpentinite massif. The main orebody was the Asuntotalo sulphide deposit, which was located within the SE margin of the massif. The Kunttisuo orebody was within the NW margin of the complex, and the Pajamalmi orebody was overlying the western tip of the Asuntotalo ore.
The deposits occurred within a peanut-shaped massif that is enclosed by a sedimentary package of gneissic greywackes, black schists and sedimentary carbonates. The massif is extensively altered (metamorphosed). Granite dykes cross-cut ultramafic rocks and sulphide ore, locally engulfing blocks of ore as xenoliths. (Peltonen et al. 2008)
The ore deposits contained on average 1.2 wt.% Cu, 0.12 wt.% Co, 0.09 wt.% Ni, 0.65 wt.% Zn and 20 wt.% S (Eskelinen et al. 1983). Ore minerals mostly consisted of pyrrhotite (about 50 wt.%), chalcopyrite (about 3.5 wt.%), sphalerite (about 1.1 wt.%), cobaltian pentlandite (about 0.2 wt.%) and cobaltianpyrite (about 1 wt.%, Eskelinen et al. 1983). The majority of sulphides occurred in sharply-bound semimassive to massive ore lenses (Peltonen et al. 2008). Gangue minerals were quartz (dominant), chlorite, talc, graphite, diopside, tremolite, calcite and magnetite (Eskelinen et al. 1983, Peltonen et al. 2008).
Mining and milling processes
The three separate orebodies (Asuntotalo, Kunttisuo, Pajamalmi) were extracted from two open pits and three underground mines having two shafts down to a depth of 524 m. During 15 years a total of 10.0 Mt of rocks and 6.85 Mt of ore were mined. The Asuntotalo sulphide deposit (main orebody) yielded in total of 6.4 Mt massive-semimassive ore, the Kunttisuo yielded 0.35 Mt, and the Pajamalmi yielded 0.14 Mt of ore.
In 1968, the flotation plant owned by the Myllykoski Oy Company started producing copper concentrate and later in the summer, also pyrrhotite concentrate. In 1969, cobalt-nickel concentrate was produced, and in 1974 production of zinc concentrate started. Talc concentrate production started in 1979. Talc production was continued by Finnminerals Oy in 1984 and by Mondo Minerals Oy from 1998 to 2006. The Mondo Minerals Oy also refined nickel concentrate as a by-product. Talc ore was imported from the Polvijärvi talc mines (Sola, Horsmanaho).
Cu ore transported from the Kylylahti mine, Polvijärvi, has been processed on-site, starting in 2012 by Altona Mining Ltd, by Boliden AB since a change of ownership in 2013. In addition to the Cu concentrate, Ni-Co and pyrite concentrates are separately produced and stored in the lined impoundment north of the tailings facility (Fig.1).
Tailings facility and water management
The Luikonlahti tailings facility is about 27 hectares (Fig. 1). At present, the facility consists of three types of tailings, pyritic tailings (potentially ARD generating) from the previous flotation of the Luikonlahti Cu-Zn-Ni-Co ore in 1968-1983, magnesite rich tailings (alkaline in nature) from the talc flotation (1979-2006), and current tailings from the Kylylahti Cu ore flotation. There are about 6 million tons of the oldest, pyritic tailings and 2.1 million tons of magnesite tailings in the facility. A major portion of the magnesite tailings covers the pyritic tailings. At present, the annual production of tailings is about 0.5 Mt and the estimate for the total tailings production during the next 10 years is about 5.4 Mt (ISAVI 2014).
Originally, the tailings were dumped in the old Petkellampi lake and on the mire surrounding the lake. Due to the topography, steep outcrops and moraine hummocks with crystalline bedrock cores naturally dam the tailings. An initial starter dam was constructed with glacial sandy till and covered with waste rocks on the outside bank. In the west, the main dam (staged conventional dam) is located in a narrow valley bordered by moraine hummocks that embank the tailings north and south of the dam. During primary mining, the initial dam was raised with pyritic tailings with a crest about 20 m above the Suursuo bog ground-level to the west. Shallow dams less than 10 m high in the north and south were constructed from the till and covered by waste rocks on the outside. A similar structure is also used to dam the settling pond in the west. (Räisänen 2003)
In 2012, the magnesite tailings were excavated and utilized to reconstruct and raise the dams in the west, north and south (light grey rim of the tailings impoundment in Fig.1). The new dam structure consists of the primary materials covered by magnesite tailings on the west side, and non-sulphide bearing rocks on the outside bank (Fig.1).
At present, the main chemical treatment of tailings water is in the settling pond located south of the tailings impoundment (Fig.1). The water from the settling pond drains via a canal into the Heinälampi Lake, from which it flows via Kylmäpuro creek into the Luikonlahti lake.
Furthermore, tailings waters are seeping from the toe of the main dam and through moraine hummocks (topmost groundwater) into the Suursuo bog to the west, and in minor amounts, from the northern shallow dam toward the north (Fig.2). In 2007, four wetland ponds were constructed as a part of the tailings facility closing operations to passively treat seepage waters (Figs.1 and 3). At that time, major seepage flowed into surface water from the toe of the main dam (toe seepage) and minor seepage from the interface of the till dam and the bank raised with tailings (upper seepage, Fig. 3). When the tailings pumping was finished, after the cessation of the talc production in 2006, the seepage flow turned toward the toe of the main dam and the upper seepage dried. After 2012 when tailings disposal restarted, the water table within the tailings raised, resulting in seepage flow from upper seepage points in addition to toe seepage flow.
At the toe seepage point waters first streamed into the open limestone drain (Fig. 3b) and then into the wetland ponds (northern pond and then pond 2) at Suursuo. During the decommissioning of the facility, the upper seepage was directed into a limestone drain bounding the road (west of the Suursuo bog) that drained to a minerogenic fen and eventually discharged into the southern wetland pond at Suursuo (Figs. 1 and 3). The water from the Suursuo bog was initially (since 2002) collected in a drain that discharged to a pit west of the mill. The pit water is pumped back into the tailings impoundment (both currently and historically). (Räisänen 2003, Räisänen 2009a, Kauppila & Räisänen 2014).
Figure 1. Location of constructed wetland ponds and water flow directions, west of the Luikonlahti tailings facility, Kaavi, eastern Finland. Blue arrows refer to surface water and white dash arrows refer to tailings seepage within the dam and neighbouring till ground. A thick arrow in south refers to a main outflow from the facility into Kylmäpuro creek and further far to the Luikonlahti lake (not seen in Figure). (Räisänen 2003 and 2009a) Airphotos © National Land Survey of Finland
Figure 2. (a) The seepage area on the slope of the dam before remediation in May 2003 (left) and (b) the construction of a limestone drain in the toe seepage area in May 2008 (right), the Luikonlahti tailings facility, Kaavi, eastern Finland. Photos © M. L. Räisänen, GTK
Major minerals of the pyritic tailings are quartz and iron sulphides. In addition, there are variable amounts of talc, chlorite, calcite, diopside, graphite, sphalerite, chalcopyrite and pentlandite. The magnesite tailings consist of magnesite (Mg carbonate, 80%), talc, serpentine, chlorite, mica, dolomite and minor amount of sulphides. (Räisänen 2003, Heikkinen & Räisänen 2009)
Mineralogical composition of the current tailings is similar to the oldest tailings except it contains less Fe and other trace metal sulphides. The carbonate mineral content is calculated to be about 5% (ISAVI 2014).
During ore refining (both historically and presently) lime is added to elevate the pH during flotation and to precipitate trace metals as hydroxides. This raises the short-term buffering capacity of tailings. The total sulphide sulphur concentration was on average 7.0% for the pyritic tailings, 1.0% for the magnesite tailings and 1.3% for the current tailings. According to Räisänen (2003), the oldest tailings are acid generating whereas the magnesite tailings have excellent buffering capacity and are non acid generating (Räisänen & Juntunen 2004). The NPR ratio of the current tailings was over 6, which indicates a non-acid-generating waste (ISAVI 2014).
Compared to the magnesite and current tailings, the pyritic tailings have markedly higher concentrations of S, Fe, Zn, Cu and Co (Table 1). The elevated concentration of Ni and moderate As were characteristics of the magnesite tailings due to their occurrence in the talc ore deposits (Räisänen 2003). The current tailings had less sulphide trace metals than the old pyritic tailings.
Table 1. Chemical composition of three tailings, the Luikonlahti facility, eastern Finland (Räisänen & Juntunen 2004, ISAVI 2014). Key: – means no data
Design of constructed wetland ponds
The passive treatment system with wetland ponds was constructed for seepage water management as a part of tailings facility decommissioning (Fig. 3, Räisänen & Juntunen 2004). The treatment of leachate was facilitated using a peat-limestone based wetland-type passive system with aerobic and anaerobic reactions (Räisänen 2009a and 2009b). Originally, seepage waters were collected via two open limestone drains (OLD) into wetland ponds. The upper seepage flows via a drain first into a natural wetland (minerogenic fen) and then via a pipe within the road embank into the southern wetland pond (1b, Figs. 1 and 3). The toe seepage flowed via the open limestone drain along the bank of the main dam into the northern wetland pond (pond 1c, Figs. 1 and 2). At present, the upper seepage flows from the south east into a drain bounding a service road and to a pipe constructed through the main dam (Fig. 1). From the pipe the water runs into the OLD collecting toe seepage waters. In the wetland system, the water runs from one pond into another via an open outlet in the middle of the dike, which is constructed of fine-grained till (Räisänen 2009a). The water from wetland pond 2 flows through a bank of limestone aggregate into a ditch and then into a pond from which it is pumped back into the facility (Fig. 1).
The thickness of the topmost peat layer and the mid limestone layer is ca. 30 cm. The thickness of the underlying peat layer is a minimum of 50 cm, increasing to a maximum of 3-4 m in the southern pond. Seepage waters collection drains and the bank of pond 2 (point of final outflow) were filled with limestone aggregate with a diameter of 10-15 cm. The diameter of the aggregate in the interlayer of the pond base is less than 10 cm. Limestone aggregates mainly consist of Proterozoic calcite-rich metasedimentary rocks with minor dolomite. (Räisänen 2009a)
At present, the wetland ponds are growing cattails and reeds in shallow water areas. The northern and southern ponds have open water in the middle of which depth varies between 0.5 and 1.5 m. Pond 1d, which collects waters from the northern and southern wetland ponds, is mostly covered by aquatic plants, whereas the final pond, pond 2 is mostly open water with fewer living plants (some die-back spruce trees, Fig. 2).
Figure. 3. Constructed wetland pond1b (foreground), and pond2 (background) looking north, September 2008, the Luikonlahti constructed wetlands, Kaavi, eastern Finland. Photo © M. L. Räisänen, GTK
Materials and methods
Study materials consist of water samples collected from 12 sites in May 2013 and their physical and chemical analyses (Fig. 4). Monitoring data from May 2003, 2007 and 2008 consisted of water analysis of seepage and outlet waters from 11 sites (excluding drain site in Fig. 4) and pore waters from 4 sites (P1.0, P3.0, P6.0, P8.0) published by Räisänen (2009b, See also Heikkinen et al. 2009). Samples were collected using a limnos sampler in 500 ml polyethylene bottles from outlets of the ponds (P10, P11, P12, P13, P14) and with syringes from a site of Pond1a, seepage sites (UpperS, ToeS1 and 2) and seepage drain sites (Sdrain 1 and 2) due to extreme shallow water (Fig. 4). A water sample from the outflow ditch was taken in a 1-l polyethylene bottle fixed onto a pole in 2013 only (See Drain in Fig. 4).
To examine the chemistry of secondary precipitates formed in the toe seepage area (seepage springs and OLD) and the outlet of the pond, 4 sediment samples were collected with a net sampler. The net sampling method is described by Räisänen et al. (1992). Sampling sites were ToeS1, ToeS2, Sdrain2 and P13 in Fig. 4.
Pre-treatment and analyses of water samples
Water samples for dissolved elements were filtered through 0.45 μm filters [GD/XP syringe filter with polypropylene prefilter and PVDF (polyvinylidene fluoride) -membrane by Whatman] into polyethylene bottles in the field. The filtered samples were immediately acidified with concentrated Suprapur® HNO3 (0.5 ml acid addition per 100 ml sample). In 2003, seepage samples were filtered with single-layer 0.45 µm filters in the laboratory within 1-2 days of collection and then the filtered samples were acidified with concentrated Suprapur® HNO3. The temperature, pH, redox, electrical conductivity, oxygen concentration and saturation were recorded with a multi-parameter field meter (YSI 556 in 2008 and YSI Plus Professional in 2013). In 2003, the pH was measured with a WTW 340i/SET field meter, the conductivity with a WTW cond330/SET field meter and the oxygen concentration and saturation with a WTW oxi330/SET field meter. Redox (ORP) values were read with Ag/AgCl electrodes using YSI and WTW meters.
Concentrations of 30 elements were determined using ICP-AES/MS techniques. Analyses were done in the FINAS-accredited laboratory of Labtium Oy (former Geolaboratory of the Geological Survey of Finland). This study presents data only for Ca, Mg, Na, K, S, Fe, Mn, Al, Zn, Cu, Ni, Co and As.
Pre-treatment and analyses of sediment samples
The pH, redox potential and electrical conductivity were measured with portable (Mettler Toledo pH/Eh, WTWCond) field instruments during the sampling.
For analyses, sediment samples were freeze-dried and sieved to <2 mm grain size. Total concentrations of nitrogen, carbon, sulphur and sulphide sulphur were determined with pyrolytic methods (Räisänen et al. 2010). The total concentration of carbonate carbon was determined by subtracting the residual carbon concentration of the sample treated with hydrogen chloride acid from the total carbon concentration (Räisänen et al. 2010). Acid extractable element concentrations were determined with concentrated nitric acid digestion and ICP-OES methods (Niskavaara 1995). In addition, acid NH4 oxalate and acetate pH 4.5 extraction methods were applied to characterize the chemical composition of secondary Fe precipitates and element concentrations chemically adsorbed by sediment particles (Kumpulainen et al. 2007, See also Heikkinen & Räisänen 2009). The concentration of the water soluble organic carbon (DOC) was determined with the Millipore water extraction method (solid:solution 1:20) by using a C analyzer for dissolved C measurements. Chemical analyses were done in the in the FINAS-accredited laboratory of Labtium Oy. This study presents ICP-OES data for only Ca, Mg, Na, K, S, Fe, Mn, Al, Zn, Cu, Ni, Co and As.
The element distribution in three geochemical fractions (tightly bound, fixed with secondary precipitates, chemical adsorption fraction) was based on the above non-sequential extraction scheme, in which separate subsamples were leached with extractive solutions of increasing effectiveness, assuming that the stronger solutions also dissolve the phases leached with the weaker solutions. Consequently, the amount of each fraction is calculated by subtracting the concentration of the previous (weaker) extraction from the next step (See also Heikkinen & Räisänen 2009).
Figure 4. Location of water sampling sites, the Luikonlahti mine area, eastern Finland. Keys for sampling sites: Pond1a for water of the minerogenic fen, P10, P11, P12, P13 and P 14 for outlet water, P1.0, P3.0. P6.0 and P8.0 for pore water of monitoring tubes, and UpperS (upper seepage), ToeS1 and 2 (toe seepages) for seepage springs and Sdrain 1 and 2 for waters of the seepage collecting drain (OLD), Drain (collecting drain) for the outflow from the wetland ponds (See also the text). Airphoto © National Land Survey of Finland
Results and discussion
Water quality of tailings seepage and wetland ponds
In 2003, effluent from the tailings was acidic (pH<3) with high trace metal (Mn, Zn, Ni, Cu, Co), Fe, Al and S content (Table 2). In addition, the seepage waters had elevated concentrations of alkaline earth (Ca, Mg) and alkaline (Na, K) metals that mainly originated from process chemical (e.g. xanthates, lime) remains in tailings. After the cessation of the talc production, the pH increased to values of near-neutral (pH>6) in 2007 and 2008 but decreased to acidic values (4.5 on average) after emplacement of the current tailings in 2013.
The upper seepage point in 2003 indicated that sulphide mineral oxidation was extensive before the decommissioning of the facility (Räisänen & Juntunen 2004, Heikkinen et al. 2009). After the decommissioning in 2006 and 2007, the greatest change in seepage quality was the increase in pH and the decrease in concentrations of trace metals (Fig. 5). However, concentrations of earth alkaline and alkaline metals, as well as those of S, were less changed. After reconstruction of the facility and refilling, pH of the seepage decreased in acidity leading to an increase in trace metal concentrations (see data in 2013 in Fig. 5). In contrast, concentrations of Ca, Mg, Na and K and Al returned almost to the same levels as they were in 2003. Concentrations of Fe, S and Mn were somewhat lower in 2013 than 2003.
Table 2. Annual variations in pH, redox (mV), electrical conductivity (EC, mS/m), oxygen concentration (mg/l) and saturation (%) of seepage waters in May 2003, 2007, 2008 and 2013, the Luikonlahti tailings facility, Kaavi, eastern Finland (See sampling sites in Fig. 4). – no measurement, n = sample amount
Figure 5. Distributions of soluble trace metals (Ni, Zn, Co, As, Cd) and major elements and Mn (a-b) in the upper seepage waters in May 2003, 2007 and 2013 and (c-d) in the toe seepage waters in 2003, 2008 and 2013, the Luikonlahti tailings facility, Kaavi, eastern Finland (See also text).
After the decommissioning of the facility, the water quality was better in the southern than northern wetland pond and slightly amended downstream in 2008 (Räisänen 2009b). The waters were slightly acidic (pH<6), consisted of some base metals and considerable concentrations of S, Ca, Mg, Na and K (Table 3). The difference between the southern and northern ponds was ten-fold for Fe and Mn, one hundred-fold for Zn, and two- or three-fold for S and Ca. Concentrations of other trace metals (Ni, Cu, Co, As, Cd), Mg, Na and K differed less between the wetland ponds. In contrast, the Al concentration was somewhat greater in the southern than northern pond and increased in the water of the collecting drain downstream.
A total reduction in concentrations of 99.6%, 97%, 96%, 95%, 92%, 88%, 86% and 70% from the upper seepage in 2003 to the southern pond in 2007 was measured for Al, Fe, Mn, Zn, Co, Ni, Cu and S, respectively (Räisänen 2009b). From the toe seepage to the northern pond, Al and Cu concentrations decrease over 99%, while Mn and Zn decreased 87% and 82%, respectively. Reduction of Ni and Co was much lower, at about 60% and 40%, respectively, whereas the decrease in S concentration was surprisingly small (<10%, 2007). In the next year 2008, the reductions were almost the same from the toe seepage to the northern pond (the upper seepage dried). In 2013, concentrations of Fe, Cu, As, Al, Cd, Zn and Co decreased 100%, 95%, 85%, 80%, 75%, 70%, 60%, respectively, from the average seepage to the outflow (Oja-site in Fig. 4). In contrast to the other trace metals, Ni removal was the lowest, at about 40%. The total reduction in concentrations of S from the average seepage to the outflow was almost the same in 2008 (and 2007) and 2013, at 45%.
Räisänen (2009b) concluded that the removal of Fe from seepage was pretty good, despite the fact that Fe bearing solids (e.g. Fe precipitates) were discharged into the collecting drain from the final wetland pond 2. Furthermore, she reported that the performance of the substrate in the ponds, especially in the southern pond, was favourable for sulphate, Fe and other metal reduction. Adsorption by sulphides, as well as to organic particles, explained the high reduction percentages observed for trace metals (Räisänen 2009b). Nevertheless, free waters of the ponds had more S related to chalcophile metals (especially to Fe) resulting in abundant release of soluble sulphur (as sulphate) into the collecting drain. In addition to S, concentrations of Ca, Mg, Na and K were fairly high in the outflow of the wetland ponds in 2008.
After the reconstruction and refilling of the facility, the water quality in the wetland ponds, except for the ponds 1a and 1b, changed from slightly acidic to acidic (Table 3). The pH ranged from 3.1 in pond 1c to 3.5 in pond 2, and was 3.6 in the outflow of the collection drain. The acidic conditions promoted the dissolution of Al, Zn, Ni, Co and Fe, while Mn concentration only increased in pond1c. Furthermore, the increase in redox potentials (except in pond 1a) revealed that ponds’ waters became more oxidative in 2013. Concentrations of S, alkaline earth and alkaline metals, however, stayed relatively constant from 2008 to 2013.
Table 3. Physical and chemical composition of waters in the wetland ponds and the outflow of the collecting drain in May 2008 and 2013, the Luikonlahti constructed wetlands, Kaavi, eastern Finland (See sampling sites in Fig. 4).
The composition of sediments at seepage sites and outlet of the collecting wetland pond
The condition of sedimentation in the seepage area ranged from acidic and oxidative to acidic and less oxidative, while the condition in the seepage collection drain was near-neutral and reducing (Table 4). It is noteworthy that pH field measurements differed from active pH (of CaCl2 extracts) measurements, with the exception of the sediment collected from the toe seepage 2 site. The active pH indicates the impact of ion exchange on acidity. For instance, the pH can be decreased by Fe and Al hydrolysis.
Total carbon concentrations ranged from 1.4-1.7 dry wt.% of seepage sediments, to 2.6 dry wt.% of the drain sediment, and 1.8 dry wt.% of the outlet sediment (Table 4). The carbon was mainly bound by carbonates in the seepage sediments, indicating dissolution of limestone rocks in the seepage area. The sediment of the outlet had abundant carbonate carbon (4.3 dry wt.%). The accumulation of organic carbon was somewhat lower, at 3.5 dry wt.%, which indicates dissolution of organic matter from the peat and the erosion of peat particles from the bottom layers of the wetland. The concentration of water soluble organic carbon (DOC in Table 4) was greatest in the sediment of the outlet drain and lowest in the seepage sediments.
Overall, sediments in the toe seepage area, seepage drain (OLD) and outlet canal (P13 in Fig. 2) of the pond 1d had abundant Fe, ranging from 31.1 to 42.1 dry wt.% (Table 5). From 75 to 95% of the Fe content was oxalate extractable indicating the dominance of secondary Fe oxyhydroxides (Fig. 6). The acetate extractable component of the oxalate extractable concentration for Fe varied between 25 and 30%. The acetate extractable Fe is assumed to be bound by non crystalline Fe hydroxides (Kumpulainen et al. 2007).
The nitric acid extractable Fe that is not oxalate extractable reflects Fe tightly bound by Mg and K bearing silicates (mica, clay minerals), and minor amounts by sulphides (See Fig. 6). Field observations during sampling of the OLD showed some indication of sulphides: The sediment smelled of hydrogen sulphur and was composed of a mixture of black-coloured particles within brownish yellow Fe precipitates. Furthermore, the redox potential of the sediment was negative, supporting the interpretation of Fe sulphide formation below the topmost layer, which was dominated by Fe oxidation and oxygen consumption (Table 4). Low redox potential was also measured from the sediment at the outlet. However, the hydrogen sulphur odour and black coloured particles were not as evident at this site as in the OLD. Because these samples were collected by net, only the topmost (likely oxygenated) sediment was evaluated and changes in sediment chemistry with depth was not assessed.
The total sulphur concentration of the sediments varied between 1.6 dry wt.% and 3.3 dry wt.% being highest at the toe seepage1 site and lowest at the toe seepage2 site (Table 4). From 40 to 75% of the total sulphur was sulphide sulphur (0.8-1.6 dry wt.%, Table 4). Relative to Fe, a high proportion (70 to 90%) of the total sulphur was dissolved in acid oxalate solution (Fig. 6). From 60 to 90% of the oxalate extractable sulphur was, however, dissolved in acid acetate solution. These findings indicate that sulphur is fixed as sulphate with Fe precipitates such as schwertmannite and poorly crystalline goethite (Kumpulainen et al. 2007). Overall, the geochemical fractions of S do not unambiguously reveal actual proportions of sulphide sulphur and sulphate sulphur. Despite the freeze-drying of the samples, part of the sulphides were obviously oxidized to sulphates not only during the sampling but also during the oxalate and acetate extractions. Therefore it is argued that the acetate and oxalate extractable S originated partly from sulphide sulphur and partly from sulphate sulphur.
Magnesium and K were mostly bound by acid extractable silicates such as micas and clay minerals (chlorite, Table 5). Their extractabilities in oxalate and acetate solutions were low (Fig. 6). The exception to this was K, of which 15% was fixed with secondary precipitates and 44% was chemically adsorbed by solids in sediment taken from the toe seepage2 site. Due to acid pH (<4) of the sediment at this site, it can be suggested that K bearing jarosite is a potential source for K.
Table 4. pH, redox potential, electrical conductivity (EC) and total concentrations of carbon, carbonate carbon, organic carbon, sulphur and sulphide sulphur and concentrations of oxalate (Ox) and acetate (Ac) extractable sulphur and concentrations of water soluble organic carbon in the sediments at sites of toe seepages, seepage drain and outlet (P13) of the wetland pond 1d, the Luikonlahti constructed wetlands, Kaavi, eastern Finland (See sampling sites in Fig. 4).
Table 5. Concentrations of acid (HNO3), oxalate (Ox) and acetate (Ac) extractable elements (Fe, Al, Ca, Mg, K, Na, Mn, Zn, Cu, Ni, Co and As) in the sediments at sites of toe seepages, seepage drain and outlet (P13) canal of the wetland pond 1d, the Luikonlahti constructed wetlands, Kaavi, eastern Finland (See sampling sites in Fig. 4).
Compared to K, concentrations of acid extractable Na were low, which was mainly the result of weak dissolution of Na bearing silicates such as plagioclase during the concentrated nitrogen acid extraction (Doležal et al. 1968). Percentage values of the oxalate extractable Na concentrations were low except for the sediment at the toe seepage1 site (Fig. 6). Obviously, the fixation of Na with Fe precipitates indicates the formation of Na jarosite at this seepage site. Percentage values of the acetate extractable Na concentrations ranged from 40 to 85%, being lowest at the toe seepage2 site and highest at the seepage drain site. It is assumed that a great deal of the acetate extractable Na originates from the process chemicals (xanthates, explosive chemicals).
The distribution of nitric acid and acetate extractable Ca is presented in Fig. 7. Concentrations of Ca in the acid oxalate extracts were inaccurate due to calcium precipitation as Ca oxalate salt during the extraction (Räisänen et al. 1992). The behaviour of Ca can, alternatively, be examined on the basis of its acetate extractable concentrations. As seen in Fig. 7, the distribution of the acid and acetate extractable Ca in the sediments had similar trend as that of the acetate extractable Mg. It is assumed that extractability of Ca and Mg in the acid acetate solution depicts their release from chemical weathering of calcite and dolomite from limestone rocks, especially in the seepage drain where the acid effluents are first discharged. The results unambiguously show that Ca (less Mg) accumulates in the sediment of the seepage drain (Fig. 7). Obviously, the acid extractable Ca reflects crystalline gypsum, but the acetate extractable Ca reflects non crystalline gypsum (CaSO4) which precipitates and consequently elevates the pH.
Acid extractable concentrations of chalcophile metals varied from one sampling site to another (Table 5). Zinc was notably accumulated in the sediment at the toe seepage2 and seepage drain sites, whereas Cu accumulated at the toe seepage1 and OLD sites, and Ni and Co accumulated at the OLD and outlet sites. The sediment of the OLD seemed to trap some Mn, whereas the accumulation of Mn at the other sites (toe seepage and outlet) was low (<100 mg/kg). A similar trend was observed for As.
According to the geochemical fractions, from 30 to 70% of chalcophile metals were tightly bound by silicates and sulphides, and oxides in the case of acid extractable Mn. Percentage values of the tightly bound fraction were somewhat higher for sediments of toe seepage1 and OLD than toe seepage2 and outlet (Fig. 6). This was especially true for Mn and Cu. Assumedly Mn oxides are adsorbed on carbonate rocks whereas Cu is mainly bound by organic matter and less by secondary sulphides (Robbins et al. 1999, Champagne et al. 2005). Mean percentage values of the oxalate extractable concentrations (fixed with secondary precipitates) for Zn, Cu, Ni, Co and As varied between 10% and 40%, and those of the acetate extractable concentrations (chemically adsorbed) between 10% and 35%. Zinc and As preferred the fixed fraction more than the chemical adsorption fraction, whereas Ni and Co were variably retained in both fractions. Compared to Cu, the proportion of Ni in the chemical adsorption fraction elevated at the site of the outlet (P13). This reflects the potential remobilization of Ni due to the increase in water acidity at that site in 2013 (cf. Table 3). Nevertheless, the geochemical fractionation of a few sediment samples did not reveal an actual reason for the weaker fixation of Ni with the solids compared to the other trace metals.
Figure 6. Percentage values of the geochemical fractions for the elements (Mg, K, Al, Na, Mn, Fe, S, Zn, Cu, Ni, Co and As), (a-b) the toe seepage area, (c) seepage collection drain (OLD) and (d) outlet (P13) of the pond1d, the Luikonlahti constructed wetlands, Kaavi, eastern Finland (See sampling sites in Fig. 2). Keys for the geochemical fractions: tightly bound = an element bound by acid extractable minerals such as silicates and sulphides (deduction of the oxalate extractable concentration from the acid extractable concentration), fixed with prec. = fixed with precipitates (such as Fe and/or Al oxyhydroxides, deduction of the acetate extractable concentration from the oxalate extractable concentration), Chem.ads. = chemically adsorbed (acetate extractable concentration) including physically adsorbed elements.
Figure 7. Distributions of the concentrations of acid and acetate extractable Ca and acetate extractable Mg in the sediments at sites of the toe seepages (Toe S1, Toe S2), seepage drain (S-drain) and outlet drain (outlet P13) of the pond1d, the Luikonlahti constructed wetlands, Kaavi, eastern Finland (see sampling sites in Fig. 4).
Efficiency of the constructed passive system
Two years after the facility’s closure, thickening of the magnesite-rich tailings had a greater impact on leachate purification than the constructed passive treatment system at Luikonlahti. The change resulted in the quick oxidation of Fe and precipitation of Fe oxyhydroxides in the constructed pond that received the most seepage in autumn 2008. As a result, the removal of Fe was greater than that of S, Al and trace metals (Räisänen 2009b).The net acidic seepage from the oxidized sulphide tailings changed to net alkaline and consequently, concentrations of metals decreased from 40% to 99%, depending on the seepage point. Moreover, the seepage converted from oxidative to reductive and thus maintained Fe dissolution.
After reconstruction and refilling of the tailings facility in 2012, the seepage quality changed to net acid and consequently concentrations of metals increased to almost the same concentrations they were before the cessation of the talc production in 2003. However, the water data from the wetland ponds in 2013 showed that chalcophile metals were trapped by wetland sediments. The total reduction in concentrations of 100%, 95%, 85%, 80%, 75%, 70% and 60% from the seepage to the outflow in 2013 was measured for Fe, Cu, As, Al, Cd, Zn and Co, respectively. In contrast, Ni removed poorly, at only about 40% of the total. The total reduction for S was 40% in 2013, which was the same as in 2008 (and 2007). Similarly, the removal of earth alkaline and alkaline metals stayed poor.
The results of the earlier study showed that pH and redox conditions, especially within the basal structures of the ponds, supported reduction of sulphate and iron resulting in the formation of Fe sulphides (Räisänen 2009b). Furthermore, Räisänen (2009b) reported that it was not reducing reactions alone that caused the excellent removal of Fe. The sediment results of the present study confirmed that aerobic precipitation of Fe and fixation of sulphate and trace metals by Fe oxyhydroxides, in addition to chemical absorption on secondary precipitates and organic matter play an important role in water treatment (Walton-Day 2003, Champagne et al. 2005). The pond waters, especially in pond 1c, 1d and 2 had abundant soluble sulphur (i.e. sulphate) related to soluble Fe and trace metals, which could limit the formation of metal sulphides. Instead, the limited adsorption of sulphates on Fe oxyhydroxides and fixation to form oxyhydroxide sulphates may explain the poor removal of sulphate (Bigham et al. 1990, Kumpulainen et al. 2007). The redox potentials of the pond waters have increased from 2008 to 2013, which obviously restrains sulphate reduction at the interface of the water and sediment (Woulds & Ngwenya 2004).
The water quality of the wetland ponds has changed from slightly acid (pH>5) to acid (pH<4), except in the southern ponds (1a and 1b) which do not receive any acidic seepage at present. It is assumed that the poor neutralization of the seepage and pond waters is caused by slow limestone weathering in the OLDs and in the basal structures (bearing limestone aggregates) of the ponds. Obviously, coatings of Fe and Mn precipitates on limestone rock surfaces retard carbonate dissolution and consequently prevent water neutralization (Fig. 8). In addition, the increase in water acidity has somewhat diminished the retention of trace metals, especially that of Ni.
Figure 8. Secondary Fe precipitates coating limestone rocks in the toe seepage area, the Luikonlahti constructed wetlands, Kaavi, eastern Finland. Photo © M. L. Räisänen, GTK
In 2008, two years after the facility’s closure, the net acidic seepage from the oxidized sulphide tailings changed to net alkaline and consequently, concentrations of metals in solution decreased from 40% to 99%, depending on the seepage point. Moreover, the seepage water altered from oxidative to reductive thus maintaining Fe dissolution. After reconstruction and refilling of the tailings facility in 2012, the seepage water returned to net acid conditions and consequently concentrations of metals increased almost to the same levels as they were before the cessation of the talc production in 2003.
The water quality of the wetland ponds changed from slightly acid (pH>5, 2008) to acid (pH<4, 2013), except in the southern ponds (1a and 1b) which do not receive any acidic seepage at present. Furthermore, water data showed somewhat lower contaminant remediation in the wetland ponds, expect for the southern pond, in 2013 than 2008. In 2013, the total reduction in concentrations of Fe, Cu, As, Al, Cd, Zn and Co from the seepage to the outflow was 100%, 95%, 85%, 80%, 75%, 70% and 60%, respectively, while in 2008 metals reduction ranged from 100 to 85%. In 2013, Ni removed poorly, at about 40% of the total, whereas about 90% of Ni was immobilized in the wetlands in 2008. The total reduction for S from the seepage to the outflow was 45% in 2013, which was the same as it was in 2008. Similarly, the removal of earth alkaline and alkaline metals stayed poor in the both monitoring years.
At present, water remediation in the wetland area is assumed to be mainly driven by aerobic precipitation of Fe and the fixation of sulphate and trace metals by Fe oxyhydroxides and to a lesser extent sulphate and metal reduction (i.e. metal sulphide formation). The limiting factors for sulphide formation are the abundance of soluble sulphur (i.e. sulphate) related to soluble Fe and trace metals, and the acidic and oxidative conditions at the interface of the water and sediment. Furthermore, the neutralization of the acidic water is limited by Fe and Mn precipitate coatings on limestone rock surfaces, which retards carbonate dissolution.
Seepage waters from the Luikonlahti tailings facility are undergoing passive treatment within constructed wetland ponds. The four-segment wetland ponds were originally constructed at the Suursuo bog in the summer, 2007 as a part of the tailings facility closure upon cessation of talc production in 2006. The basal structure of the ponds included a peat layer, limestone aggregates, and a natural peat bed. Seepage waters from the upper and toe seepage points were originally collected into two oxic limestone drains (OLDs), which discharge into wetland ponds from south and north, respectively. At present, the both seepages are discharging into the OLD bounding the main dam of the facility.
Two years after the construction aquatic plants have spread mainly in shallow water areas bounding open water in the middle of the ponds. The water level and pond volumes fluctuate according to annual precipitation and seepage discharge from the impoundment that has been actively used for storage of new tailings since 2012.
The constructed wetlands initially resulted in a fairly good water treatment rate for Fe and trace metals, as measured in 2008. From 80 to 90% of Al and several chalcophile metals were trapped into wetland peat sediments. Even though the retention of Ni and Co was almost 90% in the southern pond in 2008, the retention in the northern pond was poor overall, and was between 40 and 60% in 2008 and 2013, respectively. Similarly, the total reduction of S concentrations in the southern pond was good, at about 80%, but hardly any sulphur (<10%) was trapped in the northern pond in 2008 (and in 2007). In 2013, the total retention of sulphur in the northern pond increased slightly (to about 45 %). The retention of earth alkaline and alkaline metals was similar to the sulphur retention in 2008 and 2013. Nevertheless, the greatest change from 2008 to 2013 was that the water quality of the northern (1c), collecting (1d) and final (2) ponds changed from slightly acid to acid. Furthermore, the pond waters were more oxidative in 2013 than in 2008, which could explain the poor removal of sulphur. These findings are assumed to also explain the deterioration in water quality in 2013. In contrast to ponds 1c, 1d and 2, the southern pond (1b) showed good performance due to the fact that it does not receive acidic seepage at present.
Räisänen (2009b) reported that the main mechanism for water remediation was sulphate and metal reduction followed by the formation of Fe and trace metal sulphides. This was especially true for the southern pond. In the northern pond, chemical reduction was minor and there the main mechanism was chemical adsorption of metals and sulphate on Fe precipitates and organic matter in addition to fixation with Fe oxyhydroxides. The sediment data of the present study confirmed both mechanisms for water treatment. At present, the increase in acidity and redox potential of the seepage and pond waters, except the water of the southern pond, has caused the deterioration of S and trace metal (e.g. Ni) removal.
Overall, it is concluded that the critical factor for the effective formation of Fe and other metal sulphides is the ratio of soluble sulphur to Fe and trace metals. In addition, the pH and redox potential at the interface of the water and sediment control sulphate and metal reduction (Woulds & Ngwenya 2004, Champagne et al. 2005). At present, Fe is first oxidized and immediately precipitated as oxyhydroxides, covering limestone rock surfaces that surround seepage springs and in the seepage collection drain (OLD). This distorts the S to Fe ratio downstream in the ponds. Furthermore, coatings of secondary Fe and Mn precipitates on the rocks retard weathering of carbonates and subsequent neutralization of acidity.
The present study did not give unambiguous answers on how to improve the efficiency of the constructed wetland ponds at Suorsuo. The study was based on earlier monitoring data collected in May 2003 and 2008, and new data collected in 2013. Despite some study limitations (small amount of monitoring data, no data on sediment chemistry within the ponds), the general performance of the wetland ponds was estimated. To provide a better understanding of actual treatment mechanisms, data from the water-sediment interface and pond sediment chemistry is needed. In order to quantify metal and sulphur loads to receiving drains, annual and seasonal water velocity measurements from the seepage areas of both wetland pools and in the receiving ditch is needed. Furthermore, it is important to examine the sources and controlling factors of the current acidification in the upper and toe seepage waters before planning improvements for the passive treatment.
Bigham, J.M., Schwertmann, U., Carlson, L., Murad, E. 1990. A poorly crystallized oxyhydroxysulfate of iron formed by bacterial oxidation of Fe(II) in acid mine waters. Geochim. Cosmochim. Acta 54, 2743–2758.
Champagne, P., Van Geel, P. & Parker, W. 2005. A Bench-scale assessment of a combined passive system to reduce concentrations of metals and sulphate in acid mine drainage. Mine Water and the Environment 24, 124–133.
Doležal, J., Povondra, P., Šulcek, Z. 1968. Decomposition Techniques in Inorganic Analysis. London Iliffe books Ltd, London.
Eskelinen, E. Huopaniemi, P. & Tyni, M. 1983. Myllykoski Oy:n Luikonlahden kuparikaivos 1968-1983. Vuoriteollisuus 41, 94-98.
Heikkinen, P. & Räisänen, M L. 2009. Heavy metal and As fractionation in sulphide mine tailings – indicators of sulphide oxidation in active tailings impoundments. Applied Geochemistry 24, 1224-1237.
Heikkinen P.M., Räisänen M.L. & Johnson R.H., 2009. Geochemical Characterisation of Seepage and Drainage Water Quality from Two Sulphide Mine Tailings Impoundments: Acid Mine Drainage versus Neutral Mine Drainage. Mine Water and the Environment 28, 30-49.
ISAVI 2014. Luikonlahden kaivoksen ja rikastamon ympäristöluvan muutos ja toiminnan aloituslupa, Kaavi. Aluehallintovirasto, Itä-Suomi, Päätös Nrp 52/2014/1 (Dnro ISAVI/86/04.08/2012) 3.7.2014. 102 s.
Kauppila, P.M. & Räisänen M.L. 2014. Kohde 2. Luikonlahden suljettu vanha kuparikaivosalue ja sivukivi- ja rikastushiekka-alueiden vesien käsittely passiivisilla menetelmillä. Teoksessa: Tornivaara Anna (toim.) ja Päivi M. Kauppila (toim). Kaivoksen sulkeminen ja jälkihoito; Ekskursio Luikonlahden ja Keretin kaivosalueille. Geological Survey of Finland, Guide 60, 7-24.
Kumpulainen, S., Carlson, L. & Räisänen, M.L. 2007. Seasonal variations of ochreous precipitates in mine effluents in Finland. Applied Geochemistry 22, 760-777.
Niskavaara, H. 1995. A comprehensive scheme of analysis for soils, sediments, humus and plant samples using inductively coupled plasma atomic emission spectrometry (ICP-AES). In: S. Autio (ed.): Geological Survey of Finland, Current Research 1993-1994. Geological Survey of Finland, Special Paper 20, 167-175.
Peltonen, P., Kontinen, A., Huhma, H. & Kuronen, U. 2008. Outokumpu revisited: New mineral deposit model for the mantle peridotite-associated Cu-Co-Zn-Ni-Ag-Au sulphide deposits. Ore Geology Reviews 33, 559-617.
Räisänen, M.L. 2003. Kaavin tehtaan rikastushiekka-altaan ympäristön nykytila ja suositukset jälkihoitotoimenpiteille. Geological Survey of Finland, Regional Office for Eastern Finland. Unpublished report Dnro K40/41/02. 42 p.
Räisänen, M.L. 2009a. Luikonlahden Suursuon ja suljetun kaivosalueen kosteikkopuhdistamojen veden laatu ja toimivuus 2007-2008. Geological Survey of Finland, Regional Office for Eastern Finland. Unpublished report Dnro K464/42/2008. 40 p.
Räisänen, M.L. 2009b. Capability of natural and constructed wetlands to mitigate acidic leakage from closed mine waste facilities – cases in Eastern Finland. Conference Proceedings Securing the future, Mining, metals and society in a sustainable society and 8th ICARD International Conference on acid rock drainage Skellefteå, Sweden June 22 – June 26 2009. Electronic publication. 10 p.
Räisänen, M.L., Hämäläinen, L. & Westerberg, L.M. 1992. Selective extraction and determination of metals in organic stream sediments. Analyst 117, 623-627.
Räisänen, M.L. & Juntunen, P. 2004. Decommissioning of the old pyritic tailings facility previously used in a talc operation, eastern Finland. In: Jarvis, B.A. Dudgeon and P.L. Younger (eds.): Proceedings of the Symposium Mine Water 2004 – Process, Policy and Progress vol. 1, 91-99.
Räisänen, M.L., Kauppila, P.M. & Myöhänen, T. 2010. Suitability of static tests for acid rock drainage assessment of mine waste rock. Bulletin of the Geological Society of Finland, Vol. 82, pp 33–43.
Walton-Day, K. 2003. Passive and active treatment of mine drainage. In: J L. Jambor, D.W. Blowes & A.I.M. Ritchie (eds.): Environmental aspects of mine wastes. Mineralgical Association of Canada. Short Course Series vol. 31, p. 335-359.
Woulds, C. & Ngwenya, B.T. 2004. Geochemical processes governing the performance of a constructed wetland treating acid mine drainage, Central Scotland. Applied Geochemistry 19, 1773–1783.