You are on page 1of 74

L I C E N T I AT E T H E S I S

Department of Civil, Environmental and Natural Resources Engineering


Division of Geosciences and Environmental Engineering
Effectiveness of Reclamation by Backfilling
ISSN: 1402-1757 ISBN 978-91-7439-363-7 and Sealing at Kimheden Open-Pit Mine,
Luleå University of Technology 2011
Northern Sweden

Lucile Villain
Effectiveness of Reclamation by Backfilling
and Sealing at Kimheden Open-Pit Mine,
Northern Sweden

Lucile Villain

Luleå University of Technology


Department of Civil, Environmental and Natural Resources Engineering
Division of Geosciences and Environmental Engineering
Cover Image: Metal-contaminated seepage at Kimheden mine.

Printed by Universitetstryckeriet, Luleå 2011

ISSN: 1402-1757
ISBN 978-91-7439-363-7
Luleå 2011
www.ltu.se
Abstract
Reclamation of mine sites is a very recent concept on the scale of mining history. It in-
volves the prevention or mitigation of the environmental impacts from mining, and the establish-
ment of sustainable post-mining uses of the land. Many methods have been tested to reduce the
contamination from mining waste, but their actual performance depends highly on the nature of
the site. In this thesis, the effectiveness of reclamation is investigated at Kimheden open-pit copper
mine in Västerbotten, northern Sweden.
Sulphide-containing waste rock left from the extraction of the ore was originally dumped
on the surface close to the two open pits at Kimheden. The contact of sulphides with water and
oxygen induced the generation of copper- and zinc-rich acid mine drainage. Therefore, reclama-
tion of the mine involved back¿lling of the mine waste into the open pits and covering with a till
composite dry cover including a sealing layer, in order to reduce the oxygen ingress to the waste.
Geochemical and geophysical studies conducted in 2009-2010, complemented by moni-
toring data from the mining company, were interpreted to assess the success of the reclamation.
Long-term annual monitoring of the contaminated drainage by the mining company showed that
concentrations of copper and zinc added together have decreased by more than 87% since the
completion of reclamation in 1996. However, the decrease occurred rapidly, and metal concentra-
tions in the last ten years have remained stable at values that are still not satisfactory for discharge
into the environment. Furthermore, pH in the drainage increased only slightly after reclamation.
In 2009, seepage water from one of the pits exhibited a pH of 3.0, and copper, zinc and aluminium
concentrations of 3 mg/L, 0.3 mg/L and 20 mg/L, respectively. The partial success of the reclama-
tion in mitigating the acid mine drainage may be explained by persistent oxidation of the back¿lled
mine waste in spite of the application of the dry cover. This would imply that oxygen gas or oxy-
genated water is still accessing the waste rock, either through the dry cover, or through fractures
in the pit walls. These hypotheses were investigated in the ¿eld with two different geophysical
methods, ground penetrating radar and direct current resistivity.
With ground penetrating radar, good-quality image of the sealing layer could be obtained
at one of the two back¿lled pits, and no evident sign of disturbance in the layer could be found.
Resistivity surveys succeeded to image the drainage from the back¿ll, and suggested that it was
seeping out through the dry cover, with a risk of erosion of the sealing layer. Water discharge
measurements and resistivity data indicated that the current channelling of the mine drainage to
a treatment pond is inadequate, since a large proportion of the contaminated water disappears as
groundwater. The interpretation of the resistivity pro¿les supported by archive data suggested that
fractures in the bedrock close to the pits could explain the ingress of oxygen gas or oxygenated
water through the pit walls, and would be the primary reason for on-going sulphide oxidation.
The results of these studies may help to improve the reclamation of the mine, and in a
broader perspective, provide information for reclamation at other sites.
Acknowledgements
I am grateful to my supervisors, Lena Alakangas and Björn Öhlander, for guiding me
through an exciting research project. In addition to this guidance, a driving element of this work
was the ¿nancial support of the Georange non-pro¿t organisation. Their essential contribution is
therefore acknowledged. Financial support was also received from Boliden AB, acknowledged in
the same way.
2ther important actors who I would like to thank are Bert-Sive Lindmark, who
makes ¿eld work easier and nicer, and the people at Boliden and SG8 who helped me
¿nd out buried information about Kimheden. &olleagues who share their skills, advice,
or interest to my work are also kindly thanked. I would like to name Peter Nason,
'mytro Siergieiev, Nils Sundström, Nils Perttu, Saman Tavakoli and <u -ia, in particular.
The help I got from Milan Vnuk and (rika Resin to realise the ¿gures is also very much
appreciated.
Besides, I am naturally thinking about my family, who believes in me from far away.
Lastly, and because research is not just a few hours a day, I want to thank you, -oakim, for your
loving patience and understanding, as well as dedicated assistance in the realisation of this thesis.

Lucile Villain
Lulen, November 2011
List of papers

The following papers are included in this thesis:

Paper I
(IIHFWVRIEDFN¿OOLQJDQGVHDOLQJRIZDVWHURFNRQZDWHUTXDOLW\DW.LPKHGHQRSHQSLWPLQH
QRUWKHUQ6ZHGHQ(Manuscript).

Lucile Villain, Lena Alakangas, Björn Öhlander

Paper II
*URXQGSHQHWUDWLQJUDGDUDQGUHVLVWLYLW\PHWKRGVWRHYDOXDWHWKHUHFODPDWLRQRIDQRSHQSLW
PLQHE\EDFN¿OOLQJDQGVHDOLQJ(Manuscript).

Lucile Villain, Nils Sundström, Nils Perttu, Lena Alakangas, Björn Öhlander

The following conference papers have been published, but are not included in the thesis:

Villain, L., Alakangas, L. & Öhlander, B. (2010). Geochemical evaluation of mine water quality in
an open-pit site remediated by back¿lling and sealing. In: IMWA 2010 - proceedings of the Inter-
national Mine Water Association Symposium 2010, Sydney, Canada, Wolkersdorfer C. & Freund
A. (Eds.), Cape Breton University Press, pp. 515—518.

Villain, L., Sundström, N., Perttu, N., Alakangas, L. & Öhlander, B. (2011). Geophysical investi-
gations to identify groundwater pathways at a small open-pit copper mine reclaimed by back¿lling
and sealing. In: IMWA 2011 - proceedings of the International Mine Water Association Congress
2011, Aachen, Germany, Rüde, R. T., Freund A. & Wolkersdorfer C. (Eds.), pp. 71—76.
Contents
/ŶƚƌŽĚƵĐƟŽŶ ϯ
ZĞĐůĂŵĂƟŽŶŽƉƟŽŶƐĨŽƌǁĂƐƚĞƌŽĐŬĂƚƐƵƌĨĂĐĞŵŝŶĞƐŝƚĞƐ ϯ
^ƵůƉŚŝĚĞŽdžŝĚĂƟŽŶĂŶĚƉƌŽĚƵĐƟŽŶŽĨĂĐŝĚŵŝŶĞĚƌĂŝŶĂŐĞ ϯ

^ƉĞĐŝĮĐĐŚĂƌĂĐƚĞƌŝƐƟĐƐŽĨǁĂƐƚĞƌŽĐŬ ϰ

WƌŝŶĐŝƉůĞŽĨƉĂƐƐŝǀĞĐŽŶƚƌŽůĂƚƚŚĞƐŽƵƌĐĞ ϰ

/ŶͲƉŝƚďĂĐŬĮůůŝŶŐĂƚƐƵƌĨĂĐĞŵŝŶĞƐ ϰ

DŝŶĞǁĂƐƚĞĐŽǀĞƌƐ ϱ

tĂƚĞƌŵĂŶĂŐĞŵĞŶƚŵĞƚŚŽĚƐĂƚƐƵƌĨĂĐĞŵŝŶĞƐ ϳ

KďũĞĐƟǀĞƐĂŶĚƐĐŽƉĞŽĨƚŚĞƚŚĞƐŝƐ ϳ
<ŝŵŚĞĚĞŶƐƚƵĚLJƐŝƚĞ ϴ
>ŽĐĂƟŽŶ͕ĐůŝŵĂƚĞĂŶĚŐĞŽůŽŐLJ ϴ

DŝŶŝŶŐĂŶĚƌĞĐůĂŵĂƟŽŶĞǀĞŶƚƐ ϵ

DĞƚŚŽĚƐ ϵ
,LJĚƌŽͲŐĞŽĐŚĞŵŝĐĂůǁĂƚĞƌƐĂŵƉůŝŶŐƐ ϵ

'ĞŽƉŚLJƐŝĐĂůŝŶǀĞƐƟŐĂƟŽŶƐǁŝƚŚ'WZĂŶĚƌĞƐŝƐƟǀŝƚLJ ϭϬ

&ŝŶĚŝŶŐƐĨƌŽŵƚŚĞƐƚƵĚŝĞƐ ϭϭ
īĞĐƚƐŽĨƚŚĞƌĞĐůĂŵĂƟŽŶŽŶDŐĞŶĞƌĂƟŽŶ ϭϭ

ĞĐƌĞĂƐĞŽĨƵĂŶĚŶĐŽŶĐĞŶƚƌĂƟŽŶƐ;WĂƉĞƌ/Ϳ ϭϭ

^ŝŐŶƐŽĨŽŶͲŐŽŝŶŐDƉƌŽĚƵĐƟŽŶ;WĂƉĞƌ/Ϳ ϭϮ

ŽŶƚĂŵŝŶĂŶƚǁĂƐŚͲŽƵƚĨƌŽŵƚŚĞƉŝƚƐ;WĂƉĞƌƐ/ĂŶĚ//Ϳ ϭϯ

WŽƚĞŶƟĂůƐŽƵƌĐĞƐŽĨŽdžLJŐĞŶƚŽƚŚĞƉŝƚƐ;WĂƉĞƌ//Ϳ ϭϰ

EĞĞĚĨŽƌĨƵƌƚŚĞƌŵŝƟŐĂƟŽŶŵĞĂƐƵƌĞƐ;WĂƉĞƌƐ/ĂŶĚ//Ϳ ϭϱ

ŽŶĐůƵƐŝŽŶƐ ϭϱ
&ƵƚƵƌĞƐƚƵĚŝĞƐ ϭϲ
Z&ZE^ ϭϳ
ƉƉĞŶĚĞĚƉĂƉĞƌƐ

1
2
/ŶƚƌŽĚƵĐƟŽŶ
After closure of a mine site, rock and chemical residues, mine workings and facilities may
present long-term risks for the surrounding environment. Most perceptible risks are the disruption
of the landscape and the impossibility to re-use the land, but the toxicity of mine residues to natural
water systems and ecosystems is a subject of at least equal concern. Therefore, decommissioning
of a mine at the end of mining activities has to include reclamation of the site. If reclamation
initially involved the re-vegetation of mined-out surfaces to restore the landscape and limit the
effects of erosion (Wireman, 2001), it now has to make sure that all measures necessary to limit
pollution and allow sustainable future use of the land are taken. One major pollution problem at
hard-rock and coal mines is the production of the so-called Acid Mine Drainage (AMD) from the
reaction of sulphide minerals of exposed Fe-rich rocks with water and oxygen, usually releasing
high concentrations of metals and acidity to natural surface and ground-water systems. Many
mitigation methods have been proposed and implemented to control AMD at mine sites, but
their sustainability in the long term is still intensely debated. Reliance on small-scale tests and
prediction models are not suf¿cient to apprehend the success of the mitigation methods in real
conditions (Wilson, 2008). Therefore, assessment of the effectiveness of remediation activities at
reclaimed mine sites has to become an established means of performance assessment in the same
way as laboratory scale and meso-scale evaluations.
The present thesis focuses on the reclamation of a small open-pit site, Kimheden (northern
Sweden), where AMD mitigation has included back¿lling of waste rock into the pits and application
of a soil cover. Back¿lling of mine voids after ore extraction has been largely performed as a way
to reduce the impact area of waste deposits, but only few monitoring data are available about the
performance of this method (Tremblay & Hogan, 2001). Application of a dry cover is a well-
known method to reduce the vertical diffusion of oxygen and percolation of water into mine waste
deposits, which has shown both promising results and challenges (Tremblay & Hogan, 2001;
Wilson et al., 2003; Höglund et al., 2004; Wilson, 2008). Based on the studies included in this
thesis, the effectiveness of these methods in improving water quality at Kimheden site fourteen
years after completion of the reclamation is discussed.

ZĞĐůĂŵĂƟŽŶŽƉƟŽŶƐĨŽƌǁĂƐƚĞƌŽĐŬĂƚƐƵƌĨĂĐĞŵŝŶĞƐŝƚĞƐ

^ƵůƉŚŝĚĞŽdžŝĚĂƟŽŶĂŶĚƉƌŽĚƵĐƟŽŶŽĨĂĐŝĚŵŝŶĞĚƌĂŝŶĂŐĞ
Mining of sulphidic metal ores (e.g. Cu, Zn, Pb and Au ores) generates two main types
of waste, waste rock and tailings, where variable proportions of sulphide minerals might be left.
Waste rock (dominated by coarse material) is the rock fraction that is removed to access the ore,
while tailings are the ¿nely grinded waste produced when processing the ore. When mine waste is
exposed to oxygen and water, during and after mining operations, sulphide minerals will become
unstable and oxidise. The oxidation of Fe-sulphides such as pyrite or pyrrhotite may result in Acid
Mine drainage (AMD), an acidic water rich in sulphate, heavy metals and metalloids, unless the
buffering capacity of minerals like carbonates is high enough to neutralise the acid (Lottermoser,
2003).
The overall sulphide oxidation reaction of pyrite is usually summarised in the following reaction
(INAP, 2009):
FeS2 + 7/2O2 + H2O = Fe2+ + 2SO42- + 2H+ (1)

3
^ƉĞĐŝĮĐĐŚĂƌĂĐƚĞƌŝƐƟĐƐŽĨǁĂƐƚĞƌŽĐŬ
According to Younger et al. (2002), more than 99% of the waste rock generated world-
wide is produced by surface mining. As this mining method has developed dramatically since
the middle of the last century, so has the generation of waste rock. Therefore, the reclamation of
abandoned waste rock piles has become one of the most important issues related to environmental
management of mine sites. Younger et al. (2002) recognised two main areas of research developed
in this ¿eld: the conditions for a sustainable re-vegetation of the waste rock deposits, and the
prevention and mitigation of contaminated leachate from the waste. The latter subject has received
considerable attention, with studies on the geochemistry of waste rock, their acid generating
potential and possible mitigation options.
Waste rock mainly differs from tailings in their particle size. Waste rock is primarily coarse-
grained (1 mm-50 mm), while tailings are principally ¿ne-grained (1 mm) (Younger et al., 2002).
In the case of sulphidic deposits, they also differ in their reactivity to oxidation, with tailings
tending to be more reactive than waste rock, since they have a much larger speci¿c surface area
(Lottermoser, 2003). Waste rock on the other hand, due to coarser particle size, has much higher
permeability than tailings, and the transport of water and oxygen in the waste is therefore much
faster than in tailings. In such conditions, production of AMD may proceed faster in waste rock
compared to tailings, despite the higher reactivity of tailings (Mitchell, 1999).

WƌŝŶĐŝƉůĞŽĨƉĂƐƐŝǀĞĐŽŶƚƌŽůĂƚƚŚĞƐŽƵƌĐĞ
As an environmentally sustainable solution, prevention and mitigation of AMD at the
source have been advocated as opposed to reactive treatment of the mine drainage (Tremblay &
Hogan, 2001; Johnson & Hallberg, 2005; INAP, 2009). The aim is therefore to limit the oxidation
of sulphides and the leaching of oxidation products from the mine waste at its source, in order to
reduce the migration of contaminants to natural water bodies. As production of AMD relies on the
access of oxygen and water to the waste, prevention and mitigation methods are mainly based on
the exclusion of oxygen and/or water to the waste deposits. Traditional prevention and mitigation
methods for waste rock at surface mine sites are back¿lling in mine voids or deposition in piles,
followed at mine closure by the application of different types of dry or water covers. Back¿lling
and cover applications are further described in the two following sections.

/ŶͲƉŝƚďĂĐŬĮůůŝŶŐĂƚƐƵƌĨĂĐĞŵŝŶĞƐ
Surface mines producing sulphidic waste are opencast mines working coal, and open-pit
mines working metals. Opencast mines typically operate in areas of waste rock to coal ratios in the
order of 20:1 (Younger et al., 2002), consequently accounting for the generation of a large amount
of waste rock. Nevertheless, the extraction method from coal seams proceeds mostly horizontally,
which results in large and shallow voids and facilitates the immediate disposal of the usually
voluminous mine waste in previously mined-out areas (Phelps, 1990; European Commission,
2007). Therefore, at closure of the mine, most of the waste is already back¿lled and mine waste
management will be focused on the stabilisation and re-vegetation of the spoil. Conversely, it
has been traditionally recognised that vertical extraction methods at open pits of hard-rock mines
would not allow practical and economical back¿lling, and would generally be left as mined-out
voids or evolve as pit lakes (European Commission, 2007). Nevertheless, back¿lling is not an
uncommon practise even at open-pit sites, but relatively little scienti¿c research has covered its
implementation (Tremblay & Hogan, 2001). In certain conditions, if the ore was extracted close
to the surface and the waste rock piles are located close to the pit, back¿lling may become the
preferred option also at surface hard-rock mines (Gray & Gray, 1992; Younger et al., 2002). In an
environmental perspective, the advantages of in-pit back¿lling of waste rock have been recognised

4
by Chapman et al. (1998). Positive effects include the reduction of the area of waste to be managed,
the availability of a simple and stable containment for the waste, and the limitation of the extent
of the contaminated area.
Despite these advantages, back¿lling alone does not prevent the formation of AMD if the
waste rock is reactive and the pit does not naturally evolve as a pit lake. Tremblay and Hogan
(2001) reviewed possible alternatives to limit the ingress of oxygen and/or water to the back¿lled
waste with different con¿gurations of water covers or dry covers. The principle of these covers
is described in the next section. Other strategies may be the selective placement of waste rock in
the pit, and the addition of alkaline material to reduce the acid generation (Chapman et al., 1998).

DŝŶĞǁĂƐƚĞĐŽǀĞƌƐ
Covers on sulphidic mine waste are technologies that were introduced in the perspective of
long-term passive prevention and control of acid generation. The primary objective of all covers is
to reduce the ingress of oxygen into the waste deposit, in order to limit the oxidation of sulphides
and the production of AMD (Höglund et al., 2004). The most common way to achieve this objective
is to use the principle of reduced diffusivity and solubility of oxygen in water or water-saturated
media compared to air. Thus, in areas where a thick water cover can be maintained throughout the
year, Àooding of waste rock or tailings may develop conditions of reduced oxygen in the waste,
thereby limiting oxidation rates. Low permeability barriers (also called sealing layers), are soil
covers that exhibit high moisture content and may decrease the oxygen diffusion to the underlying
waste in the same way as a water cover. Covers including a low permeability barrier may be
useful in areas where the groundwater surface is very low in the mine waste deposit. They usually
include a protective layer on top, in order to prevent deterioration of the sealing layer from freeze/
thaw effects, root penetration, drying, etc. Other types of cover have been suggested, which are
not based on the limitation of oxygen diffusion rates, but on consumption of oxygen by organic
material, or inhibition of oxidation reaction rates.
In addition to the prevention of oxygen transport to the waste, another factor is sometimes
considered equally important when designing a cover: the reduction of water percolation through
the cover into the waste deposit (Höglund et al., 2004). Reduced water inÀow to the sulphidic
material may decrease the transport of released elements from the waste and thus the quantity
of contaminants discharged. Such objective is usually ful¿lled by low permeability barriers. A
classi¿cation of different types of covers proposed by Höglund et al. (2004) is presented in Table 1.
Placement of reactive waste under water cover is usually achieved with tailings in
impoundments. However, as mentioned in the previous section, water covers may also be
established in back¿lled open pits, if the hydrological conditions are favourable. The advantage of
in-pit disposal is that mine voids may provide long-term geotechnical stability for the placement
of the waste, as opposed to impoundments that are arti¿cially raised above the ground, and require
constant maintenance and control (INAP, 2009). As good performance can be achieved with water
covers (MEND, 1997), under-water placement of mine waste in pits may become an advantageous
solution.
A type of dry cover commonly used in Sweden is a two-layer cover comprising a sealing
layer of clayey till overlain by a protective layer of unsorted till. Other types of material may
be used for the sealing layer, e.g. bentonite or compacted organic material. The Swedish EPA
evaluated the performance of such types of covers (SEPA, 1993), and estimated that they could, in
ideal conditions, reduce sulphide weathering by 94-99%, and metal leaching by 97-99.8%. Their
results are reported in Table 2. The idea is that the potential release of metals and acidity from the
mine waste in the long term will be the same as without reclamation, but the aim is to spread the
discharge over time in such a way that nature can tolerate it.

5
Table 1 Different cover types and their primary function. Table from Höglund et al. (2004).

Table 2 Performance objective of different types of low permeability barriers, in terms of decrease
of sulphide weathering, water budget in the mine waste, and resulting metal leaching. Table from
SEPA (1993).

6
tĂƚĞƌŵĂŶĂŐĞŵĞŶƚŵĞƚŚŽĚƐĂƚƐƵƌĨĂĐĞŵŝŶĞƐ
According to equation (1), water is one of the reactants of the sulphide oxidation. However,
as water is usually in large excess compared to oxygen in mine waste deposits, water primarily acts
as a transport media for oxidation products. As mentioned by INAP (2009), control of a limited
volume of AMD-impacted water is commonly more effective than control of a larger volume of less
polluted water that still remains unacceptable according to environmental guidelines. Therefore,
water management techniques at mine sites aim at reducing the quantity of AMD discharged by
the reduction of water in¿ltration and leaching.
At closure of surface mine sites, common water management techniques include (Brady et
al., 1998):

‡ Diversion ditches. Positioned upstream of the mine waste and mine workings,
they will reduce the in¿ltration of water into the sulphidic material.
‡ Collection ditches. Positioned downstream of the AMD producing areas, they will
divert the mine drainage to treatment or sedimentation ponds.
‡ Sedimentation and treatment ponds, to control erosion and treat the mine drainage
until suf¿cient improvement of its quality.
‡ Water and dry covers as described in the previous section.

KďũĞĐƟǀĞƐĂŶĚƐĐŽƉĞŽĨƚŚĞƚŚĞƐŝƐ
The primary objective of the presented work is to assess the effectiveness of reclamation by
back¿lling and sealing in reducing water contamination from Kimheden mine, northern Sweden.
The investigations are meant to assess if the mitigation measures were adequate for improvement of
the water quality until acceptable level, and if not, facilitate the selection of appropriate additional
remedial actions. In a broader perspective, the study could be used to improve back¿lling and
sealing mitigation techniques at similar sites.
The assessment of the reclamation is based on two studies. The ¿rst study, Paper 1, is a
hydro-geochemical investigation of water quality at Kimheden, discussing the evolution of the
release of contaminants from the mine since early stages of the reclamation, and current processes
of contaminant production and transport. Paper 2 presents the ¿ndings from geophysical surveys
using Ground Penetrating Radar (GPR) and Direct Current (DC) resistivity. This study was meant
to evaluate the sources of water and oxygen ingress to the back¿lled waste, characterise water and
AMD pathways on site, and identify potential sources of failure of the reclamation in the long
term.

7
<ŝŵŚĞĚĞŶƐƚƵĚLJƐŝƚĞ

>ŽĐĂƟŽŶ͕ĐůŝŵĂƚĞĂŶĚŐĞŽůŽŐLJ
Kimheden copper mine is situated in the Kristineberg mining area in northern Sweden
(Figure 1.a). The region is characterised by a continental inland climate with cold winters and warm
summers. The snow-covered period extends from October to May. Annual precipitation at the site
is ca. 400-800 mm (Axelsson et al., 1991), and annual average temperature is 0.7 ºC (Malmström
et al., 2001). The Kristineberg area is a deformed and metamorphosed Palaeoproterozoic volcanic
domain. Kimheden is one of the smaller pyrite-rich massive sulphide deposits of the area, which are
intercalated within a succession of altered felsic volcaniclastic rocks that have been metamorphosed
to quartz-muscovite-chlorite schists (Hannington et al., 2003). Common ore minerals are pyrite
and chalcopyrite along with some sphalerite, while gangue minerals are mostly quartz, sericite and
chlorite with traces of talc. Calcareous rocks are rare. The terrain is sloping towards north-west
and south-west. The top of the hill, located north-east of Kimheden, lies at an altitude of 561 m
(Hellman & Lokrantz, 2008). Rosén and Wilske (1994) indicated that the bedrock surrounding the
open pits is fractured, and that groundwater accounts for the largest proportion of inÀow to the pits.
a. b.

ARVIDSJAUR

ABBORTRÄSK WASTE ROCK DUMP

GLOMMERSTRÄSK
N OPEN PIT 1

H
MALÅ ITC
ND H
IO TC
KIMHEDEN,
EC
T DI
KRISTINEBERG LL N
CO IO
RS
VE
20 km
DI

Luleå

Skellefteå

Shaft
WASTE ROCK DUMP

Stockholm
OPEN PIT 2
TO TR
EATM
EN
T PO
ND

Figure 1 (a) Location of Kimheden open-pit site in the Kristineberg mining area, northern Sweden. (b) Aerial view
of the site in the beginning of reclamation activities (1982). The position of open pits 1 and 2 is indicated, as well as
waste rock dumps and water management ditches.

8
DŝŶŝŶŐĂŶĚƌĞĐůĂŵĂƟŽŶĞǀĞŶƚƐ
Copper ore was extracted underground and in two open pits by Boliden AB between 1968
and 1974 (Figure 1.b). The total tonnage was 0.13 Mt with a grade of 0.95% Cu (Årebäck et al.,
2005). Waste rock left from the extraction was dumped in the proximity of the pits, and rapidly
started to generate Cu- and Zn-rich AMD. As a consequence, a network of ditches was excavated
in 1982 (Figure 1.b), in order to reduce the access of water run-off to the waste rock, and to
collect the mine drainage to a treatment pond downstream of the hill. Later on, the waste rock was
disposed of into the pits in different stages, in order to limit the area of contamination. Reclamation
activities were completed in 1995-1996, when full in-pit back¿lling of the waste rock left, along
with some remains of the mining facilities and roads was performed, and a dry cover was applied
on the waste (Figure 2). The dry cover was constructed with ca. 0.3 m clayey till (sealing layer)
overlain by ca. 1.5 m unsorted till (protective layer).



 3URWHFWLYHOD\HUPXQVRUWHGWLOO
6HDOLQJOD\HUPFOD\H\WLOO






 %DFNILOOLQJRIZDVWHURFN
LQWRWKHRSHQSLW






              

Figure 2 Cross-section of open pit 1 showing the back¿lling and sealing process completed in 1995-1996. The dry
cover consists of 0.3 m clayey till (sealing layer) overlain by 1.5 m unsorted till (protective layer).

DĞƚŚŽĚƐ
While Paper I is based on hydro-geochemical water samplings, Paper II addresses results
from geophysical investigations with GPR and DC resistivity.

,LJĚƌŽͲŐĞŽĐŚĞŵŝĐĂůǁĂƚĞƌƐĂŵƉůŝŶŐƐ
The post-reclamation water sampling programme was based on eight sessions in May-October
2009 and one in August 2010. Surface and ground-water samples were collected at 2 to 14 locations
(Figure 3). Surface water was sampled in the receiving stream from the mine (= collection ditch),
and at one location upstream of the mine (SB= background surface water). SD and 6301 are the
last surface water sampling locations in the stream, which were also used for long-term monitoring
by the mining company. Groundwater was sampled both from a groundwater well installed in the
back¿lled waste of open pit 1 (G1), and upstream of the pits (GB= background groundwater). pH
was measured on site with a Metrohm 704 portable pH meter and EC with a WTW Multi 350i
multimeter. Water was ¿ltered through 0.22-—m nitrocellulose membranes washed in 5% acetic acid
into plastic bottles washed in 50% hydrochloric acid followed by 1% nitric acid. They were sent to
the ALS Scandinavia accredited commercial laboratory for metal analysis with ICP-AES and ICP-
SFMS. Further description of the sampling and analytical methods can be found in Paper I.
The long-term water quality monitoring was performed by the mining company since 1983
through sampling of un¿ltered water in the stream.

9
Figure 3 the study area, Kimheden copper mine, and the receiving stream (= collection ditch) that was sampled. The
sampling locations are indicated (‘S’ surface water, ‘G’ groundwater). SB and GB are background surface and ground-
water, respectively.

'ĞŽƉŚLJƐŝĐĂůŝŶǀĞƐƟŐĂƟŽŶƐǁŝƚŚ'WZĂŶĚƌĞƐŝƐƟǀŝƚLJ
GPR surveys were conducted on two grids located on the two back¿lled open pits (Figure 4).
One survey grid was established for resistivity measurements, consisting of 4 survey lines of 200 to
280 m, located in the vicinity of back¿lled pit 1. The topography along each pro¿le was determined
with a levelling instrument. Resistivity data was collected with the ABEM Lund Imaging system
(Dahlin & Zhou, 2006) using the multiple gradient array with an electrode distance of 2 m. This
con¿guration with the SAS4000 terrameter permits multi-channel measurements, with 4 potential
readings for each pair of current electrodes. The data was inverted using Res2Dinv with the robust
inversion constrain L1-norm (Loke et al., 2003). The radar survey was conducted using a RAMAC
GPR system from Malå Geoscience with both a 50 MHz RTA antenna and a shielded 250 MHz
antenna. Measurements were made every 5 cm and triggered using a “hip-chain”. To facilitate the
interpretation of the radar data, dc-shift and band-pass ¿lter were applied. More information on
survey methods and data analysis can be found in Paper II.

10
Complementary data were collected at two groundwater wells situated in the back¿ll of
open pit 1 (Figure 4), whereby groundwater level was measured with an electrical dip meter and
electrical conductivity was measured with a WTW Multi 350i multimeter. Archive information
from the company was also used to support the interpretation of the results.

Y PIT 1

X PIT 1

H
ITC
ND
TIO
EC
LL
CO
LINE 3 BACKFILLED OPEN PIT 1

LINE 1 0 LINE 4

LINE 2

X PIT 2

GRP grids
Y PIT 2

Resistivity grid

Groundwater wells

Reference point for the sealing layer

BACKFILLED OPEN PIT 2

100 m
0

Figure 4 Overview of the GPR and resistivity survey grids at Kimheden mine site. The position of the two groundwater
wells installed in the back¿ll of open pit 1 and the reference point for the depth of the sealing layer are also indicated.

&ŝŶĚŝŶŐƐĨƌŽŵƚŚĞƐƚƵĚŝĞƐ

īĞĐƚƐŽĨƚŚĞƌĞĐůĂŵĂƟŽŶŽŶDŐĞŶĞƌĂƟŽŶ

ĞĐƌĞĂƐĞŽĨƵĂŶĚŶĐŽŶĐĞŶƚƌĂƟŽŶƐ;WĂƉĞƌ/Ϳ
Long-term annual monitoring of dissolved concentrations by the mining company of the
target metals Cu and Zn at the end of the receiving stream (locations 6301 and SD, Figure 3)
revealed a substantial decrease as a result of the completion of reclamation activities (Figure 5).
Complete back¿lling of the two pits, and application of the dry cover in 1996, resulted in lower
and stabilised concentrations of Cu and Zn from the mine waste. Comparison between 1991 and
2009 of average concentrations of Cu and Zn added up together ([Cu]+[Zn]) at key locations of the
receiving stream shows that the minimum objective of decrease by 87% set before reclamation has
been achieved. These results suggest a successful reclamation as a ¿rst approach.

11
84-85: partial 88-89: additional 95-96: complete backfilling
backfilling of the backfilling of of the 2 pits, application of a
2 pits, liming open pit 2 composite cover
25 5
Cu (mg/L)
pH

20 Zn (mg/L) 4

15 3
Cu
Zn
pH
10 2

5 1

0 0
83

85

87

89

91

93

95

97

99

01

03

05

07

09
19

19

19

19

19

19

19

19

19

20

20

20

20

20
Year

Figure 5 Total concentrations of Cu and Zn and pH values at the end of the receiving stream (6301-SD), between 1983
and 2010. The main remedial events are indicated in chronological order. Data was provided by the mining company.
Unknown data is represented in dashes.

^ŝŐŶƐŽĨŽŶͲŐŽŝŶŐDƉƌŽĚƵĐƟŽŶ;WĂƉĞƌ/Ϳ
Concentrations of some metals and sulphur since 2002 at SD downstream are shown in
Figure 6. The results indicate that the concentrations had been stabilised already 6 years after the
reclamation, perhaps even earlier (lack of data). While oxygen depletion in the back¿lled pits and
decrease of the release of oxidation products from mitigation actions are processes expected to
occur slowly and progressively over time, the decrease of contaminant concentrations seems to have
rapidly reached a plateau. This quick stabilisation is therefore interpreted as the establishment of a
new equilibrium in the oxidation of sulphides in the waste, at a lower rate than before reclamation.
An effective evolution towards oxygen-reduced conditions in the pits would have been expressed
by continued decrease of the metal concentrations with time.

1000

100

Cu μg/L
Concentrations

Zn μg/L
10 Fe mg/L
S mg/L
Al mg/L

1
01

15

09

10

09

07

03

06

05

02

06

08
2

0
0-

4-

5-

9-

5-

9-

5-

9-

5-

9-

5-

9-

5-

9-

5-

9-
-1

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0
02

03

03

03

04

04

05

05

06

06

07

07

08

08

09

09
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

0,1
Sampling dates (2002-2009)

Figure 6 Total concentrations of Cu, Zn, Fe, S and Al at SD downstream of the mine, between 2002 and 2009. Data
was provided by the mining company.

12
Average concentrations of sulphate in 1983 and 2009 were used to estimate the decrease in
the oxidation rate between pre- and post-reclamation conditions, according to reaction (1) (2 SO4
are produced when 1 FeS2 is oxidised). According to sulphate concentrations, weathering of pyrite
would have decreased by 80% as a result of the reclamation, which is less than expected from the
application of the dry cover according to the performance objective of 94-99% evaluated by the
Swedish Environmental Protection Agency (SEPA, 1993). The limited performance is attributed
to ineffectiveness of the mitigation measures to suf¿ciently decrease the transport of oxygen into
the back¿lled waste. The origin of the inadequacy is explored in the section “Potential sources of
oxygen to the pits”.

ŽŶƚĂŵŝŶĂŶƚǁĂƐŚͲŽƵƚĨƌŽŵƚŚĞƉŝƚƐ;WĂƉĞƌƐ/ĂŶĚ//Ϳ
Field observations showed constant seepage of contaminated water from the back¿lled
open pits, with the highest loads of contaminants to the stream being produced by back¿lled pit
1 (Paper I). Leachate from pit 1 is transported downstream over a sloping seepage area into the
collection ditch leading to a treatment pond (Figure 7).

Figure 7 Picture of the sloping area downstream of back¿lled open pit 1. The dry cover is visible in the foreground,
followed by the seepage area from the pit, and natural wood in the background. The collection ditch is situated
downstream of the seepage area, just before the trees. Photo by Yu Jia.

The DC resistivity survey conducted over back¿lled open pit 1 (Paper II) imaged the extent
of the seepage area, and suggested that the seepage from the waste occurs mostly on the ground
and in the subsurface, rather than in the deeper bedrock (Figure 8).
Water sampled in 2009 in the ditch below pit 1 (= S1a of the receiving stream) presented
high dissolved concentrations of Cu (3 mg/L), Al (20 mg/L), and Zn (0.3 mg/L). The ditch should
collect most of the mine drainage, as the mine drainage is transported at limited depth. However,
water discharge measurements in the collection ditch (Paper I) and mapping of the seepage with
resistivity (Paper II) showed that this ditch fails to properly divert the polluted water to the treatment
pond, and a large part of the seepage is disappearing as groundwater instead.

13
Intersection with
profile 2
Intersection with
profile 1
Intersection with
Possible profile 3
seepage through
the cover Collection ditch
Possible
water inflow
to the pit

Bedrock Pit 1 Seepage area

Figure 8 2D inverted resistivity pro¿le of line 4, crossing back¿lled open pit 1, and extending on the sloping terrain
downstream.

WŽƚĞŶƟĂůƐŽƵƌĐĞƐŽĨŽdžLJŐĞŶƚŽƚŚĞƉŝƚƐ;WĂƉĞƌ//Ϳ
The geochemical results suggested that the reduction of oxygen transport into the waste
was not as effective as could be expected from the remediation efforts. Sources of oxygen access to
the waste were investigated with GPR and DC resistivity. Two possible pathways for oxygen into
the waste can be recognised: (1) vertical diffusion through the dry cover, and (2) ingress through
fractures in the unsaturated zone or fractures ¿lled with oxygenated water. The geophysical
study therefore focused on imaging the integrity of the sealing layer in the dry cover, as major
deteriorations may cause its inef¿ciency as a barrier against diffusion of oxygen. Another focus
was the evaluation of the possibility for fractures in the bedrock to supply oxygen to the waste. The
GPR pro¿les of high enough quality (the ones performed on pit 2) showed no major disturbance in
the sealing layer. The geophysical data showed, however, that in the long term there may be a risk
of deterioration of the sealing layer due to frost action in zones of thinner protective layer, and also
due to seepage of water from the back¿lled waste to the surface through the dry cover (seeping
effect shown on Figure 8). Archive data from the site indicates that the bedrock close to the pits
is fractured. Resistivity pro¿les showed possible ingress of water to the waste through the high
wall of pit 1. Additionally, groundwater level measurements indicate that a large part of the pit is
not saturated with water. Therefore, there is good ground to think that oxygen and/or oxygenated
water have access to the reactive waste through fractures in the pit walls. As no evident sign of
destruction of the sealing layer could be found, oxygen transport through fractures would be the
most likely reason for on-going sulphide oxidation. A scheme of suspected oxygen and water
pathways through the pits is shown in Figure 9.
O2

OXGENATED
WATER
WASH-OUT OF
OXIDATION PRODUCTS

ON-GOING SULPHIDE OXIDATION


IN WASTE ROCK

Figure 9 Suspected pathways of water and oxygen through the back¿lled waste at Kimheden, using a cross-section of
back¿lled open pit 1 as an illustration.

14
EĞĞĚĨŽƌĨƵƌƚŚĞƌŵŝƟŐĂƟŽŶŵĞĂƐƵƌĞƐ;WĂƉĞƌƐ/ĂŶĚ//Ϳ
Stabilisation of the metal concentrations in the receiving stream since 2002 suggests that
water quality in the mine drainage is not improving further. The current water quality is judged
insuf¿cient for discharge into the natural environment (Boliden, personal communication). Table 3
displays the average concentrations of Cu, Zn, Cd, Pb, Cr, Ni and As classi¿ed in accordance with
the Swedish EPA criteria (SEPA, 2000). Concentrations of Cu, Zn and to a lesser extent Cd are
judged too high considering biological effects. Concentrations of Al (up to 20 mg/L in the stream
in 2009) are not included in this metal classi¿cation, but are still subject of concern at Kimheden.
pH values are also indicated, and are at least 0.9 pH unit lower than in the background water (SB,
4.6). These results are further described in Paper I.

Table 3 Average dissolved concentrations in —g/L of selected metals at the main sampling locations during the 2009
sampling programme. SB is background surface water; G1 is groundwater in back¿lled open pit 1; S1 is seepage from
open pit 1; S3 is seepage from open pit 2; SD is surface water downstream (cf. Figure 3). pH is shown as well. The
concentrations are classi¿ed according to the Swedish EPA criteria (SEPA, 2000). For estimation of the average As
concentrations, values below the detection limit were calculated as half the detection limit.

10 Very low concentration


10 Low concentration
10 High concentration
10 Very high concentration

The different classes are distinguished by the Swedish Environmental Protection Agency
according to biological effects.

—g/L Cu Zn Cd Pb Cr Ni As pH
SB 2.3 7.2 0.013 0.14 0.26 0.38 0.13 4.6
G1 790 410 0.58 0.73 1.0 9.5 0.32 3.7
S1 380 88 0.16 0.68 0.33 4.83 0.24 3.5
S3 1600 450 1.00 0.33 2.8 9.4 0.23 3.0
SD 400 120 0.30 0.96 0.72 3.9 0.058 3.7

Therefore, sustainable additional remedial actions of the mine waste should be considered.
Fractures in the pit walls are expected to be the main reason for the limited success of mitigation
measures (Paper II). Sealing of fractures in the pit walls seems to be an irrelevant solution, since
back¿lling and capping are completed. Nevertheless, redirection of the groundwater pathways
around the pits to avoid water Àow through the back¿lled waste has been recognised as a possible
method to reduce the discharge of contaminants from back¿lled pits (Tremblay & Hogan, 2001).
This is usually achieved by the establishment of a permeable zone (= “pervious surround”) around
the more impermeable back¿lled waste, as this approach is typically used with back¿lling of
tailings which are much less permeable than waste rock. The same type of result would be expected
in our case, if lower permeability could be achieved in the waste rock. In-situ solidi¿cation of the
back¿lled waste rock by e.g. injection could therefore be a solution at Kimheden.

ŽŶĐůƵƐŝŽŶƐ
In-pit back¿lling and sealing of waste rock at Kimheden resulted in a quick and maintained
decrease of concentrations of the target metals Cu and Zn in the mine drainage. However, quick
stabilisation of Cu, Zn and other metals also implies that their decrease soon reached a plateau,
at values that are still considered too high for discharge in the natural environment. In 2009,
concentrations of Cu, Al and Zn in the collection ditch below back¿lled pit 1 were 3 mg/L, 20
mg/L, and 0.3 mg/L, respectively. Besides, pH in the receiving stream is still relatively low [3.0-
3.7] compared to the background pH of 4.6, indicating that acid generation from the mine is still
going on.

15
Limited success of the mitigation measures at Kimheden is attributed to the continued
access of oxygen to the back¿lled waste in spite of the dry cover, and leaching of the oxidation
products. Geophysical investigations suggested that transport of oxygen to the waste might occur
through fractures in the pit walls, as oxygen gas or oxygenated water. Additional mitigation
measures could therefore involve decreasing the permeability of the back¿lled waste, in order to
divert most of the water and oxygen in the bedrock around the waste.

&ƵƚƵƌĞƐƚƵĚŝĞƐ
As on-going oxidation in the back¿lled waste is suspected from the results of the presented
studies, questions to be answered as a next step are:
‡ What is the extent of the sulphide oxidation in the back¿lled waste"
‡ How much acidity is contributed from the sulphide oxidation, compared to other
acid producing processes such as precipitation of iron in the stream"
One main process expected to generate acidity is the precipitation of Fe hydroxides from the
oxidised dissolved Fe in the mine drainage. Therefore, investigations of the speciation of Fe through
measurement of Fe(II) and Fe(III) concentrations in the water, and particulate concentrations of Fe
have been carried out. Redox potential calculations and speciation modelling are expected to give
information on the proportion of contribution from both sulphide oxidation and iron precipitation
to the acid generation.
Another important issue is to better characterise the current pathways of water Àowing to
and out from the pits, in order to validate or invalidate the theory of fractures in the pit walls as
a source of oxygenated water to the waste, and evaluate the adequacy of diverting groundwater
Àows around the pits as a future mitigation action.
Collaboration studies of waste rock leaching properties in different conditions (untreated
waste and addition of alkaline material in laboratory and meso-scale) are also planned, in order
to increase knowledge about mechanisms of contaminant release from saturated waste rock, and
possible alkaline treatment options.
The ¿ndings will be used to construct an integrated hydro-geochemical model of the
AMD production and release from the back¿lled waste rock, using Kimheden as a reference, but
attempting to serve as a support for other sites to be back¿lled.

16
Z&ZE^
Årebäck, H., Barrett, T. J., Abrahamsson, S., & Fagerström, P. (2005). The Palaeoproterozoic
Kristineberg VMS deposit, Skellefte district, northern Sweden, part I: geology. Mineralium
Deposita, 40(4), 351-367.
Axelsson, C.-L., Ekstav, A., Holmén, J., & Jansson, T. (1991). Efterbehandling av
sandmagasin i Kristineberg, Hydrogeologiska förutsättningar för åtgärdsplan: Lakvattenbalanser
och vittringsbegränsande åtgärder. Golder Geosystem AB. In Swedish.
Brady, K. B. C., Smith, M. W., & Schueck, J. (Eds.) (1998). Coal mine drainage prediction
and pollution prevention in Pennsylvania. Harrisburg, Pennsylvania: Department of Environmental
Protection.
Chapman, J., Hockley, D., Sevick, J., Dachel, R. & Paul, M. (1998). Pit back¿lling on two
continents: Comparison of recent experiences in the Wismut and Flambeau projects. In: Tailings
and mine waste ’98, International Conference on Tailings and Mine Waste (pp.55-65). Rotterdam:
Balkema.
Dahlin, T., Rosqvist, H., & Leroux, V. (2010). Resistivity-IP mapping for land¿ll
applications. Near Surface Geoscience. 28(8), 101-105.
European Commission, DG Environment (2007). Classi¿cation of mining waste facilities.
Gray, T. A., & Gray R. E. (1992). Mine closure, sealing and abandonment. In H. L. Hartman
(Ed.), SME mining engineering handbook (pp.659-674). Vol. 1. (2. ed.). Englewood, Colorado:
Society for Mining, Metallurgy, and Exploration.
Hannington, M. D., Kjarsgaard, I. M., Galley, A. G., & Taylor, B. (2003). Mineral-chemical
studies of metamorphosed hydrothermal alteration in the Kristineberg volcanogenic massive
sul¿de district, Sweden. Mineralium Deposita, 38(4), 423-442.
Hellman H., & Lokrantz H. (2008). Kimhedengruvan, Installation av grundvattenrör vid
nedlagt och igenfyllt dagbrott vid Kimheden. Berggeologiska Undersökningar AB. In Swedish.
Höglund, L. O., & Herbert, R. (Eds.), Lövgren, L., Öhlander, B., Neretnieks, I., Moreno,
L., Malmström, M., Elander, P., Lindvall, M., & Lindström, B. (2004). MiMi - Performance
assessment: main report. MiMi report 2003:3, Mitigation of the environmental impact from mining
waste programme (MiMi). Luleå, Sweden: MiMi Print.
INAP (the International Network for Acid Prevention) (2009). Global Acid Rock Drainage
Guide (GARD Guide). http://www.gardguide.com/.
Johnson, D. B., & Hallberg, K. B. (2005). Acid mine drainage remediation options: a
review. Science of the Total Environment, 338(1-2), 3-14.
Loke, M. H., Acworth, I., & Dahlin, T. (2003). A comparison of smooth and blocky
inversion methods in 2D electrical imaging surveys. Exploration Geophysics 34(3), 182-187.
Lottermoser, B. G. (2003). Mine wastes: characterization, treatment and environmental
impacts. Berlin, Heidelberg: Springer-Verlag.
Malmström, M., Werner, K., Salmon, U. & Berglund, S. (2001). Hydrogeology and
geochemistry of mill tailings impoundment 1, Kristineberg, Sweden: Compilation and interpretation
of pre-remediation data. MiMi report 2001:4, Mitigation of the environmental impact from mining
waste programme (MiMi). Luleå, Sweden: MiMi Print.

17
MEND (Mine Environment Neutral Drainage programme) (1997). Review of water cover
sites and research projects, MEND 2.18.1.
Mitchell P. (1999). Prediction, prevention, control, and treatment of acid rock drainage.
In: L. Noronha. & A. Warhurst (Eds.), Environmental policy in mining: corporate strategy and
planning for closure (pp.117-143). Boca Raton, Florida: Lewis.
Phelps, L. B. (1990). Unit operations of reclamation. In: B. A. Kennedy (Ed.), Surface
mining (pp.770-776). (2. ed.). Littleton, Colorado: Society for Mining, Metallurgy, and Exploration.
Rosén, L., & Wilske, Å. (1994). Beräkningar av metallÀöden och förslag till åtgärder vid
Kimhedengruvan, Kristineberg. Scandiaconsult Väst AB. In Swedish.
SEPA (Swedish Environmental Protection Agency) (1993). Gruvavfall från
sul¿dmalmsbrytning - metaller och surt vatten på drift -. Report 4202. Stockholm: Naturvårdsverket
Förlag. In Swedish.
SEPA (Swedish Environmental Protection Agency) (2000). Environmental quality criteria
- lakes and watercourses. Report 5050. Kalmar: Lenanders.
Tremblay, G. A., & Hogan, C. M. (Eds.) (2001). MEND Manual, Volume 4 - Prevention
and control, MEND 5.4.2d, Mine Environment Neutral Drainage programme (MEND).
Wilson, G. W., Williams, D. J., & Rykaart, E. M. (2003). The integrity of cover systems
- An update. In: ICARD 2003 - proceedings of the Sixth International Conference on Acid Rock
Drainage, Cairns, Australia, pp.445-451.
Wilson, G. W. (2008). Why are we still struggling with Acid Rock Drainage" Geotechnical
News, 26(2): 51-56.
Wireman, M. (2001). Potential water-quality impact, of hard-rock mining. Ground Water
Monitoring and Remediation, 21(3), 40-+.
Younger, P. L., Banwart, S. A., & Hedin, R. S. (2002). Mine water: hydrology, pollution,
remediation. Dordrecht, The Netherlands: Kluwer Academic.

18
I

(IIHFWVRIEDFNÀOOLQJDQGVHDOLQJRIZDVWHURFNRQZDWHUTXDOLW\
DW.LPKHGHQRSHQSLWPLQHQRUWKHUQ6ZHGHQ

Lucile Villain, Lena Alakangas, Björn Öhlander


īĞĐƚƐŽĨďĂĐŬĮůůŝŶŐĂŶĚƐĞĂůŝŶŐŽĨǁĂƐƚĞƌŽĐŬŽŶǁĂƚĞƌƋƵĂůŝƚLJĂƚ
<ŝŵŚĞĚĞŶŽƉĞŶͲƉŝƚŵŝŶĞ͕ŶŽƌƚŚĞƌŶ^ǁĞĚĞŶ
Lucile Villain1, Lena Alakangas1, Björn Öhlander1

1
Luleå University of Technology, Division of Geosciences and Environmental Engineering, 971 87 Luleå, Sweden

ďƐƚƌĂĐƚ
Evaluating water quality at reclaimed mines impacted by acid mine drainage is an essential
step to assess and improve the performance of mitigation methods. At Kimheden copper mine in
northern Sweden, reclamation involved progressive back¿lling of waste rock into the two small
open pits, and application in 1996 of a till dry cover including a sealing layer. Long-term water
quality monitoring by the mining company, and repeated surface and ground-water samplings in
2009-2010, were used to assess the success of the reclamation in mitigating the acid mine drainage
production from the mine waste.
The substantial decrease of concentrations of copper and zinc added together by more than
87% in the receiving stream suggested successful results for the reclamation as a ¿rst approach.
However, the decrease was followed by a quick stabilisation of metal concentrations at values that
are still not satisfactory for discharge of the mine drainage in the natural environment. Seepage
from one of the pits in 2009 had copper, aluminium and zinc concentrations of 3 mg/L, 20 mg/L and
0.3 mg/L, respectively, and a pH of 3.0. The mixed outcome of the mitigation actions is suspected
to be due to on-going oxidation in the back¿lled waste rock, in spite of the dry cover. Besides,
stream discharge measurements and natural tracers in the drainage showed that water management
at the site is not appropriate, whereby a large part of the contaminated drainage disappears as
groundwater, while water diverted to a treatment pond is mostly contributed by natural water.

/ŶƚƌŽĚƵĐƟŽŶ
After closure of surface mines, there are several options to handle the residues from mining
operations and to restore the landscape. In-pit back¿lling of waste rock is one option, which usually
requires the application of additional mitigation measures like capping or Àooding, in order to
reduce the release of contaminants from the waste. Waste rock back¿lling is the preferred method
at surface coal mines, where the voids are usually large and shallow, facilitating the placement
of the voluminous mined-out material (Phelps, 1990). At open-pit sites, however, this method is
commonly not economically viable, and most of the open pits will be left as mine voids or evolve
as pit lakes. In certain conditions, if the ore was mined close to the surface and the waste rock
piles are situated close enough to the pit, back¿lling can be considered for open pits as well (Gray
& Gray, 1992; Younger et al., 2002). Positive effects of this method have been recognised by
Chapman et al. (1998). They state that back¿lling open pits may help, in particular, minimising the
area of waste to be managed, offering a simple and safe containment for the residues, and reducing
the extent of the contaminated area. Nevertheless, even after back¿lling, additional mitigation
actions may have to be taken for a successful reclamation. If the back¿lled waste rock is sulphidic
and the pit is not Àooded naturally, access of water and oxygen to the rocks involves a risk of
oxidation and subsequent production of Acid Mine Drainage (AMD), which is a major threat to the
surrounding surface and ground-water. Solutions to prevent and mitigate contamination from the
mine back¿ll have been suggested both in the case of coal mining (e.g. Reed & Singh, 1986; Brady
et al., 1998) and open-pit mining (e.g. Chapman et al., 1998; Tremblay & Hogan, 2001). The

1
most common options are: water management (e.g. diversion of surface water, Àooding), selective
placement of the waste in the back¿ll according to its contamination potential, alkali injection in
the back¿ll, and surface reclamation by application of a cover preferably including a sealing layer.
These solutions have already been largely implemented during the last decades, but their actual
success at mine sites remains highly debated. Despite the logical advantage of understanding the
performance of mitigation actions in practise, studies at reclaimed sites are rather scarce in the
scienti¿c literature, in particular with regards to effects on mine water chemistry. This lack has
also been recognised by e.g. Runkel et al. (2009), in their post-remediation study of a catchment
impacted by mining activities.
At Kimheden copper mine in northern Sweden, the effects of back¿lling and sealing of
oxidised waste rock in two open pits on drainage quality after fourteen years was investigated. The
mine was operated during a few years in the 1970s, when copper was extracted by the Swedish
mining company Boliden AB. Soon after cessation of the operations, copper- and zinc-rich AMD
was identi¿ed, originating from the waste rock left on the surface. Several remedial attempts
followed, until extensive reclamation was completed in 1995-1996. Oxidised waste rock was fully
back¿lled into the two small open pits of the mine, and a composite sealing-protective till cover
was applied on top. Such type of dry cover including a low permeability barrier is largely used in
Sweden (Alakangas et al., 2005). Back¿lling and sealing of the mine waste intended to exclude the
transport of oxygen to the rocks, thus limiting the oxidation of sulphides. The mining company has
been monitoring water quality downstream of the mine annually since 1983, and also close to the
mine workings at several occasions. The data was complemented by results from repeated surface
and ground-water samplings in 2009-2010 to evaluate the effects of the mitigation actions on water
geochemistry. The objective was to determine if the decrease of metal concentrations and acidity
in the stream after reclamation met the expectations from the mitigation actions performed, both
in order to evaluate the need for further remediation at Kimheden, and to improve the performance
of applications at other sites.

^ƚƵĚLJĂƌĞĂ

>ŽĐĂƟŽŶ͕ĐůŝŵĂƚĞĂŶĚŐĞŽůŽŐLJ
Kimheden copper mine is situated in the Kristineberg mining area in northern Sweden
(Figure 1). The region is characterised by a continental inland climate with cold winters and warm
summers. The snow-covered period extends from October to May. Annual precipitation at the site
is ca. 400-800 mm (Axelsson et al., 1991), and annual average temperature is 0.7 ºC (Malmström,
2001). The summer of 2009, when most of the sampling was performed, was rather dry. Therefore,
some of the sampling locations were dried out at several occasions, and could not be sampled.
The Kristineberg area is a deformed and metamorphosed Palaeoproterozoic volcanic
domain. Kimheden is one of the smaller pyrite-rich massive sulphide deposits of the area,
which are intercalated within a succession of altered felsic volcaniclastic rocks that have been
metamorphosed to quartz-muscovite-chlorite schists (Hannington et al., 2003). 0.13 Mt ore was
extracted at Kimheden with 0.95% Cu and 0.27% Zn (Årebäck et al., 2005). Common ore minerals
are pyrite and chalcopyrite along with some sphalerite, while gangue minerals are mostly quartz,
sericite and chlorite with traces of talc. Calcareous rocks are rare.

2
ARVIDSJAUR

ABBORTRÄSK

GLOMMERSTRÄSK

MALÅ

KIMHEDEN,
KRISTINEBERG

20 km

Luleå

Skellefteå

Stockholm

Figure 1 (left) Location of Kimheden site on the map of Sweden; (right) the study area, Kimheden copper mine, and
the receiving stream that was sampled. The sampling locations are indicated (‘S’ surface water, ‘G’ groundwater). SB
and GB are background surface and ground-water, respectively.

ZĞĐůĂŵĂƟŽŶŽĨƚŚĞƐŝƚĞ
Copper ore at Kimheden was mined in the 1970s, both underground and in two small open
pits located on the slope of a hill (Figure 1). Sulphidic waste rock was directly deposited on the
surface close to the pits. Drainage from the spoil rapidly became acidic, with high copper and zinc
concentrations, revealing the production of AMD by oxidation of sulphide minerals in contact with
water and oxygen. To mitigate the contamination of surrounding natural watercourses and to limit
further oxidation, a reclamation plan was adopted (Andersson, 1988). Water management started
in 1982 with the excavation of ditches, both upstream of the mine waste to divert water run-off,
and downstream of the contaminated areas to transport the mine drainage to a limed tailings pond.
The collection ditch to the treatment pond corresponds to the receiving stream that was sampled
in the present study (Figure 1). The diversion of surface water was followed by several attempts
to back¿ll waste rock in the two open pits and to lime the area (1984-1985 and 1988-1989),
ending in 1995-1996 with the most extensive remedial event (Boliden, 1995). All waste rock left
on the surface and material from the former industrial area were back¿lled into the open pits,
and a composite dry cover including a low permeability barrier was applied on the waste. The
dry cover consisted of a 0.3 m sealing layer made of clayey till, overlain by a 1.5 m protective
layer of unsorted till. As the topography prevented the establishment of a pit lake, the dry cover
was expected to reduce the transport of oxygen to the waste, and thereby the production of AMD.
The plan was to treat the mine drainage until suf¿cient quality would be achieved. The site also
comprises a shaft of 400 m depth. Information on its rehabilitation is almost inexistent, but some
waste rock may have been back¿lled there before capping with a hydrological seal.

3
DĞƚŚŽĚƐ

^ĂŵƉůŝŶŐĂŶĚĂŶĂůLJƟĐĂůŵĞƚŚŽĚƐ
The post-reclamation sampling programme at Kimheden was performed in eight sessions
spaced out by 2-3 weeks from May to October 2009, followed by one more session in August 2010.
Water samples were collected each time over 1 to 2 days, at 2 to 14 sampling locations (Figure 1).
Surface water was sampled in the receiving stream from the mine (= collection ditch), and at one
location upstream of the mine (SB= background water). SD and 6301 are sampling locations at
the outlet of the stream, which were also used for long-term monitoring by the mining company.
Groundwater was sampled from a groundwater well installed in the back¿lled waste of open pit 1
(G1). The water was pumped continuously through a PVC tube that was previously acid-washed
with 5% nitric acid. Sampling started when electrical conductivity (EC), temperature and redox
were stabilised (after about 30 to 50 min). Background groundwater was also sampled, using an
old exploration drill casing upstream of the mine (GB).

Water sampling methods


pH and temperature were measured on-site with a Metrohm 704 portable pH meter and EC
with a WTW Multi 350i multimeter. Filtered water was sampled for analysis of dissolved elements.
The water was directly ¿ltered through a 0.22 —m nitrocellulose membrane into two 125 mL plastic
bottles. The bottles were previously washed in 50% hydrochloric acid followed by 1% nitric acid;
the syringes were washed in 5% nitric acid, and the ¿lters in 5% acetic acid. One bottle was sent
for analysis to the ALS Scandinavia accredited laboratory, while the duplicate was preserved in a
freezer in case additional results were needed. Suspended particles on ¿lters were also analysed
on the four ¿rst sampling occasions. However, the concentrations of particulate elements were
systematically very low; therefore, the ¿lters from the following sampling sessions were preserved
in freezer for potential later analysis. Un¿ltered water was sampled in 100 mL plastic bottles
for measurements of sulphate, dissolved organic carbon (or total organic carbon) and acidity at
the commercial laboratory. Alkalinity was lower than the detection limit at all locations and was
therefore not measured. The bottles of un¿ltered water were rinsed twice with water from the
sampled location before actual sampling. Laboratory and ¿eld blank samples were occasionally
collected and analysed to ensure the reliability of the analytical results. Water samples were kept
cool in ice bags and placed in a fridge or freezer within a couple of hours.
Long-term water sampling by the mining company was not part of this sampling programme.
They used acid-washed plastic bottles for sampling as well, but they did not ¿lter the water, and
total concentrations of elements were therefore determined.

nalLJƟĐal methods
At the laboratory, Ca, K, Mg, Na, S and Si were analysed with ICP-AES, and Fe, Al, As,
Ba, Cd, Co, Cr, Cu, Mn, Mo, Ni, P, Pb, Sb, Sr, U, V and Zn with ICP-SFMS. The ICP-AES and
ICP-SFMS analyses were carried out according to US EPA Methods 200.7 (modi¿ed) and 200.8
(modi¿ed), respectively. Sulphate was analysed by liquid chromatography of ions. Dissolved
organic carbon and total organic carbon were measured by IR-detection after thermal oxidation.
Total acidity was measured by titration with sodium hydroxide and the phenolphthalein indicator
to pH 8.3.

^tream disĐharge measƵrements


Estimation of the loads of elements was performed using water discharge measured at each
sampling location in the stream. At one location (6301, Figure 1), discharge was measured with

4
a 90º Thomson V-notch weir. At other locations, discharge was measured either with the bucket
and stopwatch method if all the Àow could be concentrated, or with the Àoater method otherwise.
Discharge measured with the V-notch weir was read from the corresponding height-discharge
equivalence table. As for the bucket and stopwatch method (or volumetric method), discharge is
simply calculated by the following formula:
V
Q (1)
t
where Q is the discharge in L/s, V is the volume of water ¿lled in the bucket in L, and t is time
in s. To reduce measurement errors, water was ¿lled during minimum 3 seconds except for high
Àows, and the operation was repeated at least 5 times. Stream discharge estimated from the Àoater
method was calculated on a short stream portion from the formula (Gordon, 2004):
L
Q=k A (2)
t
where Q is the discharge in m3/s, L is the stream portion length in m, t is time in s, A is the cross-
sectional area of the Àowing water in m2, and k is a correction factor (0.8) between surface velocity
and average velocity of water in the stream.

ĂůĐƵůĂƟŽŶŽĨƚŚĞƉƌŽƉŽƌƟŽŶŽĨDŝŶƚŚĞƐƚƌĞĂŵǁŝƚŚŶĂƚƵƌĂůƚƌĂĐĞƌƐ
In many hydro-geochemical studies (e.g. Hooper et al., 1990; Bazemore et al., 1994; Land
et al. 2000; Joerin et al., 2002) concentrations of natural elements have been used to estimate the
contribution of different water sources to the discharge of a watercourse at different locations. The
necessary conditions are that the selected chemical elements should have signi¿cantly contrasting
concentrations at the sources (end-members), and that they behave conservatively during their
transport. In the particular case when two end-members are identi¿ed, only one tracer is necessary,
and the equation used to calculate the proportions of the end-members is (Land et al., 2000):

C C
m j
X i
=
C C
i j

where i and j represent the two end-members, and m represent the mixture. Xi is the proportion
of the end-member i in the mixture and Ci is the concentration of the tracer in the end-member
i. In the case of a mixture from two end-members, the conservativeness of a tracer in the stream
can be evidenced if, when its concentrations along the stream are plotted against those of another
conservative tracer, a straight line is formed. The end-members are then represented on the
extremities of the mixing line.

ZĞƐƵůƚƐĂŶĚŝŶƚĞƌƉƌĞƚĂƟŽŶ

ĂƌůLJƌĞĐůĂŵĂƟŽŶǁĂƚĞƌƋƵĂůŝƚLJǀĞƌƐƵƐƉŽƐƚͲƌĞĐůĂŵĂƟŽŶǁĂƚĞƌƋƵĂůŝƚLJ
At Kimheden, Cu and Zn were identi¿ed by the mining company as the most hazardous
metals in the mine drainage before reclamation, and were therefore selected for long-term
monitoring, in addition to pH. Only in recent years, other elements were added for analysis, and
their results are presented in the next section. Figure 2 shows the total concentrations of Cu and
Zn and pH values measured since 1983 once every year at the end of the receiving stream. The
reclamation stages are indicated on the time scale. Sampling was performed at location 6301
between 1983 and 1994, and at SD between 1999 and 2010 (Figure 1). Similar water quality was
observed at those two locations in 2009, suggesting that results from the two time periods might
be compared despite of the change of monitoring location.

5
Figure 2 shows a substantial decrease of Cu and Zn concentrations between the early
reclamation and post-reclamation monitoring periods. Application of the dry cover in 1996 seems
to be the key event explaining the decrease. The increase in pH is less distinct, with pH values after
application of the cover that are slightly higher than before.

  

     


 

   

  






 
  

25 5
Cu (mg/L)
pH

20 Zn (mg/L) 4

15 3
Cu
Zn
pH
10 2

5 1

0 0
83

85

87

89

91

93

95

97

99

01

03

05

07

09
19

19

19

19

19

19

19

19

19

20

20

20

20

20
Year

Figure 2 Total concentrations of Cu and Zn and pH values at the end of the receiving stream, between 1983 and 2010.
The main remedial events are indicated in chronological order. Data was provided by the mining company. Unknown
data is represented in dashes.
In addition to the annual monitoring at the end of the stream, sampling at several other
locations of the stream was performed by the mining company at four occasions between 1987 and
1991. The results are reported in Figure 3, together with results obtained in 2009. The concentrations
of Cu and Zn in 1987-1991 are total concentrations, while concentrations in 2009 are dissolved
ones. Nevertheless, as mentioned in the methods, the suspended phase is insigni¿cant compared
to the dissolved phase. Thus, Cu and Zn occur almost totally in the dissolved phase. Therefore,
comparison between 1987-1991 and 2009 is possible.
Figure 3.a shows that pH values have increased at all sampling locations since the beginning
of reclamation. The increment between 1987 (earliest date) and 2009 is highest at S1 (seepage
from open pit 1, +0.9) and at SD (downstream, +0.7). The reference pH from background surface
water (4.6) shows that pH in the receiving stream in 2009 (3-3.7) is still low compared to natural
pH. Figures 3.b and 3.c indicate that the decrease of Cu and Zn concentrations since the beginning
of reclamation is similar at all locations, though for Cu, the decrease at S1 is higher than at the
other locations. No relevant explanation has been found for the high concentrations of Cu and Zn
at S3 in 1989 and the low pH values at all locations in 1991. These values are therefore considered
as outliers.
In 1991, a year-round monitoring of the water quality was performed along the receiving
stream. At that time, the results were used to formulate an environmental goal for the completion
of the reclamation, and decide upon the best strategy to use, which ended up being the back¿lling-
sealing option in 1995-1996. In fact, two objectives were suggested for the reclamation. The
annual average concentrations of Cu and Zn added together should be reduced to 1 mg/L either at
6301 ± later SD ± (objective 1), or at S1 (objective 2) (Rosén & Wilske, 1994). This would imply
a decrease of the concentrations of Cu+Zn of 87% (objective 1), or 94% (objective 2) along the
stream. The resulting concentrations of Cu and Zn with objective 1 were estimated at three key
locations: S1 seepage from open pit 1, S3 seepage from open pit 2, and 6301-SD downstream. The
actual decrease of concentrations of Cu and Zn in 2009 was compared to the decrease estimated
with objective 1 in Table 1.

6
5,5 (3.a) Figure 3 (3.a) pH, (3.b)
5
Cu concentrations and
Background water 2009
(3.c) Zn concentrations
4,5
1987-10-30 in the receiving stream at
4 1989-11-17 5 sampling occasions in
1990-05-18 1987, 1989, 1990, 1991
pH

3,5
1991-08-05
and 2009. The sampling
3 2009-09-02
locations can be found in
2,5 Bars show the
pH increase
Figure 1. The vertical bars
2
between show the increase in pH
1987 and 2009
(3.a), and the decrease in
1,5
S1 S1a S1c S2 S3a S3 S4 SD
Cu and Zn (3.b and 3.c),
Sampling locations between 1987 and 2009.
1000 The pH of background
(3.b) surface water measured
in 2009 is shown as a
100
horizontal line on Figure
1987-10-30
3.a. Data between 1987
Cu (mg/L)

1989-11-17
10 1990-05-18
and 1991 was provided by
1991-08-05 the mining company. No
2009-09-02 data was available for S1a
1 Bars show the in 1991. pH is shown on a
decrease in Cu
concentrations
linear scale to emphasize
between the increment between each
0,1
S1 S1a S1c S2 S3a S3 S4 SD
1987 and 2009
date, while concentrations
Sampling locations
of Cu and Zn are showed
100 on a logarithmic scale to
(3.c) emphasize the factor of
decrease between each
10
sampling date.
1987-10-30
Zn (mg/L)

1989-11-17
1 1990-05-18
1991-08-05
2009-09-02

0,1 Bars show the


decrease in Zn
concentrations
between
0,01 1987 and 2009
S1 S1a S1c S2 S3a S3 S4 SD
Sampling locations

Table 1 Average concentrations of Cu, Zn and Cu+Zn, at S1 (seepage from open pit 1), S3 (seepage from open pit 2)
and 6301-SD (downstream). Values are shown for 1991, according to objective 1 (Cu+Zn< 1mg/L at 6301-SD), and
for 2009. The reduction factor and percent decrease from 1991 are also indicated. Figure modi¿ed from Rosén and
Wilske (1994).

Cu Zn Cu+Zn
Reduction factor Percent decrease (%)
(mg/L) (mg/L) (mg/L)
Cu Zn Cu+Zn Cu Zn Cu+Zn
1991
S1 15.24 0.51 15.75
S3 16.10 1.95 18.05
6301-SD 7.04 0.59 7.63
Objective
(1) S1 2.00 0.07 2.06 8 8 8 87 87 87
S3 2.11 0.26 2.37 8 8 8 87 87 87
6301-SD 0.92 0.08 1.00 8 8 8 87 87 87
2009
S1 0.36 0.12 0.47 43 4 33 98 77 97
S3 1.60 0.45 2.06 10 4 9 90 77 89
6301-SD 0.36 0.11 0.47 20 5 16 95 81 94

7
The decrease of concentrations of Cu+Zn between 1991 and 2009 shows that objective 1
has been achieved, since all key locations have their Cu+Zn concentrations reduced by more than
87%. Objective 2 of decrease of 94% was achieved at S1 and SD, yet not at S3. These results tend
to demonstrate a successful rehabilitation as a ¿rst approach.

ǀŽůƵƟŽŶŽĨǁĂƚĞƌĐŽŵƉŽƐŝƟŽŶĚŽǁŶƐƚƌĞĂŵŽĨƚŚĞŵŝŶĞƐŝŶĐĞϮϬϬϮ
Since 2002, sampling by the mining company downstream of the mine has been extended
to twice annually. Besides, EC and concentrations of several elements have been included in the
monitoring routine in addition to pH, Cu and Zn. Figure 4 shows selected parameters.

1000 (4.a) Figure 4 (4.a) Total


concentrations of Cu,
Zn, Fe, S and Al; (4.b)
Total concentrations
100
of Ca, Mg, K, and
Cu μg/L EC and pH. The
Concentrations

Zn μg/L samples were taken


10 Fe mg/L at SD downstream
S mg/L
Al mg/L
of the mine, between
2002 and 2009. Data
1 was provided by the
mining company.
01

15

09

10

09

09

07

06

02

04

05

02

06

08
2

0
0-

4-

5-

9-

5-

9-

5-

9-

5-

9-

5-

9-

5-

9-

5-

9-
pH and EC values in
-1

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0
02

03

03

03

04

04

05

05

06

06

07

07

08

08

09

09
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
0,1
2008 are not known.
Sampling dates (2002-2009)
(4.b)
1000 10

100

Ca mg/L
Concentrations

Mg mg/L
pH

10 K mg/L
(&ȝ6FP
pH

1
01

15

09

09

09

07

03

06

02

04

05

06

08
20 4-2

20 5-1

-0
0-

5-

9-

9-

5-

5-

9-

9-

5-

5-

9-
9

9
-1

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0

-0
02

03

03

03

04

04

05

05

06

06

07

07

08

08

09

09
20

20

20

20

20

20

20

20

20

20

20

20

20

20

0,1 1
Sampling dates (2002-2009)

All parameters sampled since 2002 (concentrations of Ca, Fe, K, Mg, Na, S, Al, Ba, Cd,
Co, Cr, Cu, Mn, Ni, Pb, Zn; EC and pH values) suggest a stabilisation. No increase or decrease
has been occurring since then. The improvement of water quality occurring after reclamation (cf.
previous section) was therefore rapid and reached quick equilibrium.

īĞĐƚƐŽĨƚŚĞƌĞĐůĂŵĂƟŽŶŽŶƉLJƌŝƚĞŽdžŝĚĂƟŽŶƌĂƚĞ
Sulphate concentrations measured at several occasions in 1983 downstream of the mine
were found in archive documents from the site. Assuming that pyrite oxidation still proceeds in
the waste, though at a much lower rate that before reclamation, it was attempted to calculate the
decrease in pyrite oxidation rate following reclamation. Recent unpublished iron speciation data
at the mine supports this assumption and suggests that oxidative conditions can still be found in
the back¿lled waste, despite of capping. Archive data about the extracted ore indicates that pyrite

8
should be the dominant reactive sulphide found in the waste. Therefore, sulphate concentrations in
the receiving stream are expected to directly relate to pyrite oxidation rate in the back¿lled waste,
according to the following reaction:
FeS2 + 7/2O2 + H2O = Fe2+ + 2SO42- + 2H+ (4)
whereby 2 SO4 are produced when 1 FeS2 is oxidised. The average sulphate concentration
downstream in 1983 (458 mg/L) was compared with the average concentration at the same location
in 2009 (89 mg/L). Assuming that the annual water discharge in the receiving stream was similar
in 1983 and 2009, the oxidation rate of pyrite should have decreased by 80% between 1983 and
2009, which would be mainly explained by the reduction of oxygen transport to the waste as a
result of capping. The decrease is signi¿cant, but it does not meet the performance objective of
94-99% established by the Swedish Environmental Protection Agency for dry covers of multiple
layers (SEPA, 1993).

DŝŶĞǁĂƚĞƌĐŽŵƉŽƐŝƟŽŶŝŶϮϬϬϵ
The pH of water in 2009 was low at all parts of the receiving stream, with average values
ranging from 3.0 to 3.7 (Table 2). Background water (SB and GB) was also relatively acidic
(4.6), but the pH in the stream was 0.9 to 1.6 units lower than in the natural water. The average
concentrations of Cu, Zn, Cd, Pb, Cr, Ni and As are shown in Table 2, where they are classi¿ed in
accordance with the Swedish EPA criteria (SEPA, 2000) relating metal concentrations and biological
effects. According to this classi¿cation, Cu and Zn primarily and Cd to a lesser extent, exhibit
high concentrations in the stream, which are not observed in the background water. Aluminium
is not taken into account in this classi¿cation, but it also presented high average concentrations
in the stream in 2009 (9.6 mg/L at G1, 3.6 mg/L at S1, 4.4 mg/L at S3 and 3.1 mg/L at SD). S1a
downstream of open pit 1 was sampled at one occasion only in 2009, but showed the highest
concentration of Al (19.8 mg/L), and its value in 2010 was also the highest (19.1 mg/L).
Table 2 Average dissolved concentrations in —g/L of selected metals at the main sampling locations during 2009. SB
is background surface water; G1 is groundwater in back¿lled open pit 1; S1 is seepage from open pit 1; S3 is seepage
from open pit 2; SD is surface water downstream (cf. Figure 1). pH is shown as well. The concentrations are classi¿ed
according to the Swedish EPA criteria (SEPA, 2000). To estimate the average concentrations of As, values below the
detection limit were calculated as half the detection limit.
10 Very low concentration The different classes are distinguished by the Swedish
10 Low concentration Environmental Protection Agency according to biological
effects.
10 High concentration
10 Very high concentration

—g/L Cu Zn Cd Pb Cr Ni As pH
SB 2.3 7.2 0.013 0.14 0.26 0.38 0.13 4.6
G1 790 410 0.58 0.73 1.0 9.5 0.32 3.7
S1 380 88 0.16 0.68 0.33 4.83 0.24 3.5
S3 1600 450 1.00 0.33 2.8 9.4 0.23 3.0
SD 400 120 0.30 0.96 0.72 3.9 0.058 3.7

^ƉĂƟĂůǀĂƌŝĂƟŽŶƐŝŶǁĂƚĞƌĐŽŵƉŽƐŝƟŽŶŝŶϮϬϬϵͲϮϬϭϬ
A systematic spatial variation of the concentrations of dissolved elements, EC and acidity
was observed in the receiving stream in 2009 and 2010. This result agrees with the observation
made by Bambic et al. (2006) that spatial distribution of water chemistry in small-scale catchments
can be highly heterogeneous. Dissolved concentrations of S and Mg (elements identi¿ed to be best
related to AMD, cf. section on S-Mg mixing line) as well as EC and acidity of surface and ground-

9
water samples along the stream are shown on Figure 5. They illustrate the spatial variability of the
mine water quality on two selected sampling occasions, one in 2009 (Figure 5.a), and one in 2010
(Figure 5.b), where the largest number of locations could be sampled.
2009-09-02&03 (5.a)
250 18 Figure 5 Spatial

Backfilled pit 2
Backfilled pit 1
EC
16 variations of the mine
S
water composition on
Electrical conductivity (mS/m)

200 14
Mg
(5.a) 2009-09-02&03

Stream discharge (L/s)


Discharge 12
150 and (5.b) 2010-08-

Mg (mg/L)
S (mg/L)

10
16&17, illustrated with
100
8
EC, dissolved S and
6 Mg concentrations,
50 4 and (in 2010) acidity
2 titrated to pH 8.3.
0 0 Stream discharge is also
SB GB G1 S1a S1 S1b S1c S2 S3a S3 S4 SD indicated.
Sampling locations

2010-08-16&17
(5.b)
250 14
EC
Backfilled pit 2
Backfilled pit 1

S 12
Electrical conductivity (mS/m)

Acidity (mmol/L); Mg (mg/L)


200 Acidity

Stream discharge (L/s)


Mg 10
Discharge
150
S (mg/L)

6
100

4
50
2

0 0
SB GB G1 S1a S1 S1b S1c S2 S3a S3 S4 S5 S6 SD
Sampling locations

The spatial variations of the concentrations of S and Mg, EC and acidity exhibit similar
trends. All parameters show a considerable increase from the background water (SB, GB) to the
¿rst back¿lled open pit (G1, S1a). A more limited increase occurs from S2 to the second pit (S3a,
S3), which is followed by a decrease at S5, S6 and SD, downstream of the mine. This suggests,
as could be expected, that the areas close to the pits are the most affected by AMD. Besides, both
pits generate drainage of similar quality. In contrast, uncontaminated background water, and water
downstream, are not (respectively less) inÀuenced by AMD. Lower impact of AMD in the lower
parts of the stream may partly be explained by the effects of dilution by natural water, but might
also be due to the dispersal of contaminated water into the ground on the course of the stream.
These hypotheses were investigated further with the estimation of loads of dissolved elements
transported in the stream.
Loads of S and Mg were calculated from the multiplication of their dissolved concentrations
by the instantaneous stream discharge. The results are shown in Figures 6.a and 6.b for the two
selected sampling occasions in 2009 and 2010. Loads are indicated for both the in-stream sampling
locations and the inÀows from the back¿lled pits to the stream. However, to appreciate the variation
of loads within the stream, only in-stream samples were joined together on a line, whereas inÀow
samples were left as discrete points. The ¿rst location SB is background surface water sampled
upstream of open pit 1, and is therefore not in connection with the receiving stream, which means
that discharges at SB and at S1a coming next are not related. As such, there is in principle no
point in comparing the loads of S and Mg at these locations. Yet, since we are interested in the
contribution of AMD from back¿lled pit 1 to the stream, the concentrations at SB were multiplied
with the stream discharge at S1a for an objective comparison with the loads at S1a.

10
2009-09-02&03
(6.b)
80 10
Figure 6 Spatial variations of

Backfilled pit 2
Bacfilled pit 1

S
70
Mg
9
loads of dissolved S and Mg
8
60 on (6.a) 2009-09-02&03 and

Stream discharge (L/s)


Discharge
50
7
(6.b) 2010-08-16&17. Stream
6
discharge is also indicated. The
S (mg/s)

Mg (mg/s)
40 5
lines join the locations on the
30 4
main course of the receiving
20
3
stream, excluding the inÀows to
10
2
the stream.
1

0 0
SB S1a S1 S1b S1c S2 S3 S4 SD
inflow from backfilled pit 1 Sampling locations inflow from backfilled pit 2

2010-08-16&17
(6.b)
50 6
Backfilled pit 2
Backfilled pit 1

S
45
Mg 5
40

Stream discharge (L/s)


35 Discharge
4
30
S (mg/s)

Mg (mg/s)

25 3

20
2
15

10
1
5

0 0
SB S1a S1 S1b S1c S2 S3a S3 S4 S5 S6 SD
inflow from backfilled pit 1 inflow from backfilled pit 2
Sampling locations

As opposed to the concentrations of S and Mg, the loads of S and Mg indicate that the
contributions of AMD from the two back¿lled pits to the stream are different. Loads of S and
Mg in the stream are highest after back¿lled open pit 1 (S1c). Stream loads after the inÀow from
back¿lled pit 2 (at S4) are always lower. The drop in load may partly be explained by the smaller
inÀow of water from pit 2 compared to pit 1, but not only. Indeed, loads of elements in a stream
should be stable or increase downstream, unless dissolved elements are lost (by precipitation or
sorption), or water disappears as groundwater. The drop in stream discharge (Figures 6.a and 6.b)
between the two pits shows that loss of surface water into the ground is the dominant cause for
the drop in loads. This was con¿rmed by ¿eld observations during the driest sampling sessions,
when the stream was interrupted between the two back¿lled pits. Another sign of mine drainage
Àowing in the ground is the increase in loads of S and Mg between S4 (close to back¿lled pit 2)
and downstream locations (S5, S6 and SD). There is no surface inÀow of water feeding this stream
segment. Therefore, the increase in loads must be explained by contaminated groundwater seeping
out in the stream.
However, relatively high uncertainty in the measurement of water discharge, and risk of
temporal variability of element concentrations and stream discharge over a time span of a sampling
session, can affect the interpretation of loads in terms of sources and sinks of contaminants.
Nevertheless, the two observations presented previously (drop of loads of contaminants between
the two pits and increase in loads between the second pit and downstream locations) were veri¿ed
for all sampling occasions when water discharge data was suf¿cient. Accurate proportion of surface
water disappearing from the stream to groundwater between the two pits could not be obtained
due to uncertainties in discharge measurements. However, a rough estimation was performed, with
the ratio of cumulated discharge of water lost from the stream on cumulated discharge of water
inÀowing to the stream, in the stream section between the two pits. The loss of surface water was
estimated to range from 60% to 100% (at the driest session) of the water inÀows into the stream,
over the 6 sampling events with suf¿cient stream discharge data.

11
dĞŵƉŽƌĂůǀĂƌŝĂƟŽŶŽĨǁĂƚĞƌĐŽŵƉŽƐŝƟŽŶŝŶϮϬϬϵ͖ĚĞƉĞŶĚĞŶĐĞŽŶƐƚƌĞĂŵĚŝƐĐŚĂƌŐĞ
All in-stream sampling locations that were regularly sampled during the 2009 monitoring
programme (SB, GB, G1, S1, S2, S3, S4, SD) showed a temporal variation in ion concentrations,
to a varying extent. Calculation of the coef¿cients of variation (standard deviation divided by the
absolute value of the average) in 2009, indicated that the last sampling point downstream (SD)
had the largest variations. The high variability at SD was most probably the result of variable
contribution from different sources of water, which could be identi¿ed as contaminated water from
the mine waste and background water. A Pearson correlation coef¿cient was calculated between
the dissolved concentration of each element and discharge at SD, for the 7 sampling dates in
2009 and 2010 where instantaneous discharge was measured at SD. A positive and relatively high
correlation coef¿cient was obtained for most of the elements. Concentrations of Ca, Fe, K, Mg, S,
Al, Cd, Co, Cr, Cu, Mn, Ni, U, V and Zn (75% of all the elements measured having concentrations
over the detection limit) had a correlation coef¿cient with discharge above 0.7. Only Na, Si, Ba, Pb
and Sr had a correlation coef¿cient below 0.7. Figure 7 illustrates the variations of concentrations
of Cu, Fe, and S at SD in 2009, with reference to the stream discharge. This suggests that changes
in hydrological conditions imply changes in the source of elements, with higher concentrations at
higher Àows. This relation is explained by an increased amount of drainage water from the mine
area reaching downstream sampling locations in periods of higher Àows (wash-out of the mine),
and an increased proportion from surrounding natural water at periods of lower Àows. Temporal
variation of element concentrations in stream water due to changes in contributions from different
water sources is a phenomenon also observed by Runkel et al. (2009), in their synoptic mass
balance study of a remediated mined-out catchment. At Kimheden, however, this dependence
could only be clearly observed at SD downstream. This may be explained because, with longest
distance in the stream from the mine, the inÀuence of background water is most visible. In other
words, in-stream locations are inÀuenced by an increasing drainage area downstream.
Cu μg/L Fe mg/L
800 30
S mg/L Stream discharge m3/h

700
25
600
Fe, S concentrations
Cu concentrations

20 Stream discharge
500

400 15

300
10
200
5
100

0 0
09-05-13 09-06-04&05 09-06-17&18 09-07-01&02 09-07-22&23 09-08-12&13 09-09-02&03 09-10-22
Sampling dates

Figure 7 Concentrations of Cu, Fe and S and stream discharge vs. sampling sessions in 2009 at SD, downstream of the mine.

^ͲDŐŵŝdžŝŶŐůŝŶĞĂŶĚƉƌŽƉŽƌƟŽŶƐŽĨDŝŶƚŚĞƐƚƌĞĂŵŝŶϮϬϬϵͲϮϬϭϬ
A good tracer of AMD in a stream has to be directly related to the presence of AMD, and
conservative. To ¿nd the best tracers of AMD in the receiving stream, the Pearson correlation
coef¿cient was calculated between the concentrations of dissolved elements assumed to be related
to the production of the mine drainage. It turned out that Mg and S are the elements that are
best correlated, with a coef¿cient of 0.96 for 62 samples, indicating good conservativeness in
the stream. They also show a high correlation to titrated acidity (Pearson correlation coef¿cient
of 0.98 (S) and 0.95 (Mg), for 34 samples), which demonstrates their direct relation with AMD.

12
S is introduced in water as sulphate from the oxidation of sulphidic waste. Mg is produced by
dissolution of aluminosilicates with protons from the acidic drainage (Höglund et al., 2004). At
Kimheden, the aluminosilicate chlorite is a common Mg-bearing gangue mineral that may be
dissolved by the acidic drainage, possibly along with traces of Mg-bearing carbonates. On the
basis of the correlation data, S and Mg were chosen as natural tracers for AMD in the receiving
stream.
The results of dissolved concentrations of S and Mg at all locations in 2009 and 2010
are plotted in Figure 8.a. They display a mixing line between two end-members: SB-GB on the
one hand (background water), with the lowest S and Mg concentration values, and G1 on the
other hand (groundwater in back¿lled open pit 1), with the highest S and Mg concentrations.
Concentrations of S and Mg in the G1 population itself exhibit a linear trend, which is probably
due to varying dilution of the contaminated groundwater by natural water. At the locations close
to the pits, other processes than dilution affect the concentrations of S and Mg, manifested by a
less distinct linear relationship. In the lower parts of the stream, the processes of dilution of AMD
with natural water prevail in the variations of concentrations of S and Mg (cf. previous section).
Therefore, S and Mg can be used as natural tracers to estimate the contributions from AMD and
natural water downstream. The locations close to the pits are excluded, and the AMD end-member
chosen is the median of concentrations at S4, which is the ¿rst location in the stream after the
pits. The background water end-member is the median of concentrations at SB and GB, which
are natural surface and ground-water, respectively. Downstream concentrations of S and Mg and
positions of the end-members are presented in Figure 8.b.
(8.a)
20 SB
18 GB
G1
16
S1a
14 S1
12 S1b
Mg (mg/L)

S1c
10
S2
8
S3a
6 S3
4 S4
S5
2
S6
0 SD
0 20 40 60 80 100 120 140 160 180
S (mg/L)
(8.b) 10
S4, acid mine
9 drainage
8

7
SB+GB
6
Mg (mg/L)

S4
5 S5
4 S6
SD
3

1
SB+GB, background water
0
0 20 40 60 80 100 120
S (mg/L)

Figure 8 (8.a) Concentrations of S and Mg in all samples taken in 2009 and 2010 (8.b) S-Mg mixing line between
the background water end-member (median of concentrations at SB and GB) and the AMD end-member (median of
concentrations at S4) for sampling locations in the lower part of the stream.

13
The proportion of AMD was calculated at the 3 downstream locations and at the two end-
members, using mixing calculations with S and Mg as tracers. Median, minimum and maximum
proportions of AMD at S5, S6, SD and the two end-members are reported in Table 3. At SD,
the Pearson correlation coef¿cient between the proportion of AMD and the stream discharge is
also provided. For both S and Mg, the coef¿cient is larger than 0.7, which is explained by the
relationship between AMD contribution and stream discharge described in the previous section.
Mixing calculations with S at SD gave a median proportion of AMD of 19%; with Mg, 23%. This
implies that the proportion of background water at SD is 81% (S) or 77% (Mg). Dilution of AMD
by natural water is therefore signi¿cant downstream.

Table 3 Results of water mixing calculations with S and Mg tracers at downstream sampling locations (S5, S6 and
SD). For each tracer, the AMD end-member is represented by median concentrations at S4 (in-stream sampling
location directly after pit 2), and the background end-member is represented by median concentrations at SB and GB
(background surface and ground-water). Median, minimum and maximum proportions of AMD calculated from all
samples at each location are indicated. The Pearson correlation coef¿cient between the proportion of AMD and stream
discharge at SD is shown for each tracer. The number of samples at each location is also indicated.

%water from S4 (close to pit 2) S4 S5 S6 SD SB+GB


S Median 100 19 18 19 0

Minimum 88 14 18 10 -2

Maximum 113 24 18 30 1

Correlation with 0.82


discharge (SD)

Mg Median 100 19 19 23 0

Minimum 92 15 19 12 -4

Maximum 109 24 19 38 2

Correlation with 0.72


discharge (SD)
Number of samples 5 2 1 9 11

ŝƐĐƵƐƐŝŽŶ

džƉĞĐƚĞĚƌĞƐƵůƚƐĨƌŽŵƚŚĞƌĞĐůĂŵĂƟŽŶ
A major objective of the reclamation activities at Kimheden was to reduce the release
of metals and acid from sulphide oxidation in the waste. It was decided to apply a solution that
would mitigate AMD production at the source (abatement of sulphide oxidation), in a passive way,
and in the long term. Exclusion of oxygen transport to the waste rock was therefore the preferred
method. As the open pits are located on a hill side, establishment of a pit lake where waste rock
could be disposed of and preserved in oxygen-poor conditions was not an option. Application
of a low permeability barrier on back¿lled waste was, on the other hand, feasible, and could
also achieve reduction of vertical diffusion of oxygen into the waste. Indeed, it has been shown
that low permeability soils present increased water retention (Elander et al., 1998), which creates
water-saturated conditions in the barrier and contributes to the reduction of oxygen diffusion to
the underlying sulphidic material. In addition, dry covers including a sealing layer are expected
to reduce the vertical inÀow of water to the waste, thereby decreasing the quantity of leachate

14
released. The Swedish EPA evaluated the performance of the most common types of multiple-
layer dry covers including a sealing layer (SEPA, 1993), and they estimated that, depending on the
cover type, they could reduce the sulphide oxidation by 94% to more than 99%, and the subsequent
metal leaching by 97% to more than 99.8%. At Kimheden, the dry cover was expected to limit the
transport of oxygen into the waste, but not the transport of water, as groundwater was identi¿ed as
the major source of water inÀow to the waste (Rosén & Wilske, 1994). The speci¿c objective of
back¿lling and sealing at Kimheden was to achieve concentrations of Cu+Zn under 1 mg/L at one
of two key locations of the site, which implied a decrease by 87%-94% along the receiving stream.

ĞĐƌĞĂƐĞŽĨĐŽŶƚĂŵŝŶĂŶƚĐŽŶĐĞŶƚƌĂƟŽŶƐĨŽůůŽǁŝŶŐƌĞĐůĂŵĂƟŽŶ
The expected reduction of metal release from mine waste as a result of reclamation did
occur. Results show that concentrations of both Cu and Zn in the receiving stream were rapidly
and signi¿cantly reduced after back¿lling and application of the dry cover. The decrease occurred
in similar proportions along the sampled stream, which shows that the effects of the reclamation
were similar at both open pits. The minimum performance objective of 87% decrease of Cu+Zn
concentrations in the stream has been achieved, which suggests the success of the mitigation
measures as a ¿rst approach.
Comparison of the performance of back¿lling and sealing of waste rock between Kimheden
and other open-pit sites is complicated by the fact that, as mentioned by Tremblay and Hogan
(2001), in-pit back¿lling is a widespread practise but it lacks scienti¿c monitoring data. Much
more information in the scienti¿c literature can be found on the design and sometimes short-term
assessment of covered back¿ll than actual performance in the longer term. Dry covers including
a sealing layer have been more documented, and some performance data can be found in e.g.
Skousen et al. (2000), Tremblay and Hogan (2001), Wilson et al. (2003) and Höglund et al. (2004).
Although the reported examples mostly deal with small-scale trials or short-term performance at
mine sites, they usually point out good results of the dry covers, manifested by substantial decrease
of oxygen diffusion and water in¿ltration to the mine waste, and lower release of contaminant
leachate.

^ŝŐŶƐŽĨŽŶͲŐŽŝŶŐƉƌŽĚƵĐƟŽŶŽĨDĂŶĚŶĞĞĚĨŽƌĨƵƌƚŚĞƌƌĞĐůĂŵĂƟŽŶ
Water samples downstream of Kimheden mine from 2002 to 2010 exhibited stabilised
concentrations of metals and sulphur, pH and EC. It seems therefore that the improvement of water
quality mentioned previously had already ceased 6 years only after completion of the rehabilitation
in 1996, maybe even earlier, as data is lacking in the ¿rst years after reclamation. Copper and zinc
are available from 1999, and were already stabilised by that time. Thus, release of metals and
acid from the waste quickly reached equilibrium after completion of the reclamation. However,
establishment of oxygen-reduced conditions in the back¿lled pits and full wash-out of oxidation
products are expected to be progressive and rather slow. Therefore, this quick stabilisation
suggests that production of AMD from the back¿lled waste is still going on, though at lower
rate than before reclamation. According to concentrations of sulphate in 1983 and 2009, pyrite
oxidation rate would have been reduced by 80% as a consequence of back¿lling and sealing. The
decrease is notably lower than the performance criteria of 94-99% evaluated by the Swedish EPA
for application of multiple-layer dry covers (SEPA, 1993). Besides, pH along the receiving stream
has remained relatively low until now. The values range between 3.0 and 3.7, which is not much
higher than ¿rst available measurements of pH along the stream in 1987 (2.5-3.0). Additionally,
current values are at least 0.9 pH unit lower than the background pH on site (4.6). Improvement
of the water quality at Kimheden was therefore limited, and this might be due to the incapacity of
reclamation activities to suf¿ciently reduce oxygen transport to the back¿lled waste.

15
Two types of inadequacies of the mitigation measures may be recognised. The dry cover
could constitute a poor barrier against oxygen, resulting in persistent diffusion of oxygen into the
waste. This de¿ciency may, in that case, be related to insuf¿cient water retention in the sealing layer,
or its early deterioration by freeze and thaw cycles, subsidence etc. However, unsaturated fractures
in the pit walls, or inÀow of oxygenated groundwater into the pits are an equally reasonable source
of oxygen to the waste, as archive data indicate that the rock substratum surrounding the pits is
relatively fractured (Rosén & Wilske, 1994).
Due to on-going acid generation at Kimheden, additional remediation measures should be
considered. Though concentrations of Cu and Zn have been signi¿cantly reduced after completion
of the reclamation, current concentrations of metals in the mine drainage are not satisfactory for
discharge into the natural environment (Boliden, personal communication). Results showed that if
the concentrations of contaminants from open pit 1 and open pit 2 are similar, the loads from open
pit 1 are much higher, as water discharge is also much higher. Consequently, additional mitigation
actions should focus on back¿lled open pit 1 in the ¿rst place.
If the continued generation of acidity may be explained by persistent oxidation of the
back¿lled sulphidic waste, another source of acid production might be secondary processes like
the formation of iron hydroxides from Fe dissolved in the drainage. The inÀuence of iron hydroxide
precipitation on the generation of acidity is currently under investigation with additional ¿eld data
on iron speciation.

WĂƚŚǁĂLJƐŽĨŵŝŶĞĚƌĂŝŶĂŐĞŽŶƐŝƚĞ
The aim of the ditches constructed around the back¿lled pits at the early stages of
reclamation was to limit the inÀow of water to the mine waste, and to divert the contaminated
drainage to a downstream pond treated with lime. It seems, however, that the ditches are not
entirely ful¿lling their role. Field observations indicate that most of the ditches upstream of
the waste remain dry soon after snow melting, which means that they must have a very limited
inÀuence on the reduction of water inÀow to the pits. Besides, discharge measurements show that
a large volume of contaminated water running in the collection ditch disappears as groundwater
between the two pits, demonstrating that the base of the ditch is not impermeable enough. Thus, a
considerable part of the leachate from the back¿lled pits may end up as groundwater, which makes
it very dif¿cult to monitor and treat. Furthermore, mixing calculations with natural tracers show
that the water sampled in the ditch downstream is highly diluted, whereby ca. 80% of the drainage
is natural water. As the ditch leads to the treatment pond, dilution of the drainage adds a useless
amount of natural water to be treated with lime, when much of the contaminated water is instead
disappearing in the surrounding ground.
Despite the dilution of mine drainage with natural water, concentrations of critical metals
(Cu, 400 ȝg/L; Zn, 120 ȝg/L; and to a lesser extent, Cd, 0.3 ȝg/L) in the lower part of the stream
are still considered relatively high by the Swedish EPA (SEPA, 2000). This indicates that areas
downstream of the mine that are impacted by mine drainage (in the form of surface or ground-
water), should still be considered polluted. The transport of contaminants from the mine varies
with the Àow, and it is expected to be lowest during winter, and highest during the spring Àood.

>ŝŵŝƚĂƟŽŶƐŽĨƚŚĞƐĂŵƉůŝŶŐŵĞƚŚŽĚƐ
Results from the 2009-2010 sampling programme are based on one to two-day sampling
sessions along the receiving stream, attempting to capture spatial variability. A possible limitation,
in this case, is that steady-state conditions required to conclude on the spatial variability may
not be ful¿lled over the sampling event (Runkel et al., 2009). If the loads of contaminants at one
location considerably change between the beginning and the end of the sampling along the stream,

16
variations of loads between locations and determination of sources and sinks of contaminants may
become misleading. Temporal variability on the scale of the sampling event is expected to be a
problem in systems where hydrological changes are brutal. Investigation of the temporal variability
over a sampling event was not included in the scope of the present work. However, repetition of
the sampling sessions in 2009 and 2010 alleviated this problem by comparison between sessions
and determination of trends. Another issue related to temporal variability raised by the same author
(Runkel et al., 2009), is the possibility that very different hydrological conditions between pre- and
post-reclamation sampling sessions might bias the comparison of water quality between the two
periods. Very different hydrological conditions may involve a change of sources of contaminants
in the stream, arti¿cially affecting the concentrations and loads. In the present study, nevertheless,
care has been taken to compare not only pre- and post-reclamation sampling occasions between
each others, but also two sampling years (1991 for early reclamation stage and 2009 for post-
reclamation stage), by averaging data from several sampling sessions performed at similar dates
of the year.

ŽŶĐůƵƐŝŽŶƐ
Back¿lling waste rock into the open pits and capping with a till low permeability barrier
was, to some extent, a successful measure at Kimheden. Long-term monitoring of Cu and Zn
concentrations in the receiving stream showed that covering the back¿lled waste succeeded
in reducing concentrations of Cu and Zn added together by more than 87%. The decrease was
relatively rapid, and has been sustained until today.
However, the water quality in the mine drainage is still not suf¿cient to allow discharge into
the natural environment. Average dissolved concentrations of Cu, Zn, Cd and Al at key locations
of the stream were respectively in the range [380-1600] ȝg/L, [88-450] ȝg/L, [0.16-1] ȝg/L and
[3.1-9.6] mg/L in 2009. pH values were in the range [3.0-3.7] at the same locations. Trends from
metal concentrations, pH and EC since 2002 indicate a quick stabilisation of these parameters, and
suggest that the mitigation of AMD has reached a plateau.
Discharge measurements and mixing calculations with S and Mg natural tracers suggest
that a substantial part of the mine drainage ends up as groundwater instead of reaching the treatment
pond as it was aimed. The collection ditch (receiving stream) is probably not impermeable enough
to ful¿l its role.
In order to improve the rehabilitation, there will be a need of self-sustained additional
remedial measures taking into account the topography of the site, and the reclamation already
performed. The actions should focus on decreasing further the oxidation rate, and the leaching of
contaminants.
The persistent discharge of relatively high concentrations of metals and acidity is suspected
to be due to continued access of oxygen to the mine waste in spite of capping, and wash-out of the
oxidation products by inÀow of groundwater. These hypotheses are currently under investigation,
with the help of additional ¿eld studies at Kimheden.

ZĞĨĞƌĞŶĐĞƐ
Alakangas, L., Lundberg, A., & Öhlander, B. (2005). Pilot scale studies of different dry
covers on sulphide-rich tailings in northern Sweden. In: proceedings of Securing the Future:
International conference on Mining and the Environment, Metals and Energy Recovery, Skellefteå,
Sweden, pp.9-18.
Andersson, G. (1988). Kimheden, miljöåtgärder. Boliden AB. In Swedish.

17
Årebäck, H., Barrett, T. J., Abrahamsson, S., & Fagerström, P. (2005). The Palaeoproterozoic
Kristineberg VMS deposit, Skellefte district, northern Sweden, part I: geology. Mineralium
Deposita, 40(4), 351-367.
Axelsson, C.-L., Ekstav, A., Holmén, J., & Jansson, T. (1991). Efterbehandling av
sandmagasin i Kristineberg, Hydrogeologiska förutsättningar för åtgärdsplan: Lakvattenbalanser
och vittringsbegränsande åtgärder. Golder Geosystem AB. In Swedish.
Bambic, D. G., Alpers, C. N., Green, P. G., Fanelli, E., & Silk, W. K. (2006). Seasonal and
spatial patterns of metals at a restored copper mine site. I. Stream copper and zinc. Environmental
Pollution, 144(3), 774-782.
Bazemore, D. E., Eshleman, K. N., & Hollenbeck, K. J. (1994). The role of soil water
in stormÀow generation in a forested headwater catchment - synthesis of natural tracer and
hydrometric evidence. Journal of Hydrology, 162(1-2), 47-75.
Boliden (1995). Efterbehandlingsplan för sandmagasin 1B samt Kimhedengruvan i
Kristineberg. Boliden AB. In Swedish.
Brady, K. B. C., Smith, M. W., & Schueck, J. (Eds.) (1998). Coal mine drainage prediction
and pollution prevention in Pennsylvania. Harrisburg, Pennsylvania: Department of Environmental
Protection.
Chapman, J., Hockley, D., Sevick, J., Dachel, R. & Paul, M. (1998). Pit back¿lling on two
continents: Comparison of recent experiences in the Wismut and Flambeau projects. In: Tailings
and mine waste ’98, International Conference on Tailings and Mine Waste (pp.55-65). Rotterdam:
Balkema.
Elander, P., Lindvall, M. & Håkansson, K. (1998). MiMi - Prevention and control of
pollution from mining waste products - State-of-the-art-report. MiMi Report 1998:2, Mitigation
of the environmental impact from mining waste programme (MiMi). MiMi Print.
Gordon, N. D. (Ed.) (2004). Stream hydrology: an introduction for ecologists. (2. ed.)
Chichester: Wiley.
Gray, T. A., & Gray R. E. (1992). Mine closure, sealing and abandonment. In H. L. Hartman
(Ed.), SME mining engineering handbook (pp.659-674). Vol. 1. (2. ed.). Englewood, Colorado:
Society for Mining, Metallurgy, and Exploration.
Hannington, M. D., Kjarsgaard, I. M., Galley, A. G., & Taylor, B. (2003). Mineral-chemical
studies of metamorphosed hydrothermal alteration in the Kristineberg volcanogenic massive
sul¿de district, Sweden. Mineralium Deposita, 38(4), 423-442.
Höglund, L. O., & Herbert, R. (Eds.), Lövgren, L., Öhlander, B., Neretnieks, I., Moreno,
L., Malmström, M., Elander, P., Lindvall, M., & Lindström, B. (2004). MiMi - Performance
assessment: main report. MiMi report 2003:3, Mitigation of the environmental impact from mining
waste programme (MiMi). Luleå, Sweden: MiMi Print.
Hooper, R. P., Christophersen, N., & Peters, N. E. (1990). Modelling streamwater chemistry
as a mixture of soilwater end-members - an application to the Panola mountain catchment, Georgia,
U.S.A. Journal of Hydrology, 116(1-4), 321-343.
INAP (the International Network for Acid Prevention) (2009). Global Acid Rock Drainage
Guide (GARD Guide). Chapter 2.4.3. http://www.gardguide.com/.

18
Joerin, C., Beven, K. J., Iorgulescu, I., & Musy, A. (2002). Uncertainty in hydrograph
separations based on geochemical mixing models. Journal of Hydrology, 255(1-4), 90-106.
Land, M., Ingri, J., Andersson, P. S., & Ohlander, B. (2000). Ba/Sr, Ca/Sr and 87Sr/86Sr ratios
in soil water and groundwater: implications for relative contributions to stream water discharge.
Applied Geochemistry, 15(3), 311-325.
Malmström, M., Werner, K., Salmon, U. & Berglund, S. (2001). Hydrogeology and
geochemistry of mill tailings impoundment 1, Kristineberg, Sweden: Compilation and interpretation
of pre-remediation data. MiMi report 2001:4, Mitigation of the environmental impact from mining
waste programme (MiMi). Luleå, Sweden: MiMi Print.
Phelps, L. B. (1990). Unit operations of reclamation. In: B. A. Kennedy (Ed.), Surface
mining (pp.770-776). (2. ed.). Littleton, Colorado: Society for Mining, Metallurgy, and Exploration.
Reed S. M., & Singh R. N. (1986). Groundwater recovery problems associated with
opencast mine back¿lls in the United Kingdom. International Journal of Mine Water, 5(3), 47-74.
Rosén, L. & Wilske, Å. (1994). Beräkningar av metallÀöden och förslag till åtgärder vid
Kimhedengruvan, Kristineberg. Scandiaconsult Väst AB. In Swedish.
Runkel, R. L., Bencala, K. E., Kimball, B. A., Walton-Day, K., & Verplanck, P. L. (2009).
A comparison of pre- and post-remediation water quality, Mineral Creek, Colorado. Hydrological
Processes, 23(23), 3319-3333.
Skousen, J. G., Sexstone, A., & Ziemkiewicz, P. F. (2000). Acid mine drainage control and
treatment. In: R. I. Barnhisel, R. G. Darmody, and W. L. Daniels (Eds.), Reclamation of drastically
disturbed lands (pp.131-168). American Society of Agronomy and American Society for Surface
Mining and Reclamation. Agronomy No. 41, Madison, Wisconsin.
SEPA (Swedish Environmental Protection Agency) (1993). Gruvavfall från
sul¿dmalmsbrytning - metaller och surt vatten på drift -. Report 4202. Stockholm: Naturvårdsverket
Förlag. In Swedish.
SEPA (Swedish Environmental Protection Agency) (2000). Environmental quality criteria
- lakes and watercourses. Report 5050. Kalmar: Lenanders.
Tremblay, G. A., & Hogan, C. M. (Eds.) (2001). MEND Manual, Volume 4 - Prevention
and control, MEND 5.4.2d, Mine Environment Neutral Drainage programme (MEND).
Villain, L., Alakangas, L., & Öhlander, B. (2010). Geochemical evaluation of mine water
quality in an open-pit site remediated by back¿lling and sealing. In: IMWA 2010 - proceedings
of the International Mine Water Association Symposium, Sydney, Canada, Wolkersdorfer C. &
Freund A. (Eds.), Cape Breton University Press, pp.515-518.
Wilson, G. W., Williams, D. J., & Rykaart, E. M. (2003). The integrity of cover systems
- An update. In: ICARD 2003 - proceedings of the Sixth International Conference on Acid Rock
Drainage, Cairns, Australia, pp.445-451.
Younger, P. L., Banwart, S. A., & Hedin, R. S. (2002). Mine water: hydrology, pollution,
remediation. Dordrecht, The Netherlands: Kluwer Academic.

19
II

*URXQGSHQHWUDWLQJUDGDUDQGUHVLVWLYLW\PHWKRGV
WRHYDOXDWHWKHUHFODPDWLRQRIDQRSHQSLWPLQH
E\EDFNÀOOLQJDQGVHDOLQJ

Lucile Villain, Nils Sundström, Nils Perttu, Lena Alakangas, Björn Öhlander
'ƌŽƵŶĚƉĞŶĞƚƌĂƟŶŐƌĂĚĂƌĂŶĚƌĞƐŝƐƟǀŝƚLJŵĞƚŚŽĚƐƚŽĞǀĂůƵĂƚĞƚŚĞ
ƌĞĐůĂŵĂƟŽŶŽĨĂŶŽƉĞŶͲƉŝƚŵŝŶĞďLJďĂĐŬĮůůŝŶŐĂŶĚƐĞĂůŝŶŐ
Lucile Villain1, Nils Sundström1, Nils Perttu1, Lena Alakangas1, Björn Öhlander1

1
Luleå University of Technology, Division of Geosciences and Environmental Engineering, 971 87 Luleå, Sweden

ďƐƚƌĂĐƚ
Evaluation of the performance of reclamation at acid generating mine sites is essential to
improve mitigation methods, and for this purpose, geophysical surveys may be a useful complement
to geochemical water quality assessments. The effects after fourteen years of in-pit back¿lling of
waste rock and capping at a small copper mine in northern Sweden were investigated using ground
penetrating radar and direct current resistivity. The dry cover on the pits consists of a sealing layer
expected to limit oxygen diffusion to the waste, overlain by a protective layer. Special attention
was drawn on identifying potential pathways of oxygen to the waste. No sign of existing damage in
the sealing layer could be found with the radar, but access of oxygen to the waste through fractures
in the pit walls was considered possible. Risks of deterioration of the sealing layer in the long term
were identi¿ed. The radar localised regions of thinner protective layer which could be exposed to
frost action, and resistivity indicated a risk of erosion of the sealing layer from seepage Àowing
through the dry cover. Despite the inÀuence of resistivity side effects, the heterogeneous back¿lled
material could be successfully imaged with resistivity. We showed, however, that the groundwater
table in the back¿ll could not have been identi¿ed, even in perfect conditions of measurement. In
the surrounding bedrock, resistivity suggested that the contaminated water occurs mostly on the
ground and in the subsurface.

/ŶƚƌŽĚƵĐƟŽŶ
Solutions to control contaminated drainage from mines have been developed during the last
decades. Prevention and mitigation methods at the source have been promoted as an economic and
practical alternative to treatment of the polluted water. In this approach, efforts are concentrated on
limiting the reactions generating contaminants, and reducing leaching and transport of the reaction
products by water (INAP, 2009), rather than removing the contaminants from the drainage.
At coal and hard-rock mines where toxic metal-rich Acid Mine Drainage (AMD) is produced
from the oxidation of sulphide rocks as they are exposed to water and oxygen, prevention and
mitigation may consist in reducing water and/or oxygen access to the mining residues. This may be
achieved by e.g. interception and diversion of surface water upstream of waste rock piles and mine
workings, application of various types of dry covers on mine waste, or Àooding them with water
to limit oxygen diffusion. Monitoring the effects of such reclamation methods on the quality of the
mine drainage after on-site application gives essential data to evaluate the need for improvement
of the rehabilitation, and upgrade its implementation at other sites. Monitoring programmes may
consist of regular hydro-geochemical studies at the reclaimed sites, by e.g. sampling of surface
water, groundwater and sediments, and measurement of water Àows and groundwater levels.
Examples of such post-reclamation studies can be found in Holmström et al. (2001), Brake et al.
(2001), Jeffree et al. (2001), Bambic et al. (2006) and Runkel et al. (2009). When the sampling
programme is extensive enough, these methods may provide abundant and detailed data to assess
the effects of rehabilitation on the release of contaminants from the mine. Common hydro-
geochemical studies, however, present the disadvantage of generating local results that may be

1
dif¿cult to interpolate in order to appreciate evolution of pathways and quality of water across the
site. Indirect geophysical methods that have the potential to image the contaminated drainage may
help picture spatial variations and support local sampling results. Moreover, structural information
about the site can be supplied by geophysical techniques, in addition to water-related data.
Geophysical methods may be used for hydro-geological mapping, estimation of hydrological
parameters, and monitoring of hydrological processes (Rubin & Hubbard, 2005). Due to the
ability of electrical and electromagnetic techniques to image groundwater, and even sometimes
its quality, they have proven useful in environmental studies. Methods like resistivity, ground
penetrating radar, and induced polarisation have been largely employed in studies of leachate
transport from land¿lls (Nobes et al., 2000; Abu-Zeid et al., 2004; Porsani et al., 2004; Dahlin
et al., 2010). They have also been used in mining environmental studies, in order to image AMD
plumes (Buselli & Lu, 2001; Rucker et al., 2009); tailings ponds (Faz et al., 2005; Placiencia-
Gymez et al, 2010; Marttn-Crespo et al., 2010); or waste rock piles (Van Dam et al., 2005; Poisson
et al., 2009; Anterrieu et al., 2010). The present investigation is in line with these studies, but its
singularity is the focus on imaging back¿lled pits and their dry cover. Advantages of geophysical
techniques in such studies are practical application and relatively limited cost. A major limitation,
however, is that they give indirect images of the studied objects, and the data may therefore lead
to misinterpretations without good knowledge of the underground characteristics of the site, in
particular the variations of lithologies.
Ground Penetrating Radar (GPR) and Direct Current (DC) resistivity surveys were
conducted in 2010 at Kimheden open-pit mine, northern Sweden, in order to characterise the
effects of the reclamation completed 14 years before, and to complement the information obtained
from hydro-geochemical studies. Mitigation of AMD at Kimheden was mainly achieved through
the back¿lling of waste rock into two small open pits, and application of a composite till dry cover
on top. A previous geochemical study (Villain et al., 2010) demonstrated that, in spite of the large
decrease of Cu and Zn concentrations in the mine drainage following reclamation, water quality
at the mine is still insuf¿cient, and this might be due to inadequacies in the mitigation methods
used. One possibility is that control of oxygen and water access to the back¿lled waste has failed
to decrease the rate of sulphide oxidation in the rocks to an acceptable level.
Therefore, two main objectives of this geophysical study can be recognised; (1) Identify
potential sources of oxygen to the back¿lled waste and possible de¿ciencies of the mitigation
measures that might challenge their success in the short and long terms (2) Image groundwater and
contaminated water pathways in the proximity of one of the pits which generates the largest amount
of contaminants. While DC resistivity was expected to give information about groundwater, AMD
and fractured media, GPR would document structural characteristics like thickness of the dry
cover, discontinuities in the rock substratum surrounding the pits and in the cover. Results related
to the dry cover are presented and discussed ¿rst, followed by ¿ndings about mine waste and water
pathways.

^ƚƵĚLJƐŝƚĞ

^ŝƚĞůŽĐĂƟŽŶ͕ŐĞŽůŽŐŝĐĂůĂŶĚŚLJĚƌŽůŽŐŝĐĂůĐŽŶƚĞdžƚ
Kimheden site is a small copper mine situated in the Kristineberg mining area (Figure 1.a)
in the region of Västerbotten, northern Sweden. The regional climate is cold, with a yearly average
air temperature of 0.7 ºC, and 5 months with an average temperature below 0 ºC (Malmström et
al., 2001). Annual precipitation at the site is ca. 400-800 mm (Axelsson et al., 1991), accumulating
in the form of snow from October to May.

2
a b
ARVIDSJAUR

WASTE ROCK DUMP


ABBORTRÄSK

GLOMMERSTRÄSK
N OPEN PIT 1

H
MALÅ ITC
ND H
TIO TC
KIMHEDEN,
KRISTINEBERG EC DI
LL N
CO IO
RS
20 km
VE
DI
Luleå

Skellefteå

Shaft
WASTE ROCK DUMP

Stockholm

OPEN PIT 2
TO TR
EATM
EN
T PO
ND

Figure 1 (a) Location of Kimheden open-pit site in the Kristineberg mining area, northern Sweden. (b) Aerial view
of the site in the beginning of reclamation activities (1982). The position of open pits 1 and 2 is indicated, as well as
waste rock dumps and water management ditches.

Kristineberg belongs to a deformed and metamorphosed Palaeoproterozoic volcanic


domain. Kimheden is one of the smaller pyrite-rich massive sulphide deposits of the area, which
are intercalated within a succession of felsic volcaniclastic rocks that have been altered to quartz-
muscovite-chlorite schists (Hannington et al., 2003). Archive electromagnetic data showed that
the mineralization is striking in the north-eastern direction, which is also the direction of the two
open pits of the site. The bedrock is covered by a one to two metre thick glacial till cover, locally
overlain by a thin layer of peat. The terrain is sloping towards north-west and south-west. The
top of the hill, located north-east of Kimheden, lies at an altitude of 561 m (Hellman & Lokrantz,
2008).
The surface-near groundwater Àows partly in the till cover, and partly through fractures in
the bedrock. According to Rosén and Wilske (1994), fractures around the pits are mainly oriented
in the direction of the pits. They also estimated that transmissivity in the bedrock surrounding the
open pits was ca. 9.10-5 m2.s-1. The same report states that almost all water inÀows to the pits occur
through fractures in the ground. Re-logging for geotechnical properties of drill cores left from
exploration investigations corroborated that the bedrock is relatively fractured.
Data from Rosén and Wilske (1994) and ¿eld observations were compiled into a hydrological
model of back¿lled open pit 1 (Figure 2) where the resistivity survey was performed. Suspected
main surface and ground-water Àow directions are indicated.

3
Figure 2 Conceptual hydrological model in the vicinity of back¿lled open pit 1, established from archive data and
¿eld observations. The size of the arrows is proportional to the importance of the Àows into and out from the pit.

DŝŶŝŶŐĂŶĚƌĞĐůĂŵĂƟŽŶĞǀĞŶƚƐ
Copper ore was extracted underground and in two open pits by the Swedish mining
company Boliden AB between 1968 and 1974 (Figure 1.b). The eastern pit or pit 1 (respectively
western pit or pit 2) is 210 m long (respectively 140 m long); both are approximately 20 m wide
and less than 10 m deep in average. The total tonnage was 0.13 Mt with a grade of 0.95 % Cu
(Årebäck et al., 2005). Waste rock left from the extraction was dumped in the proximity of the pits,
and rapidly started to generate Cu- and Zn-rich AMD. As a consequence, a network of ditches was
excavated in 1982 (Figure 1.b), in order to reduce access of water run-off to the waste rock, and to
collect the mine drainage to a treatment pond downstream of the hill. Later on, the waste rock was
disposed of into the pits in different stages, in order to limit the area of contamination. Reclamation
activities were completed in 1995-1996, when full in-pit back¿lling of the waste rock left, along
with some remains of the mining facilities and roads was performed, and a dry cover was applied
on the waste (Figure 3). The dry cover was meant to include a layer of 0.3 m clayey till (sealing
layer) overlain by 1.5 m unsorted till (protective layer). The actual thickness of the protective
layer is investigated in this study. Geotechnical tests before application of the dry cover showed
that the sealing layer contained about 8% clay, and had a hydraulic conductivity of ca. 1.10-9 m.s-1
(Edström & Schönfeldt AB, 1996).



 3URWHFWLYHOD\HUPXQVRUWHGWLOO
6HDOLQJOD\HUPFOD\H\WLOO






 %DFNILOOLQJRIZDVWHURFN
LQWRWKHRSHQSLW






              

Figure 3 Cross-section of open pit 1 showing the back¿lling and sealing process completed in 1995-1996. The dry
cover consists of 0.3 m clayey till (sealing layer) overlain by 1.5 m unsorted till (protective layer).

4
DĞƚŚŽĚŽůŽŐLJ

The GPR survey was carried out over three days in the beginning of June 2010. The ground
was free from snow, but remains of frost in the subsurface soil cannot be excluded. The resistivity
survey, on the other hand, was performed in the beginning of October 2010 when no frost
should have been left in the soil, and before establishment of the snow cover. For both methods,
measurements were performed along lines of constructed grids.

^ƵƌǀĞLJŐƌŝĚƐĂŶĚƌĞĨĞƌĞŶĐĞŵĞĂƐƵƌĞŵĞŶƚƐ
Survey grids were created in the areas of interest as guidance for the geophysical
measurements. Two grids were established for the GPR survey on the surface of back¿lled pits 1
and 2 (Figure 4). One survey grid was established for resistivity, consisting of lines 1 to 4, between
200 and 280 m long, on back¿lled pit 1 and the surrounding bedrock. The central lengthwise line
of the GPR grid on back¿lled pit 1 is a part of the longer line 1 of the resistivity grid. Elevation
measurements were performed with a levelling instrument on all grids. Correspondence with the
Swedish geographic coordinate system was realised with a Garmin eTrex GPS.

Y PIT 1

X PIT 1

H
ITC
ND
IO
E CT
LL
CO
LINE 3 BACKFILLED OPEN PIT 1

LINE 1 0 LINE 4

LINE 2

X PIT 2

GRP grids
Y PIT 2

Resistivity grid

Groundwater wells

Reference point for the sealing layer

BACKFILLED OPEN PIT 2

100 m
0

Figure 4 Overview of the GPR and resistivity survey grids at Kimheden mine site. The GPR coordinate systems (x,y)
used in the method and result sections is presented here. The position of the two groundwater wells installed in the
back¿ll of open pit 1 and the reference point for the depth of the sealing layer are also indicated.

Some ¿eld measurements were performed during the geophysical surveys to serve as
reference for data interpretation. Groundwater level was measured with an electric dip meter inside
two groundwater wells situated in the back¿ll of open pit 1 (Figure 4). Electrical conductivity was
measured with a WTW Multi 350i multimeter in groundwater samples taken from the same wells,
and surface water samples taken close to back¿lled pit 1.

5
'ƌŽƵŶĚƉĞŶĞƚƌĂƟŶŐƌĂĚĂƌ
The radar survey was performed using a RAMAC GPR system from Malå Geoscience.
Measurements were conducted every 5 cm along each survey line in the two GPR grids placed
over the two back¿lled open pits (Figure 4); the measurements were trigged using a “hip-chain”.
Each line was investigated twice, once using a 50 MHz RTA antenna, and then using a shielded
250 MHz antenna. Two different frequencies were used, as our intention was to map the sealing
layer at approximately 1.5 m depth (with the 250 MHz antenna), but also the groundwater table
and the Àoor of the pits, which were both assumed to be found at greater depths (with the 50 MHz
antenna). A 100 MHz antenna was also tested, but only 250 MHz and 50 MHz were selected as
they best ¿tted our objectives. In addition, reference measurements in air, necessary for the time
zero correction, were made with both antennas.

Analysis of radar data


This section describes the models of propagation velocity in the ground, and the procedure
used for the identi¿cation of interesting reÀections. Before interpretation of the radar data, dc-
shift, band-pass ¿lter, and altitude correction were applied to the data.
Models of propagation velocity in the till protective layer above the sealing layer were
required for each back¿lled open pit to estimate the thickness of the protective layer from the two-
way travel time of the signals. The models were based on velocity estimated from the measured
direct waves (waves travelling from the transmitter to the receiver without reÀections - in our case
travelling in the ground close to the surface -) as follows.
In back¿lled open pit 1, humidity of the till was observed to be highest in the south-west
side of the pit (xpit_1=0), and decrease in the x-direction (Figure 4). As it is reasonable to assume
that any larger variation in the velocity is mainly dependent on the amount of pore water in the soil,
the velocity model of this pit was established as a function of xpit_1 (Figure 5). This function was
obtained by ¿tting a 5th order polynomial curve to velocity values estimated from the direct waves
measured at all measurement points of the central lengthwise survey line (Figure 4).

Figure 5 Model of propagation velocity for back¿lled open pit 1.

6
In back¿lled open pit 2, the direct waves were also used to estimate the propagation velocity
at all measurement points of the central lengthwise survey line. However, no large variation in the
velocity was obtained, which may be explained by visibly drier conditions in the protective layer,
and the velocity model was simply taken as a constant calculated as the mean of the velocities
estimated from the direct wave at each measurement point along the line (vtill = 0.1053 m.ns-1).
A major objective with the GPR survey was to create maps of the depth of the sealing layer.
Therefore, attempts were made to identify the reÀection of the sealing layer for each radar pro¿le
obtained with the 250 MHz antenna. With the data from back¿lled open pit 2, the sealing layer
could easily be detected as the only clearly visible reÀection in the radar pro¿les. On the other
hand, data obtained from back¿lled open pit 1 showed many reÀections that could originate from
the sealing layer. To be able to pick the correct reÀection, some reference point was required. Such
a reference point was found from archive data, where the thickness of the protective layer was
measured to be 1.5 m during a ¿eld geotechnical control in 1996 (the reference point is indicated
on Figure 4). Providing that no signi¿cant compaction of the protective layer occurred since then,
and using this reference point together with the velocity model, we could identify the reÀection
from the sealing layer on the closest survey pro¿le. In the rest of the pro¿les, the reÀection from the
sealing layer was identi¿ed by comparing its depth at the intersection with the previous crossing
pro¿le, to make sure that the same reÀection was picked. This was an iterative process that was
pursued until all intersection points ¿tted together. Finally, the data obtained from all survey lines
were used to create 3-D plots picturing the depth of the sealing layer for both open pits.
The penetration of the 50 MHz antenna was not deep enough to map the groundwater
table and the Àoor of the pits, probably due to the high electrical conductivity in the back¿lled
waste rock. Therefore, 50 MHz data was used exclusively to identify areas of higher electrical
conductivity in relation with the open pits (stronger attenuation of the signals).

ƌĞƐŝƐƟǀŝƚLJ
Resistivity data was collected with the ABEM Lund Imaging system (Dahlin & Zhou,
2006) using the multiple gradient array with an electrode distance of 2 m. This con¿guration with
the SAS4000 terrameter permits multi-channel measurements, with 4 potential readings for each
pair of current electrodes. Each measurement was stacked 3-5 times. In back¿lled areas covered
by resistive till, problems were encountered to inject suf¿cient current in the ground. In order to
decrease the contact resistance between electrodes and the ground, salt water and extra current
electrodes were therefore used when needed. In the data analysis stage, bad data points were
manually removed. The data was then inverted using Res2Dinv with the robust inversion constrain
L1-norm (Loke et al., 2003). To support the interpretation of the inverted resistivity sections,
conceptual models of resistivity con¿gurations were constructed with Res2Dmod.

ZĞƐƵůƚƐĂŶĚĚŝƐĐƵƐƐŝŽŶ
While the ¿rst section about the dry cover discusses the results obtained at back¿lled open pits
1 and 2, the second section on mine waste and water pathways focuses on back¿lled pit 1, where the
resistivity survey was speci¿cally conducted because of a higher discharge of contaminants from this pit.

7
ƌLJĐŽǀĞƌ

epth of the sealing layer ʹ thiĐŬness of the proteĐƟǀe layer


The results from all radar pro¿les were plotted together on a 3D map showing the depth
of the sealing layer for each of the back¿lled open pits. Figure 6.a shows the aerial view of the
plot for back¿lled open pit 1, and Figure 6.b exhibits the map for back¿lled open pit 2. It must
be noted that areas plotted outside of the pit boundaries are not interpretations of the depth to the
sealing layer, but rather depth to the bedrock. The horizontal extent of the sealing layer outside of
the pits is unknown and cannot be determined by the radar, as the sealing layer outside lies directly
on the bedrock, and both sealing layer and bedrock have the same signature with the radar. It is
therefore relevant to consider the depth of the sealing layer only within the boundaries of the pits.
Besides, the radar survey lines do not cover the whole surface of back¿lled open pit 2 (Figure 6.b);
consequently there are blank areas that cannot be interpreted even inside the pit boundaries.

Figure 6 Aerial view of the 3D plot of the depth of the sealing layer on (a) back¿lled open pit 1 and (b) back¿lled open
pit 2. The coordinate systems employed are the same as for the GPR grids on the pits presented in Figure 4. The depth
of the sealing layer in metre is referred to by a colour bar. Measurement points on the survey lines, pit boundaries, and
the reference point used for identi¿cation of the sealing layer on back¿lled open pit 1 (cf. methods) are indicated on
the ¿gures.

8
According to the constructed models presented in Figure 6, the depth of the sealing
layer varies from ca. 1 m to ca. 2.3 m on back¿lled open pit 1, and from ca. 1 m to ca. 2 m on
back¿lled open pit 2. Sources of error are, however, affecting these results, which arise mainly
from uncertainties in the interpretation of the data, and possible inaccuracies in the models of
propagation velocity; they will be discussed here separately.
Uncertainties in the interpretation of the data lie in the choice of the right reÀection for the
sealing layer. Multiple super¿cial reÀections were observed on most of the GPR pro¿les, which,
according to the velocity model and the theoretical thickness of the protective layer (1.5 m), turned
out to be too close to the surface to be interpreted as the sealing layer. These reÀections should
therefore belong to the protective layer, and a deeper reÀection was selected for the sealing layer.
Note that the choice of the correct reÀection relies on an accurate model of propagation velocity;
otherwise, the risk is to pick one of the super¿cial reÀections as the sealing layer. The risk is
highest for back¿lled open pit 1, where attenuation of the radar signals towards depth occurred
more rapidly than for back¿lled open pit 2, and the reÀection from the sealing layer is much less
distinct. Lower penetration depth of the radar signals on back¿lled open pit 1 could be explained
by more humid conditions in the protective layer, resulting in higher conductivity of the soil. This
would explain why the super¿cial reÀections are more visible than the lower reÀection from the
sealing layer (Figure 7). To determine the position of the reÀection from the sealing layer, depth at
the reference point (1.5 m) and the velocity model were used (cf. methods). The uncertainty due
to the choice of one particular reÀection that is not clearly de¿ned while multiple reÀections are
visible is increased by the fact that the selection of the right reÀection for all pro¿les relies on a
unique reference point. On back¿lled open pit 2, the radar data is of higher quality, and the sealing
layer signature was more distinct (Figure 8). The interpretation model for the sealing layer is
therefore more reliable. Even though super¿cial reÀections were also encountered in the protective
layer of back¿lled open pit 2, the risk of confusion with the sealing layer is minor, since the lower
reÀection interpreted as the sealing layer is more visible.
Several interpretations were considered for these super¿cial reÀections. They could, for
example, arise from heterogeneities in the composition of the till protective layer, which may
naturally present variations in its clay content. Deposition of the protective layer in several stages
during reclamation could have enhanced this phenomenon. Another explanation could be the
presence of local water tables from perched aquifers lying on the sealing layer, but no investigation
of the existence of such aquifers has been carried out. A third possibility is that frost could still
have affected the subsurface of the protective layer at the time when the GPR measurements were
conducted. The super¿cial reÀections would then originate from the interface between super¿cial
unfrozen soil and an underlying frozen soil layer. Such a frost interface has been recognised earlier
with GPR by Bergström (1997) in a similar protective layer on sulphide tailings in central Sweden.
The uncertainty in the model of propagation velocity arises mostly from the assumption
that the protective layer is vertically homogeneous, as it is supposed to represent the protective
layer all the way down to the sealing layer, while it is based on the travel times of the direct waves
on the surface. We believe that this is a good hypothesis for back¿lled open pit 2, which was
characterised by fairly dry conditions (i.e. the possibility of signi¿cant variation in the propagation
velocity across the protective layer is low), but for the more humid back¿lled open pit 1 (i.e. the
possibility of large variations in the propagation velocity across the protective layer is higher),
the validity of the assumption is on weaker grounds. It should also be noted that the direct waves
travelling close to the ground surface are always inÀuenced by air, which could lead to some over-
estimation of the propagation velocity. However, we believe that at least the horizontal trend in the
velocity along the length of the pits depicted by the constructed velocity models (cf. methods), i.e.
horizontal variation for back¿lled open pit 1 and constancy for back¿lled pit 2, should have been
captured in a satisfactory way.

9
Figure 7 Ground penetrating radar pro¿le over the central lengthwise line of back¿lled open pit 1 (ypit 1= 10 m on
Figure 6.a). The highlighted trace represents the interpretation of the sealing layer.

140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

Figure 8 Ground penetrating radar pro¿le over the central lengthwise line of back¿lled open pit 2 (ypit 2 = 10 m on
Figure 6.b). The highlighted trace represents the interpretation of the sealing layer.

Due to the discussed uncertainties, conclusions should not be drawn about exact thickness
of the protective layer from the models presented in Figure 6, though it would have been interesting
to compare the values obtained with GPR with the planned thickness of the protective cover (1.5
m). This limitation concerns above all open pit 1, where the selection of the reÀection from the
sealing layer is less reliable, and the velocity model is more subject to errors. Thus, the models
should rather be used as a picture of general trends, like the variation of depth of the sealing layer.
They show that the depth varies up to one metre inside the boundaries of the pits.
The dry cover was successfully imaged by the resistivity survey. As it is composed of till
with low content of sulphides, the protective layer is contrastingly more resistive than the sulphide-
containing waste rock underneath. However, the sealing layer could not be imaged since it is very
thin (ca. 0.3 m) and its resistivity has been suppressed between the resistive uppermost layer and
the conductive spoil. Only one long pro¿le on the cover has been surveyed with resistivity (pro¿le
1, central lengthwise line on pit 1, cf. Figure 9.a in the next section), and it supports the assumption
of variation in thickness of the protective layer formulated from GPR data. According to this
¿gure, the protective layer in the central part of the pro¿le, between 130 m and 180 m, is thinner

10
and slightly less electrically resistive than from 180 m to the end of the pro¿le. In the beginning of
the pro¿le up to 100 m, the resistivity survey line is outside of the pit or on its edge, therefore the
dry cover does not lie on the waste rock but on the bedrock, and the variation in resistivity cannot
be used to delimit the bottom of the cover. Pro¿le 4 crossing pit 1 (Figure 12) also shows a portion
of the cover on the width of the pit. The pro¿le was 5 m away from the reference point for the
depth of the sealing layer (of 1.5 m), and the cover thickness on this portion is 2 m, which seems
a good estimation given the uncertainty of absolute depth values from the resistivity inversion.
The variations in thickness observed with both methods could be the result of an attempt
to level the ground from the rippled surface of the spoil, at the deposition of the dry cover. If they
actually are as large as indicated by the GPR models, they could in the long term present a risk for
the integrity of the sealing layer. The south-eastern lengthwise GPR pro¿le on back¿lled open pit 2
(ypit 2 = 0 m on Figure 6.b) exhibits a reÀection of good quality from the sealing layer, and its depth
towards the end of the pro¿le (ypit 2 = 0 m; xpit 2= 80 m to 140 m) is less than 1 m. From 80 m to
120 m, the pro¿le is within the back¿lled pit; consequently, a suf¿cient thickness of the protective
layer is required. A minimum thickness of 1.5 m has been established as a guaranty of preservation
of the sealing layer against deterioration from frost in the regional climatic conditions (Höglund,
2004). Adu-Wusu and Yanful (2006) tested the performance of three soil covers including a 0.9 m
protective layer of gravely sand at Whistle mine, Ontario. They showed that adequate thickness of
the protective layer was critical for the protection of the underlying low permeability layer from
frost in rigorous climate.

Integrity of the sealing layer


After construction of the 3D plots of the sealing layer from GPR results, the reÀection
selected as the sealing layer was re-investigated on each radar pro¿le, in order to detect major
irregularities like fractures and deformations. Small-scale fractures can be recognised as hyperbolas
in radargrams. Such features could be found close to the sealing layer in a few pro¿les. However,
since this hyperbola signature may also characterise bigger rocks in the uppermost till layer or
in the waste rock, and no additional sign in favour of a fracture like vertical displacement of the
layer was found, none of these features could ¿rmly be identi¿ed as fracture in the sealing layer.
The same precaution not to over-interpret hyperbolic patterns was expressed by Bergström (1997).
This suggests that individual fractures in the sealing layer may be very risky to interpret, unless
vertical displacement of the sealing layer is visible.
As the sealing layer reÀections were only clearly recognisable on back¿lled pit 2, the
continuity of the reÀections was investigated for this pit solely. No evident sign of interruption
or displacement of the sealing layer could be acknowledged. On the central lengthwise line of
back¿lled pit 2, an extra clear reÀection can sometimes be observed directly below the reÀection
from the sealing layer (Figure 8). This reÀection is interpreted as the bottom of the sealing layer,
which marks the interface between clayey till and waste rock. The GPR signals are rapidly
attenuated under the sealing layer, due to the higher conductivity of the water-saturated clayey till
(Bergström, 1997).
Although the thin sealing layer cannot be visualised with the resistivity method due to
limited resolution, resistivity variations in the uppermost protective layer may indicate zones of
erosion in the dry cover. Variations in resistivity in the protective layer can be observed in pro¿le
1 from the central lengthwise line on pit 1 (Figure 9.a). Nevertheless, without further information,
zones of lower resistivity may as well originate from perched aquifers on the sealing layer as from
washing out of the till. On pro¿le 4, the resistivity survey line crossing open pit 1, a zone of lower
resistivity is observed in the protective layer close to the low wall, directly in contact with the
back¿lled mine waste and extending up to the surface (Figure 12 in next section). Because of the

11
position and shape of the resistivity anomaly, this zone is interpreted as drainage water seeping
out from the mine waste through the dry cover up to the surface, or at least to the subsurface. This
assumption is legitimated by the sloping topography, which makes it possible for the drainage to
Àow over the low wall of the pit. In this area, patches of oxidation, or actual seeps in periods of
higher water Àow, can be observed on the ground surface, supporting the seepage interpretation as
well. It is believed that these patches are caused by precipitation of iron dissolved in the drainage
from the waste rock when it comes into contact with air and oxidises. Seepage of mine drainage
throughout the dry cover presents a risk of erosion of the sealing layer. This is true for back¿lled
open pit 1 where seepage was detected with resistivity on pro¿le 4, but also for back¿lled open pit
2 though no resistivity measurement has been performed there, since seepage from the back¿lled
pit is visible on the surface downstream of the pit. Similar case of erosion risk has been described
by Tremblay and Hogan (2001), whereby base-Àow discharge from sloping covered mine waste
dumps may cause deterioration of the overlying cover. Solutions to avoid such source of erosion
have received little attention. One possibility at Kimheden would have been to include a capillary
break layer underneath the sealing layer to allow the drainage from the mine waste to occur freely,
as suggested for all types of mine waste deposits by Lottermoser (2007).
Another potential risk for the integrity of the sealing layer has been observed directly in the
¿eld, in the form of local caves-in in the protective layer close to the high wall of open pit 1. As
the terrain is sloping (ca. 10% crosswise and 7% lengthwise the pit), this type of cave-in may be
explained by subsidence of the protective layer at the edge of the pit between bedrock and waste
rock, and by the additional effect of water run-off on the surface. The latter effect is probably
enhanced when spring Àood from snow melting occurs. Tremblay and Hogan (2001) recognised
the contrast in performance between a soil cover for mine waste placed on a horizontal terrain and
a sloping terrain. The difference is conditioned by climatic conditions, characteristics of the slope
and of the material. Installation of a vegetation cover on the surface could help stabilising the
sloping ground (Tremblay & Hogan, 2001), but attempts to seed the protective layer with trees at
Kimheden have met a limited success, and the vegetation is scarce on the back¿lled pits.

ĂĐŬĮůůĞĚŵŝŶĞǁĂƐƚĞĂŶĚǁĂƚĞƌƉĂƚŚǁĂLJƐ;ĨŽĐƵƐŽŶďĂĐŬĮůůĞĚŽƉĞŶƉŝƚϭͿ

Dine ǁaste and groƵndǁater taďle in ďaĐŬĮlled open pit 1


Figure 9.a exhibits pro¿le 1, the inverted resistivity pro¿le over the length of back¿lled
open pit 1. The known limits of the pit and the measured level of the groundwater table are
indicated. The electrical resistivity in the back¿lled waste varies in an interval of ca. [10-630]
ohm-m, contrasting with resistivity values of ca. 1 300 ohm-m and higher in the surrounding
bedrock and in the dry cover on top of the waste. Between 30 m and 70 m from the beginning
of the pro¿le, line 1 is situated on the bedrock but very close to the pit. Between 70 m and 100
m, the line is on the edge of the pit. The adjacent low resistive mine waste creates an impression
of decreasing resistivity with depth but is merely an artefact. The same type of side effect is
observed in Figure 11 (in the next sub-section), where resistivity goes down to 700 ohm-m
at the bottom of the pro¿le while the whole line was surveyed in resistive bedrock of 15 000
ohm-m and higher. Side effects close to the pit could be best identi¿ed when plotting all pro¿les
together. Low resistivity (or high conductivity) in the back¿lled waste rock may be explained
by the sulphide content of the rocks, which are partially crushed to ¿nes and moisturised. The
resistivity variations within the waste may be explained by heterogeneous material. In addition
to the petrologic variability of the rocks, particle size and moist content may affect the electrical
properties of the medium. The production or storage of acidic drainage in the sulphide material
may also affect the bulk resistivity by decreasing its values (Campbell & Fitterman, 2000). Most

12
of the studies of tailings ponds and waste rock piles with geo-electrical methods recognise the
heterogeneity of the resistivity in the deposits, and provide the same possible interpretations (e.g.
Placiencia-Gymez et al, 2010; Marttn-Crespo et al., 2010; Anterrieu et al., 2010). Another factor
explaining resistivity variations the back¿lled material at Kimheden is that, along with waste rock,
other types of material such as remains of the industrial facilities from the mine, roads and even
trees were dumped. This additional source of heterogeneity makes it dif¿cult to associate low
resistivity areas in the back¿lled waste to geotechnical properties (particle size and pore water
content), or processes (AMD generation). Lastly, as mentioned earlier, side effects may have a
signi¿cant inÀuence on the resistivity variations in the pro¿le. Nevertheless, areas with very low
resistivity values (under 25 ohm-m) may be, with relatively high con¿dence, correlated to zones
of moisturised ¿ne-grained waste rock, which certainly also are regions of more intense sulphide
oxidation (Strömberg & Banwart, 1999). The origin of the high resistivity zone of over 1 000
ohm-m in the middle of the back¿lled waste (distance between 130 and 170 m), from a depth of
-6 m and downwards, has not been established. It could represent highly resistive material like
concrete from the former mine facilities, or another case of side effect from the adjacent resistive
bedrock, as the pit is narrower in this section.
According to aerial mapping over the mine site before completion of the reclamation, the
Àoor of the pit should be found at a depth of 10-15 m from the current surface of the dry cover.
Image of the rock substratum underneath the pit with resistivity may provide information about
the integrity of the pit Àoor and indicate potential fracturing or leakage. Unfortunately, the pit Àoor
has not been imaged by resistivity, on either of the resistivity pro¿les over the back¿lled pit 1
(Figure 9.a, and Figure 12 in the next sub-section). Therefore, the actual penetration depth of the
signals has to be lower than indicated on the inverted sections.
Since GPR signals are attenuated when they travel through materials of high electrical
conductivity, the radar data may also serve to localise highly conductive bodies in a more resistive
background. Use of GPR at 50 MHz to identify the upper limit of conductive zones by signal
attenuation has also been noted by Paterson (1997). In our study, 50 MHz radar data revealed the
position of the back¿lled pit (Figure 9.b). Probably due to the rapid decline of the radar waves
in the conductive material, no signature of the pit Àoor could be observed with GPR either. The
same theory is proposed to explain why no reÀection from the groundwater table in the spoil could
be detected in the radargrams. Limited results of the GPR application in highly conductive mine
waste deposits have also been observed by Campbell and Fitterman (2000) and Gymez-Ortiz et al.
(2010).
The groundwater table in back¿lled pit 1 was not imaged by resistivity (pro¿le 1 on
Figure 9.a). In an attempt to evaluate the performance of the DC resistivity method to map the
contaminated water in the back¿lled pit, a conceptual model was created with the modelling
software Res2Dmod (Figure 10.a). The model was built from reference data of the depth of the
pit Àoor, dry cover and the groundwater table together with the interpreted conductive anomalous
zones of sulphidic waste in the inverted resistivity section (Figure 9.a). The bulk resistivity ȡ of the
saturated zone in the back¿lled pit was estimated from measured conductivity of the pore water
using Archie’s law (Archie, 1942):

 = o   m (1)

where ȡo denotes pore water resistivity in ohm-m, Ș denotes the porosity of the waste rock, and m
denotes the cementation grade. Estimations of ȡ were done with ȡo varying in an interval of [14-
20] ohm-m according to conductivity measurements in the groundwater wells in the back¿ll, Ș in
an interval of [20-40] % and m in the interval [1.3-1.7]. The mean bulk resistivity obtained was ca.

13
a
-20 20 60 100 140 180 220 260
Intersection with Groundwater wells
profile 4
Dry cover
Groundwater
table

Bedrock Backfilled open pit 1 Bedrock

Bedrock Edge Backfilled open pit 1

Figure 9 (a) 2D inverted resistivity pro¿le of line 1, central lengthwise line on back¿lled open pit 1. Note that the
resistivity scale is the same as for pro¿le 4 on Figure 12. (b) 50 MHz GPR pro¿le of the same line with GPR. Both
pro¿les were corrected for topography. High conductivity in the back¿lled waste marks the position of the back¿lled
pit on both ¿gures. Note that depths with resistivity and with GPR were estimated according to different models
(inversion model and average of the velocity model, respectively), and cannot be directly compared.

100 ohm-m. The resistivity of the unsaturated zone was chosen a little higher (250 ohm-m). The
apparent resistivity values were then calculated for the multiple gradient array and later inverted
in Res2Dinv. The modelling results (Figure 10.b) indicate that the groundwater table cannot be
resolved, since the intermediate resistivity values of the water-saturated and unsaturated waste
rock are suppressed between the low and high resistivity bodies of the pit. Placiencia-Gymez
et al. (2010) observed that low resistivity in mine waste is more likely to be related to high ion
concentrations in the pore water rather than to moisture of the waste. It is therefore reasonable to
assume that regions of active sulphide oxidation and generation of ion-rich pore water in the waste
will be better imaged by resistivity than groundwater.
Unlike waste rock piles, mine tailings are contained deposits, and are as such interesting
for comparison with the back¿lled waste rock. Geo-electrical investigations of tailings have
succeeded in imaging the Àoor of the tailings impoundments and the underlying bedrock (Faz et
al., 2005; Placiencia-Gymez et al., 2010; Marttn-Crespo et al., 2010; Gymez-Ortiz et al. 2010).
Resistivity investigations on back¿lled pit 1 at Kimheden did not reveal the pit Àoor, neither the
underlying bedrock. The fact that modelling under perfect conditions allowed detection of the
underlying bedrock (Figure 10.b) shows that this is probably a practical issue. The most reasonable
explanation is that due to high resistivity in the upper layer, the contact resistance between the
electrodes and the ground was very high, making it impossible to inject the necessary current in
order to penetrate beneath the conductive zone of the pit.

14
a Resistivity in ohm.m Unit electrode spacing 2.0 m

Dry cover
Unsaturated
Saturated Conductive
waste rock
waste Conductive body
Resistive Saturated
rock body
body waste rock

Bedrock

b
Dry cover
Concuctive bodies

Underlying
bedrock

Figure 10 (a) Conceptual model of back¿lled open pit 1 surrounded by homogeneous bedrock on a 200 m pro¿le. (b)
Inverted model with DC resistivity from the previous conceptual model, expected to represent the conditions observed
on pro¿le 1 (Figure 9.a). Note that the resistivity scale goes up to 1 000 ohm-m only, as the conceptual model used a
maximum resistivity of 1 000 ohm-m for the bedrock and the dry cover.

&raĐtƵres and groƵndǁater Ňoǁ paths


Geotechnical drill core logging and archive documents from Kimheden indicate that the
rock substratum close to the pits is generally fractured, and presents individual larger scale cracks
accounting for the major part of water inÀow to the pits. Evidence of fractured bedrock or single
fractures was therefore a special focus of the geophysical investigations. The objectives were to
identify potential zones of water and oxygen ingress into the back¿lled material, and pathways of
contaminated drainage from the pits. With GPR, no evident signature of fractures in the bedrock
could be found. With resistivity, some suggestions may be formulated.
A hydrological report from the mining company (Rosén & Wilske, 1994) indicates that
before back¿lling, a major source of water inÀow to open pit 1 occurred in the north-eastern end
of the pit through a large fracture of same direction as the pit. Before completion of back¿lling, it
was intended to seal this fracture with clayey till directly on the wall of the pit. Resistivity data in
this area (Figure 9.a) indicates that the limit between the pit and the surrounding bedrock is sharp.
A small spot of relatively lower resistivity can be observed at the end of the pro¿le, 12 m under
the surface (= 0 m depth on the scale of the ¿gure), which may originate from a groundwater Àow
that would be disconnected from the pit. The quality of the data at the end of the pro¿le does not
allow reliable conclusions, but the available model is, however, not indicative of an inÀow of
water through the pit wall there.
Line 2 is situated on the rock substratum upstream of back¿lled open pit 1 (Figure 4). The
resistivity in the bedrock and the overlying natural till cover varies on average between 15 000

15
and 100 000 ohm-m (Figure 11), whereas the topsoil layer is less resistive (ca. 4 500 ohm-m), and
the side effect from the nearby pit generates resistivity values down to 700 ohm-m in the bottom
of the pro¿le. Since the lithology is rather uniform along the pro¿le, the relatively large horizontal
variations in the resistivity of the bedrock may indicate weaker zones of fractured medium, but no
evident single water-saturated fracture could be observed.

Intersection with Topsoil


profile 4

Weaker
bedrock?

side effect from


backfilled pit 1

Figure 11 2D inverted resistivity pro¿le of line 2, on the bedrock upstream of back¿lled open pit 1, and parallel to it.
Note that the resistivity scale goes up to 100 000 ohm-m.

Line 4 crosses the pit on its width and extends along the slope of the hill downward the pit
(Figure 4). On the inverted section (Figure 12), the low pit wall is marked by a sharp limit between
low resistivity in the back¿lled waste and higher resistivity in the bedrock, but the signature of
the high wall is indistinct. A zone of relatively low resistivity (150 ohm-m) is found just before
the known position of the high wall, and suggests that the high wall is a much weaker barrier
compared to the low wall, at least in this section of the pit. Such feature suggests that the high wall
could be fractured and constitute a source of water inÀow to the pit in this section.

Intersection with
profile 2
Intersection with
profile 1
Intersection with
Possible profile 3
seepage through
the cover Collection ditch
Possible
water inflow
to the pit

Bedrock Pit 1 Seepage area

Figure 12 2D inverted resistivity pro¿le of line 4, crossing back¿lled open pit 1, and extending on the sloping terrain
downstream. Note that the resistivity scale is the same as for pro¿le 1 on Figure 9.a.

16
Still on pro¿le 4, the high-resistivity layer on top of the pit (ca 10 000 ohm-m) corresponds
to the dry cover. Downstream of the pit (from 40 m to the end of the pro¿le), the resistivity
on the surface is much lower (less than 2 500 ohm-m). This area is covered by peat and rock
outcrops, and is constantly humid. It is interpreted as the seepage area from the back¿lled pit.
Water on the surface is ion-rich, with electrical conductivities of ca. 600 ȝS.cm-1. This seepage area
is also visible on pro¿le 3, which is situated on the bedrock downstream of the pit, parallel to it
(Figure 13). The two resistivity sections allowed delimiting the seepage area: 0±110 m on line 3,
and 40±200 m on line 4. Pro¿les 3 and 4 show that the contaminated drainage occurs mostly on
the surface and in the subsurface. Resistivity variations suggest that the deeper bedrock might be
fractured, but larger pathways of contaminated water as observed on the surface do not seem to
occur deeper in the ground. One lower resistivity anomaly is observed in the deeper bedrock, at
a horizontal distance of ca. 170 m at the bottom of pro¿le 3 (Figure 13). However, the pseudo-
section of apparent resistivity data suggests that this is likely to be an artefact. Pro¿le 4 intersects
a ditch supposed to divert contaminated water to a treatment pond at the bottom of the hill
(Figure 12). Low resistivity values of less than 100 ohm-m are, nevertheless, found on the surface
of the slope even after this ditch (from 150 m to the end of the pro¿le), indicating that the ditch
fails to retain water entirely, as a part of it runs farther down the slope.

Intersection with
profile 4

Artefact?

Seepage area Bedrock

Figure 13 2D inverted resistivity pro¿le of line 3, on the bedrock downstream of back¿lled open pit 1, and parallel
to it. Note that the resistivity scale starts from 200 ohm-m (unlike pro¿les 1 and 4).

Reference data and the resistivity sections tend to show that the rock substratum
surrounding the pit is fractured, allowing groundwater transport to and from the pit, though mostly
in the subsurface. These conditions may contribute to an insuf¿cient con¿nement of the waste
from water and oxygen. As it is known from groundwater level measurements, a large part of
the back¿lled waste is not saturated with water. Consequently, the presence of fracture zones
in the subsurface may allow oxygen access to the pit. Continuous Àow of water through the pit
may therefore result in persistent oxidation of the sulphidic waste, and wash-out of the oxidation
products. The problem of fractured bedrock surrounding the pit had already been raised before
completion of the reclamation by Rosén and Wilske (1994), who even mentioned sealing of the
fractures as a reclamation alternative to application of a dry cover. They recognised that this option
would result in a much more effective reduction of the water inÀow to the pit compared to the dry
cover, as most of the inÀow occurs as groundwater. They considered, however, that decrease of
the oxygen diffusion to the waste with the dry cover was more determining in decreasing AMD
generation than decreasing water inÀow. They did not mention, however, the possibility of oxygen
ingress through the fractures in the subsurface.

17
ŽŶĐůƵƐŝŽŶƐ
Sources of oxygen to the back¿lled waste could not be clearly identi¿ed with the
geophysical methods. However, as no evident sign of deterioration in the sealing layer was found,
and we know the existence of fractures connected to the pits, the hypothesis of oxygen transport
through unsaturated fractures or oxygenated groundwater inÀow would best explain the persistent
sulphide oxidation in the waste. Although the sealing layer did not show any indication of major
irregularity, risks of damage in the long term were recognised. Models of the depth of the sealing
layer constructed with GPR showed variations of up to 1 m in the thickness of the protective layer
on back¿lled open pits. These variations imply that certain zones may have insuf¿cient thickness
to protect the sealing layer from frost action. With DC resistivity, seepage from the back¿lled
waste through the cover has been identi¿ed as a source of erosion of the sealing layer. Another
de¿ciency of the mitigation actions recognised with resistivity is the failure of the collection ditch
to properly divert the mine drainage for treatment.
Groundwater inÀows to back¿lled pit 1 surveyed with resistivity could not be clearly
imaged. Nevertheless, indications of fractured bedrock upstream of the pit and a possible inÀow of
groundwater through the pit wall in one pro¿le could be observed. The seepage of contaminated
water from the pit was, on the other hand, successfully imaged. The seepage area from the pit
could be delimited, and the mine drainage was observed to Àow mostly on the ground and in the
subsurface. The lateral Àow of water from the pit causes the wash-out of oxidation products from
the waste. Resistivity also evidenced the heterogeneity of the waste, but we showed that even in
perfect conditions of measurement, the groundwater could not have been identi¿ed in the back¿ll,
due to the suppression of its signature by surrounding high and low conductive anomalies.

ZĞĨĞƌĞŶĐĞƐ
Abu-Zeid, N., Bianchini, G., Santarato, G., & Vaccaro, C. (2004). Geochemical
characterisation and geophysical mapping of Land¿ll leachates: the Marozzo canal case study
(NE Italy). Environmental Geology, 45(4), 439-447.
Adu-Wusu, C., & Yanful, E. K. (2006). Performance of engineered test covers on acid-
generating waste rock at Whistle mine, Ontario. Canadian Geotechnical Journal, 43(1), 1-18.
Anterrieu, O., Chouteau, M., & Aubertin, M. (2010). Geophysical characterization of the
large-scale internal structure of a waste rock pile from a hard rock mine. Bulletin of Engineering
Geology and the Environment, 69(4), 533-548.
Archie, G. E. (1942). The electrical resistivity log as an aid in determining some reservoir
characteristics. Metallurgical and Petroleum Engineers 146, 54-62.
Årebäck, H., Barrett, T. J., Abrahamsson, S., & Fagerström, P. (2005). The Palaeoproterozoic
Kristineberg VMS deposit, Skellefte district, northern Sweden, part I: geology. Mineralium
Deposita, 40(4), 351-367.
Axelsson, C.-L., Ekstav, A., Holmén, J., & Jansson, T. (1991). Efterbehandling av
sandmagasin i Kristineberg, Hydrogeologiska förutsättningar för åtgärdsplan: Lakvattenbalanser
och vittringsbegränsande åtgärder. Golder Geosystem AB. In Swedish.
Bambic, D. G., Alpers, C. N., Green, P. G., Fanelli, E., & Silk, W. K. (2006). Seasonal and
spatial patterns of metals at a restored copper mine site. I. Stream copper and zinc. Environmental
Pollution, 144(3), 774-782.

18
Bergström, J. (1997). Development of a geophysical method for investigating and
monitoring the integrity of sealing layers on mining waste deposits: ¿nal report. Stockholm: AFR,
Swedish Environmental Protection Agency.
Brake, S. S., Connors, K. A., & Romberger, S. B. (2001). A river runs through it: impact of
acid mine drainage on the geochemistry of West Little Sugar Creek pre- and post-reclamation at
the Green Valley coal mine, Indiana, USA. Environmental Geology, 40(11-12), 1471-1481.
Buselli, G., & Lu, K. L. (2001). Groundwater contamination monitoring with multichannel
electrical and electromagnetic methods. Journal of Applied Geophysics, 48(1), 11-23.
Campbell, D. L., & Fitterman, D. V. (2000). Geoelectrical methods for investigating mine
dumps. In: ICARD 2000 - proceedings from the Fifth International Conference on Acid Rock
Drainage, Denver, USA, pp.1513-1523.
Dahlin, T., & Zhou, B. (2006). Multiple-gradient array measurements for multichannel 2D
resistivity imaging. Near Surface Geophysics, 4(2), 113-123.
Dahlin, T., Rosqvist, H., & Leroux, V. (2010). Resistivity-IP mapping for land¿ll
applications. Near Surface Geoscience. 28(8), 101-105.
Edström & Schönfeldt AB (1996). Kristinebergsgruvan, Efterbehandlingsarbeten. Edström
& Schönfeldt AB and Boliden AB. In Swedish.
Faz, A., Aracil, E., Marttnez-Pagin, P., Acosta, J. A., Marttnez-Marttnez, S., Maruri,
U., & Marttnez-Segura, M. A. (2005). Geochemical and geophysical techniques application to
characterize a mining silt pond from Cartagena- Union (Murcia): potential pollution pathways. In:
IMWA 2005 - proceedings of the International Mine Water Association Congress, Oviedo, Spain,
J. Loredo & F. Pendis (Eds.), University of Oviedo, pp.295-302.
Gymez-Ortiz, D., Marttn-Crespo, T., & Esbrt, J. M. (2010). Geoenvironmental
characterization of the San 4uintin mine tailings, Ciudad Real (Spain). Dyna-Colombia, 77(161),
131-140.
Hannington, M. D., Kjarsgaard, I. M., Galley, A. G., & Taylor, B. (2003). Mineral-chemical
studies of metamorphosed hydrothermal alteration in the Kristineberg volcanogenic massive
sul¿de district, Sweden. Mineralium Deposita, 38(4), 423-442.
Hellman H., & Lokrantz H. (2008). Kimhedengruvan, Installation av grundvattenrör vid
nedlagt och igenfyllt dagbrott vid Kimheden. Berggeologiska Undersökningar AB. In Swedish.
Höglund, L. O., & Herbert, R. (Eds.), Lövgren, L., Öhlander, B., Neretnieks, I., Moreno,
L., Malmström, M., Elander, P., Lindvall, M., & Lindström, B. (2004). MiMi - Performance
assessment: main report. MiMi report 2003:3, Mitigation of the environmental impact from mining
waste programme (MiMi). Luleå, Sweden: MiMi Print.
Holmström, H., Salmon, U. J., Carlsson, E., Petrov, P., & Öhlander, B. (2001). Geochemical
investigations of sul¿de-bearing tailings at Kristineberg, northern Sweden, a few years after
remediation. Science of the Total Environment, 273(1-3), 111-133.
INAP (the International Network for Acid Prevention) (2009). Global Acid Rock Drainage
Guide (GARD Guide). Chapter 6.2. http://www.gardguide.com/.
Jeffree, R. A., Twining, J. R., & Thomson, J. (2001). Recovery of ¿sh communities in the
Finniss River, northern Australia, following remediation of the Rum Jungle uranium/copper mine
site. Environmental Science & Technology, 35(14), 2932-2941.

19
Loke, M. H., Acworth, I., & Dahlin, T. (2003). A comparison of smooth and blocky
inversion methods in 2D electrical imaging surveys. Exploration Geophysics 34(3), 182-187.
Lottermoser, B. G. (2007). Mine wastes: characterization, treatment and environmental
impacts. (2. ed.). Berlin, Heidelberg: Springer-Verlag.
Malmström, M., Werner, K., Salmon, U. & Berglund, S. (2001). Hydrogeology and
geochemistry of mill tailings impoundment 1, Kristineberg, Sweden: Compilation and interpretation
of pre-remediation data. MiMi report 2001:4, Mitigation of the environmental impact from mining
waste programme (MiMi). Luleå, Sweden: MiMi Print.
Marttn-Crespo, T., De Ignacio-San José, C., Gymez-Ortiz, D., Marttn-Velizquez, S., &
Lillo-Ramos, J. (2010). Monitoring study of the mine pond reclamation of Mina Concepcion,
Iberian Pyrite Belt (Spain). Environmental Earth Sciences, 59(6), 1275-1284.
Nobes, D. C., Armstrong, M. J., & Close, M. E. (2000). Delineation of a land¿ll leachate
plume and Àow channels in coastal sands near Christchurch, New Zealand, using a shallow
electromagnetic survey method. Hydrogeology Journal, 8(3), 328-336.
Paterson, N. (1997). Remote Mapping of Mine Wastes. In: proceedings of Exploration
1997 - Fourth Decennial International Conference on Mineral Exploration, A. G. Gubins (Ed.),
pp.905-916.
Placencia-Gymez, E., Parviainen, A., Hokkanen, T., & Loukola-Ruskeeniemi, K. (2010).
Integrated geophysical and geochemical study on AMD generation at the Haveri Au-Cu mine
tailings, SW Finland. Environmental Earth Sciences, 61(7), 1435-1447.
Poisson, J., Chouteau, M., Aubertin, M., & Campos, D. (2009). Geophysical experiments
to image the shallow internal structure and the moisture distribution of a mine waste rock pile.
Journal of Applied Geophysics, 67(2), 179-192.
Porsani, J. L., Filho, W. M., Elis, V. R., Shimeles, F., Dourado, J. C., & Moura, H. P. (2004).
The use of GPR and VES in delineating a contamination plume in a land¿ll site: a case study in SE
Brazil. Journal of Applied Geophysics, 55(3-4), 199-209.
Rosén, L., & Wilske, Å. (1994). Beräkningar av metallÀöden och förslag till åtgärder vid
Kimhedengruvan, Kristineberg. Scandiaconsult Väst AB. In Swedish.
Rubin Y., & Hubbard S. S. (2005). Introduction to hydrogeophysics. In: Y. Rubin & S. S.
Hubbard (Eds.), Hydrogeophysics (pp.3-21). Dordrecht, The Netherlands: Springer.
Runkel, R. L., Bencala, K. E., Kimball, B. A., Walton-Day, K., & Verplanck, P. L. (2009).
A comparison of pre- and post-remediation water quality, Mineral Creek, Colorado. Hydrological
Processes, 23(23), 3319-3333.
Strömberg, B., & Banwart, S. (1999). Weathering kinetics of waste rock from the Aitik
copper mine, Sweden: scale dependent rate factors and pH controls in large column experiments.
Journal of Contaminant Hydrology, 39(1-2), 59-89.
Tremblay, G. A., & Hogan, C. M. (2001). MEND Manual, Volume 4 - Prevention and
control, MEND 5.4.2d, Mine Environment Neutral Drainage programme (MEND).
Van Dam, R., Gutierrez, L., McLemore, V., Wilson, G. W., Hendrickx, J., & Walker, B.
(2005). Near surface geophysics for the structural analysis of a mine rock pile, northern New
Mexico. In: paper presented at the 2005 National Meeting of the American Society of Mining and
Reclamation, ASMR, 3134 Montavesta Rd., Lexington, KY 40502, pp.1178-1201.

20
Villain, L., Alakangas, L., & Öhlander, B. (2010). Geochemical evaluation of mine water
quality in an open-pit site remediated by back¿lling and sealing. In: IMWA 2010 - proceedings
of the International Mine Water Association Symposium, Sydney, Canada, Wolkersdorfer C. &
Freund A. (Eds.), Cape Breton University Press, pp.515-518.

21

You might also like