Water management Research and Development

Antti Pasanen & Anu Eskelinen, Geological Survey of Finland, P.O. Box 1237, FI-70211 Kuopio, FINLAND, e-mail: antti.pasanen(at)gtk.fi, anu.eskelinen(at)gtk.fi

Introduction

Water management Research and Development presents R&D studies carried out during Closedure project. In the water management work package, the objective was to develop a new method to recognize and study groundwater transport paths in crystalline bedrock using geophysical methods and isotopic composition of groundwater and mine water.

The study area for the case study was the Talvivaara Mining PLC in Finland. The original study focused on the impacts of the Talvivaara mine gypsum pond leakage to groundwater and the results for this are presented in Finnish in Forss et al. (2013), Eskelinen et al. (2013) and Pasanen et al. (2014). In addition to these reports, GTK continued to study and develop the methodology on the use of isotopic composition of water in identifying the preferable groundwater flow in hard, almost impermeable, crystalline bedrock through fracture zones.

The methodologies and the results for the isotopic studies and geophysics used are briefly summarized in this section and in Water management research methods section. The description of geophysical methodologies can also be found in the Monitoring section. The objectives of the study were:

  1. To develop a combination of geophysical methods that can be used to identify the bedrock fracture zones and to pinpoint the diamond coring locations with great precision so that the water samples taken would represent the groundwater in fracture zones rather than in almost impermeable, unfractured bedrock.
  2. To study the usability of oxygen (18O/16O), hydrogen (2H/1H), strontium (87Sr/86Sr), sulphur (34S/32S) and uranium (234U/238U) isotopes in identifying the hydraulic connections in bedrock between sampling points.

The gypsum bond, in which so called gypsum waste from the metal recovery is stored, leaked in late 2012 and again in early 2013 releasing ca. 1,2 Mm3 of acidic, metal bearing water to the environment, of which most is currently dammed in the mining area. The leak and the dammed waters caused risks to the groundwater which were studied in the original study (Forss et al. 2013, Eskelinen et al. 2014, Pasanen et al. 2014). The preliminary study suggested, and later focused studies supported the hypothesis, that the sediment thickness and permeabilities are low compared to possible bedrock fracture zones and the studies were focused to identify the possible hydraulic connections in bedrock fracture zones and to study the groundwater flow and transport paths in bedrock. In addition to this, hydraulic conductivities and heads were measured and water geochemistry was sampled and analyzed.

This article summarizes the performed studies and their results. More detailed descriptions of the methodology and results of the research is presented in respective pages:

Study area

The study area is located in the central part of Eastern Finland in the Sotkamo municipality (Fig. 1).

Talvivaara mine utilizes low grade polymetallic black schist ore which is extracted using bioheapleaching method, the first of a kind developed for extraction of nickel in Nordic climatic conditions. Mine started its operation in 2008. In the bioheapleaching the ore is quarried and transported to vast heaps where naturally occuring bacteria are used to extract the metals. The heaps are fed with a low-pH solution which accelerates the naturally occuring phenomena of bacterial leaching. The solution is pumped to the metal extraction factory where metals are precipitated and the so called gypsum waste produced in the precipitation is transported to the gypsum pond for disposal. After certain time at the first stage bioheapleaching the ore is transported to the second stage bioheapleaching piles, where the process continues at the slower pace and the heaps are closed when appropriate.

The main product of the mine is nickel and by-products are zinc, copper and cobalt. The uranium extraction from the multimetal ore has also been permitted and planned but the financial and water management problems have prevented the start of the uranium production.

Figure 1. Location of Talvivaara mining district. © GTK, Basemaps © National Land Survey of Finland. 

The bedrock in the area can be divided in two distinct parts. The western side of the area, where the gypsum pond leak occurred, consists of Archean basement rocks (pegmatites, tonalithic migmatites, quartzwackes, greywackes, diabases and quartzites, Fig. 2.,Bedrock of Finland – DigiKP. Digital map database [Electronic resource]. Espoo: Geological Survey of Finland [referred 05.05.2014]. Version 1.0.). The eastern side of the area consists of schists and paraschists of which the blackschist is being quarried (Fig 2.). The effect of the Archean basement rocks to groundwater is thought to be minimal during the time of mining (since 2008) compared to the schist area. Therefore, it can be assumed that possibly anthropogenically altered groundwater has not been significantly altered in contact with Archean basement rocks giving a good starting point in recognizing the geochemistry from the gypsum pond leak and other anthropogenic contamination and using these as geochemical tracers. The Quaternary sediments overlay the bedrock in most places in the mining area. They consist mainly of peat and till. Geotechnical drillings, made in gypsum pond area, strengthens this interpretation, but also sand units were observed. It was interpreted that the groundwater movement in sediments is less significant than in bedrock fracture zones and the investigations were directed to bedrock groundwater.

Figure 2. Bedrock lithology in Talvivaara mining area ( © GTK). The main study area is south of gypsum pond and west of first stage bioheapleaching. Basemaps © National Land Survey of Finland. 

Methodology

The methodology used in the study is briefly described and further discussed in respective web pages (see the links in the Introduction).

As described earlier the study focused on bedrock groundwater in fracture zones. This approach needed a use of several different methods of which the data was interpreted first individually to guide more detailed investigations and in later stage the data was also interpreted against each other to reduce the ambiguity in interpretation. The methods used varied from coarse areal methods at the beginning to detailed methods at the end.

The areal methods used consisted of map and topography interpretation and interpretation of aerogeophysical data. The map and topography interpretation consisted of interpretation of sedimentary landforms, bedrock outcrops and surface water flow patterns. The aerogeophysical interpretation was based on the low altitude flying data. The aeroradiometric and aeroelectromagnetic data was used to interpret the rough sediment thickness and bedrock outcrops. The aeromagnetic data in conjunction with Lidar-DEM data was used to interpret the lineations at the study area. The interpretation of lineations gives an estimate of the bedrock fracture zone locations. Ground based geophysics and drillings are always needed to redefine the results and to reduce the ambiguity.

The detailed, ground based geophysical methods consisted of gravimetric measurements, magnetic measurements, refraction seismics, electrical resistivity tomography (ERT) and ground penetrating radar (GPR). The data from these methods were interpreted both separately and in conjunction with other methods and existing groundtruth data. The data was used to redefine the aerial interpretations and to pinpoint the coring locations.

The bedrock wells used for groundwater sampling and verification of geophysical interpretation were done using a diamond drilling rig and monitoring wells installed in the Quaternary sediments using a rotary drill. Eight new wells were installed to 100 meters depth below surface in previously interpreted fracture zones and five to Quaternary sediments. The depth of the bedrock wells was selected to balance the cost and areal scope. The cores were logged and photographed and it was notified that six out of eight drillings were in fracture zones.

The hydraulic conductivities of the fractured bedrock were measured using slug-tests in four bedrock wells to estimate the water flow velocities. In addition, velocities were also estimated between two bedrock wells P7 and FID28 (Fig. 3), based on the observation of similar S, Ca and K concentration ratios in both wells, suggesting a a hydraulic connection between the wells. The distance between the wells and  times from the beginning of the mining process (from the receiving of environmental permit) to sampling date and from the filling of the first stage bioheapleaching to sampling date was used to calculate the flow velocities.

Geochemistry of the water was sampled in 29 sampling points with a total of 42 samples (Fig. 3). The sampling was done in bedrock and sediment groundwater, surface water and process water. In bedrock wells cored for this study and already at the study area, and consisting of a core log, a sample was taken with bailer in several depths interpreted to represent the highest hydraulic conductivities i.e. high bedrock fracturing. In other localities bulk sampling was used. The field analyses were made with YSI Professional Plus field meter consisting of measurements of pH, temperature, redox potential, electrical conductivity and dissolved oxygen. The alkalinity of the samples was measured in field with titration. The laboratory analyses consisted of dissolved and total concentrations of elements, Fe2+, TOC, DOC, anions, suspended solid content and alkalinity.

Figure 3. Sampling sites in Talvivaara study area. The vp+ indicates the phreatic surface of the groundwater (m asl). The groundwater observation wells marked with FID XX and RX were installed in this study. The inset map shows the background sampling points ca. 5 km north of the open pit. Basemaps © National Land Survey of Finland. 

The isotopes were sampled at the same locations and depths for O and H isotope analyses than for geochemical analyses. 25 samples were selected for Sr, U, S and Pb isotope analyses of which the concentration of Pb was deemed low and unusable in the study.

Results and discussion

The results of the methodologies used are presented in respective web pages (see the links in the Introduction). In this section the main results of the study are summarized and discussed.

When studying the possible contamination of the bedrock groundwater and movement of groundwater in bedrock fracture zones it is uttermost important to pinpoint the coring locations such that the water samples taken represent the water in fracture zones rather than in less fractured bedrock. The geochemical composition of the waters can be totally different, even though, the distance between the sampling points is only metres or few tens of metres. This is due to high variability in hydraulic conductivities between fracture zones and less fractured bedrock and different residence times. In isotopic studies, especially, the sampling of less fractured bedrock is also necessary to gain enough background information for the interpretation. Coring is expensive compared to ground based geophysics and, therefore, an areally comprehensive geophysical study is needed to reduce the costs. The geophysical methods need to be selected according to the problem and the geological setting and preferably several methods should be used to reduce the ambiguity in the interpretation. For example, if this study would have been done in the eastern part of the area at the schist bedrock, ERT method would have been useless due to the high electrical conductivity of the strata.

The interpretations of the geophysical measurements indicate that there are several fracture zones in the study area. Three fracture zones are located under the gypsum ponds (fracture zone 1, 2 and 3, Fig. 3). There is a NE-SW oriented fracture zone (fracture zone 4, Fig. 3) located south of the gypsum ponds. According to the geochemical results and isotopic composition this fracture zone collects and possibly transports groundwater outside of the study area.

According to the water level measurements the groundwater flow in bedrock occurs along fracture zones 1 and 3 towards fracture zone 4 (Fig. 3). Along fracture zone 4 groundwater flows from primary heaps to the SW. The groundwater flow directions are based on the assumption that the fracture zones are open and the connections between different fracture zones actually exists.

In this study it was not possible to calculate the exact flow velocities for the groundwater movement in bedrock. The flow velocities and the transport times were estimated based on the water level measurements and K-values. The K-values were estimated from the slug-tests in several wells. The drill cores indicate that the bedrock is highly fractured in places. Therefore, it seems possible that groundwater flow also occurs via by pass routes. Within these by pass routes the K-values can be notably larger than is estimated based on the slug-tests. The mean effective velocities calculated from hydraulic conductivities, effective porosity and ground water head differences is estimated to vary between 22 m/a and 282 m/a. The velocity calculated between wells P7 and FID 28 (Fig 3.) from the receiving of the environmental permit and filling of first phase bioheapleach to sampling date in fracture zone 4 vary between 148 m/a and 339 m/a which can be assumed to be minimum velocities.

Based on the geochemical analysis of groundwater, there is a similar S, Ca and K concentration ratios in two bedrock wells P7 and FID28. Also these similarities in chemical composition suggest the hydraulic connection between these wells. The waters from the primary heap have partially leaked in bedrock groundwater and transported along fracture zone 4 to the SW.

In this study the analysis of the stable isotopes of oxygen and hydrogen did not indicate surface water mixing with bedrock groundwater. The similarities in isotopic composition suggested hydraulic conductivies between wells P7 and FID 28 and P7 and FID 0 (Figs. 3 and 4). Other analyzed isotopes (S, Sr and U) did show some influence of the mining process on lake waters and shallow groundwater. The latter results cannot be applied for identifying bedrock groundwater flow paths. The lack of strong evidence for mixing based on isotope data was not expected because geochemical data had indicated contamination with process chemicals in some of the bedrock wells.

The geophysical methodologies applied proved to be useful and cost efficient for locating water bearing fracture zones. Further research is needed to study the hydrogeological properties of the fracture zones and the connection between the fractures. Also on-line monitoring should be applied to receive continuous groundwater data. The geophysical methodology in studies of groundwater in fracture zones is discussed in respective page (Geophysical methods in bedrock groundwater studies).

In bedrock groundwater the effect of the gypsum pond leak could not be detected. Minor contamination was observed but that is interpreted to have been caused by seepage through the pond structures. In sediment groundwater the contamination is stronger than in bedrock groundwater, but they don’t indicate the effect of the leak either.

Figure 4. Interpreted groundwater flow routes in bedrock fractures in Talvivaara mine gypsum pond area (solid arrows) and their estimated continuations (dashed arrows). Basemaps © National Land Survey of Finland. 

Conclusions

The following conclusions can be drawn from the study.

  • Pinpointing the coring sites and sampling locations is crucial when studying hydraulic connections and transport in crystalline bedrock groundwater
    • Geophysical methods should always be used to lower the costs before drilling
  • Similar traces of chemical contamination in P7 and FID 28 indicate possible long term seepage from primary heap and transport in bedrock groundwater through fracture zone
    • Isotopic composition of water (δ18O, δ2H) supports this interpretation
  • Isotopic composition (δ18O, δ2H) suggests a hydraulic connection and water flow from FID 0 to P7
    • Chemical analyses do not indicate the connection between the observation points
  • Fracture zones can act as preferential flow paths in bedrock and can be significant in transporting possible adverse substances outside the mining area

References

Eskelinen, A., Forsman, P., Hendriksson, N., Pasanen, A. & Kittilä, A. 2013. Vesinäytteenotto Talvivaaran kaivosalueella. Geological Survey of Finland, archive report. 36 p.

Forss, H., Lerssi, J., Huotari-Halkosaari, T., Pasanen A., Eskelinen, A. & Kittilä, A. 2013. Talvivaara – Geofysikaaliset tutkimukset 2013. Geological Survey of Finland, archive report. M21K2013. 41 p.

Kittilä, A. 2015. Groundwater flow paths in the bedrock fracture zones revealed by using the stable isotopes of oxygen and hydrogen in the Talvivaara mine gypsum pond area, northeastern Finland. Master’s thesis, University of Helsinki, Department of Geoscienses and Geography, Division of Geology. 60 p.

Pasanen, A., Eskelinen, A., Räisänen, M-L., Lerssi, J. & Kittilä, A. 2014. Talvivaaran kipsisakka-altaan vuodon pohjavesivaikutusten selvitys ja leviämisen ja haitan arviointi. Geological Survey of Finland, archive report. M21K2013. 52 p.