A Laboratory Study of ARD Neutralization and Prevention by Alkaline Rest Products: Waste rock from Maurliden

A Laboratory Study of ARD Neutralization

and Prevention by Alkaline Rest Products

Waste rock from Maurliden

Elin Andersson

Master of Science in Engineering Technology Natural Resources Engineering

Luleå University of Technology

II

ABSTRACT

Boliden’s Maurliden mine, located in Northern Sweden, is a polymetallic VMS deposits with pyrite as the dominant sulfide mineral. At closure, waste rock will be backfilled into the open pit, covered, and the pit will then be allowed to flood to reduce oxidation. Boliden is considering the addition of alkaline material as a safeguard to buffer potential acid rock drainage (ARD). Therefore, characterization and leaching tests of waste rock were completed as part of this study. Characterization of alkaline rest products and leach tests with mixtures of waste rock and alkaline rest products were performed. Acid generation potential of the waste rock was determined together with neutralization experiments with alkaline rest products. Leach tests of the waste rock were conducted in 1m3 tanks under saturated and dry conditions, dewatered once a week. The pH in the water saturated tank was observed to increase from 2.4 to 5.6 as maximum over a 4 months period. pH was observed to decrease in the water saturated tank after 2 months of sampling to pH 3.5, most likely due to a higher evaporation and a lowered water table. A decrease of the pH in the dry tank from 6.7 to 1.4 over an 8 months period, with a minimum pH of 0.9 was observed. ICP-analysis of the leachate from the dry tank showed a significant increase in concentrations of elements such as S and Fe from µg/L to g/L, while the concentrations were constant at µg/L in the saturated tank. This shows the importance of saturated conditions to reduce the oxidation of sulfides.

Alkaline rest products mesa lime, fly ash (FA) and green liquor dregs (GLD) from paper and pulp mills together with slag from steel industry were evaluated for their ARD neutralization potential. Waste rock crushed down to <6mm were mixed with different fractions of alkaline waste, 0.5%-5%, and leached in small scale for 1, 3, 8, 21 and 65 days. The 5% mixtures were leached for 115 days. After 65 days the analyzed pH was circum-neutral for most materials expect AOD slag with pH >10, mesa lime with one sample with a pH at 4 and Iggesund FA with 2 samples with a pH at 3. Samples with pH <7 showed increased concentrations of elements such as As, Cr, Fe and S, and low element concentrations were reported in all samples with pH >7. This shows how important it is to keep pH in the solution at 7 to 9, to minimize the leaching of elements. After 65 days of leaching, almost all water had been replaced with fresh water. It takes many years before all water in an open pit is replaced with new water. The material in the pit is significantly coarser grained than the size fraction 6mm used in the experiment, which also influence the oxidation of the waste rock. A material with fine particle size has a larger surface area, which increases the oxidation.

The results of this master thesis are a preliminary step in evaluating the neutralization capacity of alkaline rest products and their effect on leachate quality when mixed with waste rock. The investigation could be used in further studies of neutralization. Rest products other than mesa lime can be used with the purpose to neutralize ARD when mixing the alkaline material with waste rock and backfill it to an open pit. Further investigations will be undertaken to determine the long-term ARD neutralization capacity of the alkaline wastes for use in ARD mitigation.

III

SAMMANFATTNING

Maurliden-gruvan ägs av Boliden Mineral AB och är en av tre gruvor i det så kallade Skellefte fältet beläget i norra Sverige. Malmförekomsten innehåller koppar, bly, silver och guld. Pyrit är det dominerande sulfidmineralet. Efter stängning av gruvan kommer gråberg att återföras till dagbrottet, täckas och tillåtas vattenfyllas naturligt. Detta kommer att reducera oxidationen av det sulfidrika gråberget. För att säkerställa att oxidationen reduceras överväger Boliden att även tillsätta alkaliskt material till gråberget, för att förhindra uppkomsten av surt lakvatten. Inom denna studie, har en karakterisering av gråberg från Maurliden gruvan gjorts för att bestämma dess kemiska sammansättning och syrabildande potential. Laktest i 1m3 tankar har utförts under vattenmättade och torra förhållanden. De torra tankarna vattnades en gång i veckan för att efterlikna naturliga regnförhållanden. Varje vecka analyserades pH och elektrisk konduktivitet. Den kemiska sammansättningen på lakvattnet provtogs och analyserades varannan vecka under en 8 månaders period. Resultatet visade en pH ökning från 2,4 till maximalt pH 5,6 under en 4 månaders period, i den vattenmättade tanken. Efter ytterligare två månader minskade pH till 3,5; troligtvis på grund av minskad vattennivå i tanken till följd av ökad avdunstning. I den torra tanken minskade pH från 6,7 till 1,4, med ett minimum på pH 0,9; under en 8 månaders period. ICP-analyser visade konstant låga koncentrationer i den vattenmättade tanken, medan element som Al, As, Fe och S ökade i koncentrationer från µg/l till g/l i den torra tanken. Detta visar tydligt hur viktigt vattenmättade förhållanden är för att minimera sulfid oxidationen.

En karakterisering av alkaliska restprodukter som mesa-kalk, flygaska (FA) och grönlutslam (GLS) från papper- och massabruk, samt AOD slag från stålindustrin har gjorts. Restprodukternas potential att neutralisera surt lakvatten har undersökts. Gråberg krossat till <6mm har blandats med de olika restprodukterna i 0,5 % - 5 % blandningar, och lakats i liten skala (50ml prover) under 1, 3, 8, 21 och 65 dagar. 10ml lakvatten har bytts ut vid varje tillfälle. 5 % blandningarna har lakats i 115 dagar. Efter 65 dagars lakning var pH nära neutrala förhållanden för alla material förutom AOD slag med pH >10, mesa-kalk där ett prov sjunkit till pH 4, och Iggesund FA där två prover sjunkit till pH 3. Analyser av de blandningar som höll ett pH <7, visade ökande koncentrationer av element som As, Cr, Fe och S, medan låga koncentrationer rapporterades i blandningar med ett pH >7. Detta visar hur viktigt det är att hålla pH mellan 7 och 9, för att minska urlakningen av element. Kemisk analys av lakning med Billerud FA och GLS inkluderades inte i denna studie. Efter 65 dagars lakning, hade nästan allt vatten ur 50ml proverna bytts ut mot nytt milli-Q vatten. Det tar många år innan allt vatten är utbytt i ett vattenfyllt dagbrott. Materialet i dagbrottet är också betydligt mer grovkornigt än materialet krossat till <6mm som användes vid lakningen. Kornstorleken på materialet påverkar oxidationen, då mer yta är tillgängligt för oxidation för finkornigare material.

Resultaten från detta examensarbete är ett första steg i att undersöka alkaliska restprodukters neutraliseringsförmåga, samt visa på effekter då blandningar med restprodukter och gråberg lakas. Resultaten i denna studie kan användas i vidare undersökningar, och visar att andra restprodukter än mesa kalk kan användas i syfte att begränsa uppkomsten av surt lakvatten, i syfte att blandas tillsammans med gråberg och återförs till ett dagbrott. Vidare undersökningar krävs för att undersöka långtidseffekterna av användning av dessa restprodukter.

IV

ACKNOWLEDGEMENT

This master thesis is the final part within the Master Programme in Natural Resources Engineering, with a focus on environment and water, at Luleå University of Technology. The master thesis corresponds to 30hp and has been running from January to September 2012, and is a project in collaboration with the division of Geosciences and Environmental Engineering LUT andBoliden Mineral AB.

I am very grateful to my supervisors Lena Alakangas (LUT) and Seth Mueller (Boliden Mineral AB) for guiding and supporting me in this project. Thank you for your motivation, inspiration and great ideas to form this master thesis in the best of ways.

I want to say Thank you to my examiner Björn Öhlander for your support to this report. Lu Jimnei for your assistance and help with the laboratory work. Jaana Ekblom, my friend and colleague, for your help and support both professionally and personally, as well as your thoughts as opponent to my work. Others I would like to thank are Anton Lundkvist (Boliden Mineral) for helping me with questions about at Maurliden mine site, Bengt Jonasson for guiding me at Maurliden mine site, Danil Korelskiy for assisting me with the XRD equipment, Milan Vnuk for providing figures to this report and Fredrik Engström and Ida Strandkvist for helping me with SEM preparation. I am also grateful to all of you at the environmental laboratory (LUT) and ALS Environmental in Luleå.

I want to thank my family for your love and that you believe in me and what I do. And finally, a huge Thanks to my beloved Frida and Petter for your love, patience and support.

Elin Andersson

V

TABLE OF CONTENT

1.1. Acid Rock Drainage ... 1

1.2. Alkaline rest products ... 3

1.3. Objectives ... 3

1.4. Limitations ... 4

2. BACKGROUND ... 5

2.1. Maurliden mine site ... 5

2.1.1. Geology ... 5

2.1.2. Waste rock ... 6

2.1.3. Overburden and till ... 6

2.1.4. Water treatment ... 6

2.1.5. Effects from the remediated mine ... 7

3. MATERIALS ... 9

3.1. Waste rock ... 9

3.2. Alkaline rest products ... 9

4. METHODS ... 10

4.1. Sampling of waste rock ... 10

4.2. Characterization of waste rock ... 10

4.2.1. Mineralogy ... 10

4.2.2. Chemical composition ... 10

4.2.3. Neutralization potential ... 10

4.3. Sampling and characterization of ARD leachate ... 11

4.3.1. Sampling of leachate ... 11

4.4. Laboratory Leaching Experiments ... 12

4.4.1. Experiment 1, Neutralization of ARD ... 12

4.4.2. Experiment 2, Leaching of waste rock mixed with alkaline rest products ... 12

5. RESULTS ... 13

5.1. Characterization of Waste Rock ... 13

5.1.1. Mineralogy ... 13

VI

5.1.3. Neutralization potential ... 14

5.2. Characterization of leachate ... 15

5.2.1. Leachate from 1m3 tanks ... 15

5.3. Laboratory leaching experiments ... 18

5.3.1. Titration with ARD ... 19

5.3.2. Leaching experiment ... 20

6. DISCUSSION ... 26

6.1. Characterization of waste rock ... 26

6.2. Characterization of leachate ... 27

6.2.1. Dry conditions ... 27

6.2.2. Water saturated conditions ... 27

6.3. ARD neutralized by alkaline rest products ... 28

6.3.1. Alkaline rest products mixed with waste rock ... 30

7. CONCLUSIONS ... 34

7.1. Further research ... 35

8. REFERENCES ... 36

9. APPENDIX ... 39

9.1. Alkaline rest products ... 39

9.1.1. SCA mesa lime and AOD slag ... 39

9.1.2. AOD slag ... 39

9.1.3. Billerud and Iggesund fly ash and green liquor dregs ... 40

9.2. XRF ... 41 9.3. XRD ... 42 9.3.1. Sample no 1 ... 42 9.3.2. Sample no 2 ... 43 9.3.3. Sample no 3 ... 44 9.4. Experiment 1... 45 9.4.1. L/S 110 ... 45 9.4.2. L/S 27 ... 46 9.5. Experiment 2... 47

9.5.1. Mixing waste rock with alkaline materials ... 47

9.5.2. pH and electrical conductivity ... 48

9.5.3. ALS analyzes ... 50

VII

DEFINITIONS

ARD Acid Rock Drainage, low pH (pH <6) water formed from the

oxidation of sulfide minerals.

Waste rock Rock and overburden (i.e. till/moraine) excavated and mined from

surface and underground operations, to access the ore at a mine site.

AOD slag Argon Oxygen Decarburization, a slag from steelmaking

FA Fly ash, a rest product from paper and pulp mills

ML Mesa lime, a by-product from where mesa is burned in the pulp

mill processes

GLD Green liquor dregs, a rest product from paper and pulp mills

ELEMENTS Al Aluminum As Arsenic Ba Barium Be Beryllium C Carbon Ca Calcium Cd Cadmium Co Cobalt Cr Chromium Cu Copper Fe Iron Hg Mercury K Potassium Mg Magnesium Mn Manganese Mo Molybdenum N Nitrogen Na Sodium Nb Niobium Ni Nickel P Phosphorus Pb Lead S Sulfur Sb Antimony Sc Scandium Si Silicon Sr Strontium Ti Titanium U Uranium V Vanadium W Tungsten Y Yttrium Zn Zinc Zr Zirconium CARBONATES CaCO3 Calcite SILICATES SiO2 Quartz SULFIDES CuFeS2 Chalcopyrite FeAsS Arsenopyrite FeS2 Pyrit

1

1. INTRODUCTION

The mining industry generates large volumes of waste rock, which is non-valuable rock that has to be removed to be able to access the ore, in both underground and open pit mining. The amount of waste rock generated is dependent on which phase the mine is in. More waste rock is generated in the beginning of a mine life (Eriksson et al. 2008). More than 4,700M tons of mine wastes were generated in the European Union during 2000 (BRGM 2001). Open pit mining generates more waste rock, overburden and tailings than underground mining. Waste rock is often deposited in heaps at the mine site, and where possible backfilled after closure. Not all mine wastes contain or release contaminants, and therefore do not require any treatment (Lottermoser 2003) and lime is sometimes added to the wastes to prevent acid rock drainage (ARD).

Sweden is a major metal producer. The ore quality is often low graded, which requires larger mine to make a profit. This is turn generates large volumes of mine waste (Lotermoser 2003). 59 million tons of waste rock and tailings were landfilled in 2008, and of that more than half is of a sulfidic nature (Naturvårdsverket 2010). The mined bedrock most often contains sulfides. If sulfidic waste is brought to the surface, it can cause long term environmental problems due to formation of ARD (INAP, 2012). The overburden consists mainly of till and is normally of no environmental concern (Eriksson et al. 2008).

1.1. Acid Rock Drainage

Weathering of the mineral pyrite (FeS2), the most common sulfide mineral present, can

potentially cause long-term environmental issues due to formation of ARD. A leachate with low pH and elevated concentrations of metals and metalloids, can result in significant ecological disruption, decrease the water quality and make it toxic to aquatic life (Lottermoser 2003). When rocks containing sulfide minerals are excavated due to mining and exposed to the atmosphere, the oxidation accelerates because of the exposure of sulfide minerals to air, water and microorganisms. Pyrite is naturally weathered, or oxidized, when it is exposed to atmospheric conditions. It can be geological processes or anthropogenic activities as mining. When the bedrock is crushed and brought to the surface, the waste is no longer in equilibrium with the oxidizing environment. There are several sources for weathering processes: tailings and waste rock piles are just two examples. The oxidation and transport products need water, like rainfall or groundwater, for transport to the recipient. ARD can be transported long distances and cause problems far away from the mine site due to contamination of waters and the receiving environment (INAP 2012).

Oxidation of sulfide minerals present in mine and process waste can be rapid, and in extreme cases the oxidation can result in self-heating and combustion (INAP 2012). The overall pyrite oxidation reaction requires oxygen and water and is usually expressed by reactions (Eq. 1) and (Eq. 2) in oxygenated acidic (approximately pH 4.5 and lower) environment (Lowson 1982):

FeS2 + 7/2O2 +H2O  Fe2+ + 2SO42- +2H+ (Eq. 1)

Fe2+ + 1/4O2 + H+  Fe3+ + 1/2H2O + energy (Eq. 2)

If pH increases more than approximately pH 3, precipitation of ferric hydroxides (Fe(OH)3)

2

Fe3+ + 3H20  Fe(OH)3 + 3H+ (Eq. 3)

Precipitation of dissolved Fe3+ will lower the pH by the release of H+, and less Fe3+ will remain in solution. If Fe3+ are stable in solution (low pH9 it can oxidize pyrite via (Eq. 4), which forms a continuing cycle of Fe2+ production (Eq. 4) and conversion to Fe3+ (Eq. 2) and oxidation of pyrite by Fe3+ (Eq. 4):

FeS2 +14Fe3++ 8H2O  15Fe2+ + 2SO42- + 16H+ (Eq. 4)

The oxidation of pyrite (Eq. 4) continues until either Fe3+, oxygen or pyrite is no longer provided (Singer and Stumm 1970).

The content of sulfides and carbonates (e.g. calcite (CaCO3 and dolomite (CaMg(CO3)2) in

the waste rock are an important factor in the formation and extent of ARD production. Degradation of pyrite is an acid producing reaction (i.e. generation of H+), whereas weathering of calcite is an acid buffering reaction (i.e. consumption of H+). The capacity of water to resist pH changes is commonly known as “buffering capacity”. A buffer is a solution to which an acid can be added without changing the concentration of available H+ ions, and subsequently without changing the pH. Rapid dissolution of carbonates minerals makes them effective acid consumers, and thereby in some circumstances they will neutralize the drainage. The minerals buffer at different pH ranges, buffering of calcite occurs around neutral pH (6.5-7.5):

CaCO3 + H+   Ca2+ + HCO3- (Eq. 6)

Below pH 6.3, the dominant carbonate species is carbonic acid (H2CO3) instead of

bicarbonate (HCO3-). Hence, bicarbonate might form carbonic acid as follows:

HCO3- + H+   H2CO3 (Eq. 7)

Silicates can also contribute to buffering capacity, but have a slow dissolution rate. The silicate minerals provide neutralizing capacity between pH 5 and 6. Other factors that influence the ARD formation other than the type of mineral is the surface area of the waste rock, pore space, infiltration of water and runoff (INAP 2012).

Once the problems with ARD have started it is difficult and costly to make it stop. It is often more cost and material effective to prevent or minimize sulfide oxidation at the source than to treat ARD contaminated waters. When ARD has spread to larger areas more waters will be contaminated and will require treatment. Higher cost for monitoring and treatment of leachate at long distances from the source is needed. Dry or wet covers are also effective ways to control and reduce the oxidation rate of the mine waste. ARD treatment can be classified as an active treatment, where continued addition of chemical reagents to neutralize the waters, active maintenance and monitoring is required. Water treatment techniques often generate large volumes of sludge, due to addition of reagents such as lime. The formed sludge can contain precipitated heavy metals that require additional handling. Other treatment techniques are classified as passive treatment like wetlands which use chemical and biological processes to neutralize acidity and reduce dissolved metal concentrations (Lottermoser 2003). Backfilling and flooding of sulfidic waste rock to an open pit is an additional measure to prevent ARD generation.

3

1.2. Alkaline rest products

Large volumes of alkaline rest products such as mesa lime (ML), fly ash (FA) and green liquor dregs (GLD) are produced at the paper and pulp mills in Sweden (Naturvårdsverket 2010). Those materials are often landfilled at the site. Land filling of material is the least desired option as waste treatment. It is always more desirable to use rest products than use virgin materials, both due to lower costs and from an environmental perspective (Avfall Sverige 2012). Before rest products can be used for some application they need to be, characterized by their chemical composition and leaching properties.

Boliden is considering the addition of alkaline material as a safeguard to buffer potential acid rock drainage (ARD). Mixing of alkaline rest product when backfilling mine waste to the open pit, could be an alternative use for the materials instead of deposit them on a landfill. For this application the materials need to buffer pH and prevent ARD formation. On the other hand the material itself should not leach elements that can be of environmental concern, or cause other problems that are not originated from the waste rock. The positive effects should be greater than the negative. It is important to investigate the properties of the materials before they are used, so they do not cause unexpected problems later on (Eriksson et al. 2008).

Mesa lime is an alkaline material with high carbonate content. It is used today at the Maurliden mine site to buffer acid rock drainage (ARD) in the water treatment process, as an addition in the settling ponds. Mesa lime is a by-product from where mesa is burned in the pulp mill processes, from the so called causticization process, where sodium sulfide and sodium hydroxide are formed together with calcium carbonate, CaCO3 (mesa). The mesa is

then burned in rotary kiln (generally referred to as mesa kiln), in which mesa lime (CaO) is obtained (SCA 2009).

Green liquor dregs (GLD) are a rest product from the combustion of black liquor from the recovery boiler (Billerud 2010). Fly ash is another rest product, which together with sludge are today most often landfilled. Argon Oxygen Decarburization (AOD) slag originates from a converter in the stainless steel production (Ghan 2009).

1.3. Objectives

Maurliden mine site generates large amounts of sulfidic mine waste that needs to be treated and one part in this thesis is to evaluate the acid producing potential of sulfidic mine waste from the site. This master thesis is also a first step to examine if alkaline rest products could be used for the purpose to prevent ARD formation, when backfilling the waste rock to the open pit after mine closure. The study includes following parts:

 Characterization of sulfidic waste rock (from Maurliden mine site)

 Characterization of leachate generated from sulfidic waste rock, both under dry and water saturated conditions

 Investigation of different alkaline rest products (i.e. mesa lime, fly ashes, green liquor dregs and AOD slag) ability to prevent and neutralize ARD generated from sulfidic mine waste, when mixed and leached for 65 days.

4

1.4. Limitations

This master thesis is a project in collaboration with Luleå University of Technology (LUT) and Boliden Mineral AB. The project has formally been running from January to September but it started with two visits to the Maurliden mine site, the 20th of October and 7th of November 2011. Laboratory work has been ongoing from December 2011 to August 2012, and the report has been written continually from January to October 2012. Leach tests, XRD and ABA tests have been performed at LUT, and all ICP analyses have been performed at ALS Environmental in Luleå. Waste rock has been crushed at ALS in Piteå. Waste rock from Maurliden mine site has been chosen due to its high content of sulfur.

5

2. BACKGROUND

2.1. Maurliden mine site

Maurliden Cu-Zn-Au-Ag mine is one of three mine sites in the mineral-rich Skellefte field, northern Sweden operated by Boliden Mineral AB (Boliden). Almost 30 mines have been operated by Boliden in this area since 1920s (Eriksson, et. al., 2008).

Maurliden mine site consist of two open pits named Maurliden and Maurliden Östra (main pit and east pit). Maurliden is an open pit mine and the production started in year 2000. The ore contains zinc, copper, gold and silver. A new push-back began in 2008 and the final depth of the open pit will be approximately 150 m. Maurliden mine is situated in a region characterized by a continental inland climate with cold winters and warm summers. Average annual temperature is about +1°C. The annual rainfall varies between 500-900mm/year with a corrected mean value of about 750mm/year. Approximately 50% of the precipitation falls as snow, with a snow cover from early November to late May. About 350-400mm/year are potentially evaporated, which is an amount significantly below the annual rainfall. The area, which is relatively hilly, is in Maurbäckens catchment and consists mainly of peat and till soil. The till is sandy silty and has a thickness of about 6-18 meters (Eriksson et al. 2008).

At closure of Maurliden mine, waste rock will be backfilled to the open pit and be allowed to flood. The waste rock will be mixed with some alkaline material to prevent acid generation from the sulfidic mine waste and keep the pH at relatively high levels, i.e. at a pH between 7 and 9 (Eriksson et al. 2008).

2.1.1. Geology

Main source for the metals mined in Maurliden (zinc, copper, lead, silver and gold) has its origin from volcanic-associated massive sulfide (VMS) ores. Maurliden area consists of an approximately 8x6 km wide area with four polymetallic network to massive sulfide deposits. These volcanic rocks belong to the so-called Skellefte Group, which is a sequence of volcanic rocks in the Skellefte field that was deposited at sea bottom about 1.89 billion years ago. The host rock is quartz-feldspar pohyritic rhyolite, and strongly quartz-feldspar porphyritic rhyolite also occurs as isolated bodies scattered through most of the Maurliden (Montelius 2005). Pyrite is the dominating sulfide while the main ore minerals are sphalerite and chalcopyrite. Other identified ore minerals are in decreasing order arsenopyrite, Pb-Sb-sulfosalts, tetrahedrithe, chalcopyrite, galena, pyrrhotite, gudmundit and pyrargyrit. Silver occurs mainly in tetrahedrithe(Eriksson et al. 2008).

The ore body in Maurliden mine falls almost vertically, maximum extension is to the north and to the south. It consists of an approximately 300m long and up to 60m wide body of

Fig. 1 Maurliden mine site in Boliden mine

6

pyrite that contains value metals zinc (about 4.7%), copper (about 0.2%), gold (about 1g/ton) and silver (about 60g/ton). From an environmental perspective it is the ore sulfur content (up to about 35%) and arsenic content (about 1%) which is of importance in addition to metals. The side dip is about 90° and the ore body is not defined at depth. The ore had an up to 20m thick till cover that was removed before the mining started (Eriksson et al. 2008).

2.1.2. Waste rock

During 2011, 220ktons of waste rock was produced in Maurliden. In total 4.9million tons of waste rock will be removed to access the ore, and 60% of the waste rock is expected to be acid producing. This waste rock contains high quantities of sulfides, from 1% to 10% (Eriksson et al. 2008). The sulfur content at the area are at some points extremely high (>30%), mostly in pyrite. This sulfidic waste rock has potential to produce acidic drainage with high content of metals both in short and long term, if it is not treated right. A smaller part of the waste rock, about 40%, is not considered to be potentially acid producing, due to lower sulfur content. Sampling and analysis of the drill cuttings make it possible to determine the acid producing potential of the waste rock. If it is not acid producing then the waste rock can be handled separately as non-acid producing waste rock. The potentially acid forming waste rock is loaded and transported to the waste rock piles located north west of the open pit. The deposit is extended to the north and is planned to have an area of 16 hectares and be about 40 m in height. The ground underneath the deposit area is compacted with till layers to significantly reduce the permeability of the ground (Eriksson et al. 2008). Stored waste rock generates an acidic drainage which contains metals, due to acid formation from pyrite.

At closure of the Maurliden mine, waste rock will be backfilled to the open pit and the open pit will be allowed to flood. The waste rock will be mixed with alkaline material with high buffering capacity to prevent acid generation from the sulfidic mine waste and keep the pH near-neutral. Large volumes of alkaline materials are needed to neutralize the ARD. Rest products can be used for this purpose which has the potential to be lower costs. This will also precipitate already dissolved metals and neutralize the acid produced. The open pit will then be covered with till. This is expected to effectively reduce the sulfide weathering rate. The till will provide a diffusion barrier between the water volume above and the waste rock, which further reduces the sulfide weathering rate and also the rate which erosion products are transported out to the surroundings. Potentially contaminated water will be collected and treated before it is transported to the recipient Lilltistelmyrbäcken (Eriksson et al. 2008).

2.1.3. Overburden and till

The overburden, which consist mainly of till, with lesser quantities of soil and peat, are stored at the mine site. Geochemical characterization of the till shows that it contains low sulfur and metal content, and therefore unlikely to be any source of pollution. Storages of till are surrounded by collecting ditches and no eroded material is discharged to the recipient. Collected drainage water will be transported to the treatment plant. Deposited till will occupy an area of 3ha, and 10-15 meters high (Eriksson et al. 2008).

2.1.4. Water treatment

During operation of the mine, the groundwater level is lowered by pumping and the main direction of the groundwater is towards the pit. The mining area is surrounded by shielding and collection ditches which minimize the inflow of fresh water into the mining sector. Less fresh water is therefore mixed with potentially contaminated water from the piles, ore deposit

7

and industrial areas. In this way the amount of contaminated water (i.e. water from deposits, drainage water from open pits and direct water from the industrial areas) that needs to be treated is minimized.

The annual amount of water collected from the deposit, industrial sites and drainage water from the open pit is about 525,000m3, which corresponds to 60m3/h (medium flow). Of this 16m3/h or about 65-70% of the water is drainage water from the open pit (Eriksson et al. 2008).

Table 1 Temporary guidance values for treated water in Maurliden (Environmental permit 2010-06-23)

Object Parameter Condition

Purified mine water Guideline for emissions:

Suspended material ≤ 15mg/L Dissolved content As ≤ 30µg/L Dissolved content Cu ≤ 30µg/L Dissolved content Pb ≤ 30µg/L Dissolved content Zn ≤ 400µg/L Dissolved content Cd ≤ 5µg/L Mineral oil ≤ 0.5mg/L N-tot ≤ 10mg/L Limit: ∑(Cu+Pb+Zn+As) 500kg/year

A larger resultant flow together with unchanged metal concentrations implies that the overall quality of emissions (not levels) increases correspondingly. The current discharge limit for ∑(Cu+Pb+Zn+As) = 500kg/year may be difficult to follow.

Leachate from the waste rock pile can generate problems in the treatment plant because of its composition. The water contains high levels of metals and sulfate, that at certain percentage involvement (>35% leachate and <65% mine water) interferes with the liming step significantly. Water with a better quality is continually mixed with leachate from the open pit and treated under controlled conditions (Eriksson et al. 2008).

When treating acidic and metallic leachate water from the mine site, CaOH2 is added to the

water in the treatment plant. The slaked lime raises the pH whereby the solubility of the dissolved metals is reduced and precipitates in the lime slurry. Generated slurry is pumped into centrifugation, resulting in a dewatered sludge with a consistency like soil with about 25% dry matter. Sludge from the treatment plant in Maurliden will sediment in sedimentation basins and will be moved and deposited together with the waste rock. Treated water is distributed to Skellefte River via the recipient Lilltistelmyrbäcken, north of the mine.

2.1.5. Effects from the remediated mine

When post-treatment of the Maurliden mine is completed, the pumping and treatment of the water will stop. Drainage water from the mine will drain to Maurbäcken, as before the mining. Even after treatment the mine will produce some residual metal loads to the recipient, due to leaching from waste rock piles and also the backfilled open pit. The annual water emissions generated from the post-processed Maurliden mine, over the long term is estimated to be: <0.5kg Cu, <0.5kg Pb, <5kg Zn, <0.5kg As and <0.005kg Cd. These emissions are expected to be handled by the recipient Maurträsket. Assuming no sorption of weathering products either in the open pit or Maurträsket gives a theoretical level in Maurträskets outlet at <0.6mg/L Cu, <0.6mg/L Pb, <5mg/L Zn and <5mg/L As (Eriksson et al. 2008).

8

9

3. MATERIALS

3.1. Waste rock

Waste rock used in this study originates from a waste rock heap at Boliden’s Maurliden mine, chosen for its high sulfide content (Eriksson et al. 2008). Different size fractions of the waste rock were used (Fig. 3).

Fig. 3 Waste rock heap at Maurliden mine site, selected for its high sulfur content (E. Andersson)

3.2. Alkaline rest products

The neutralizing potential of several alkaline materials are studied (Table 2). The alkaline wastes come from pulp and paper mills in northern Sweden. This is due to costs for long distance transportation. Mesa lime is the alkaline material that Boliden use today to neutralize their acid producing waste rock. The Argon Oxygen Decarburization (AOD) slag is a rest product from the steelmaking industry and consists of mainly of relatively soluble calcium silicates. This is also compared with the mineral calcite. The alkaline wastes were characterized according to their mineralogy, chemical composition and alkaline properties. All materials have a relatively high concentration of buffering calcium compounds.

Table 2 Rest products used in analysis and experiments.

Company Type of material Sample ID Produced (tons/year)

SCA Mesa lime SCA ML

- AOD Slag AOD SG 105 000

Iggesunds Bruk Green liquor dregs Iggesund GLD 6 500

Iggesunds Bruk Fly ash Iggesund FA 4 000 - 5 000

Iggesunds Bruk Fly ash “släkt” Iggesund FAS

Billerud Karlsborg Green liquor dregs Billerud GLD 11 200

Billerud Karlsborg Fly ash Billerud FA 2 300

Smurfit Kappa Kraftliner Green liquor dregs Smurfit GLD 9 000

10

4. METHODS

4.1. Sampling of waste rock

Waste rock from a dump at Maurliden mine site was manually sorted into four different size fractions: 0-5cm, 5-10cm, 10-20cm and 20-30cm. XRF equipment was used for selection of an area with as high content of sulfur as possible (Table 15). The waste rock was send to ALS Minerals in Piteå for crushing to <6mm.

4.2. Characterization of waste rock

4.2.1. Mineralogy

X-ray diffraction (XRD) was used for mineralogy studies at Luleå University of Technology. The instrument used is called PANanalytical Empyrean, PIXCEL3D Amedipix 2 collaboration. The software used is called High Score Plus. Three samples (from <0.125mm fraction) was analyzed between 5-90 2θ for approximately 5 minutes each.

4.2.2. Chemical composition

Two samples from each size fraction, 8 in total were analyzed for their chemical composition at ALS Environmental in Luleå. Following elements were analyzed: Al, As, Ba, Be Ca, Cd, Co, Cr, Cu, Fe, Hg, K, Mg, Mn, Mo, Na, Nb, Ni, P, Pb, S, Sc, Si, Sb, Sr, Ti, U, V, W, Y, Zn and Zr. Analyzes has been done with the EPA-methods (modified) 200.7 ICP-AES and 200.8 ICP-SFMS.

4.2.3. Neutralization potential

Acid Base Accounting (ABA) is a static test where the acid generation potential of a rock is determined. This is determined in three steps: determination of acid production (AP), determination of acid consumption (NP) and calculation of net acid production or consumption (NNP). The test was performed at Luleå University of Technology.

The neutralizing potential was determined using the Standard SS-EN 15875:2011,

Characterisation of waste - Static test for determination of acid potential and neutralization potential of sulfidic waste. 1mol/L HCl acid was added to 2 grams of waste rock mixed with

90 ml of milli-Q water to test the ability to neutralize acid. Duplicates of waste rock grinded down to <0.125mm was used. The mix was left on a shaker for 24 hours and pH was measured. Samples with a pH between 2 and 2.5 were titrated with 0.1mol/L NaOH to a pH of 8.3.

The acid potential (AP) is the maximum potential acid generation from a sample assuming that all sulfur occurs as pyrite and that acidity will result from complete oxidation of the pyrite. This is expressed as carbonate equivalents (CaCO3) in kg/t of solid were calculated

using the formula in Eq. 1.

AP = 31.25*WS (Eq. 8)

Where

11

The neutralization potential (NP) is the capacity of the waste rock to neutralize generated acidity and is expressed as carbonate equivalents (CaCO3) in kg/t was calculated using the

formula in Eq. 7.

(Eq. 9)

Where

c(HCl) is the concentration of the HCl (1 mole/liter)

VA is the volume added of HCl

c(NaOH) is the concentration of the NaOH (0.1 mole/liter)

VB is the volume added of NaOH

Md is the dry mass of test portion

The resulting neutralization potential ratio (NPR) (potential for neutralization of acidic drainages) was assessed using the formula in Eq. 8. In theory, a NRP > 1 should be enough to avoid acid drainage.

NPR = NP/AP (Eq. 10)

The net neutralization potential (NNP) was calculated from AP and NP using the formula in Eq. 6-7.

NNP = NP-AP (Eq. 11)

In theory, a rock with a negative NNP value has potential of acidification, meanwhile rocks with positive NNP does not have the potential. To be more certain a safety factor is applied. A negative NNP <-20 or <-30 kg CaCO3/t are considered to be potentially acid generating and

a positive NNP >+20 or >+30 kg CaCO3 are not considered to be potentially acid generating.

Rocks with NNP between -20 and +20 are uncertain acid generating potential (Mitchell 1999 and White et al. 1999). Mineralogical composition of the mine waste is of importance when interpreting NP values, because different minerals neutralize acid drainage at different rates and in different pH ranges. The test used does not distinguish between mineralogical differences and the NP is often overestimated (Lawrence and Scheske 1997).

4.3. Sampling and characterization of ARD leachate

4.3.1. Sampling of leachate

Sulfide-rich waste rock from Maurliden was filled into 1m3 tanks and leached in the lab at Luleå University of Technology. One tank was leached under water saturated conditions and the other tank leached under dry conditions. One filtered (22µm) and one unfiltered sample were collected, through taps located at the bottom of the tanks, once a week from December 2011 to April 2012, and twice a week from May to the end of August 2012. Water from melted snow was used in the leaching from December to the end of March, and then milli-Q water was used instead. Syringes and filter holders were washed in 5% nitric acid (HNO3).

0.22µm filters washed in 5% acetic acid were used to filter the samples. pH and electric conductivity was determined every week using the Metrohm 704 pH Meter and HANNA instruments HI 8733 Conductivity meter, respectively.

12

The leachate samples were analyzed at the accredited ALS Environmental in Luleå, using EPA-methods (modified) 200.7 ICP-AES and 200.8 ICP-SFMS for elements: Al, As, Ba, Ca, Cd, Co, Cr, Cu, Fe, HCO3, Hg, K, Mg, Mn, Mo, N-tot, Na, Ni, P, Pb, P-tot, S, Sb, Si, SO4,Sr

and Zn.

4.4. Laboratory Leaching Experiments

4.4.1. Experiment 1, Neutralization of ARD

Automated titrations were performed using an Autoburette (ABU901) and Radiometer Analytical 150ml. The software used was called Tim Talk 9. Method used in this study was based on Young et al. (1990). For each test, 0.5g of alkaline waste (e.g. rest products from Table 2) was mixed with 55gs deionized water, which gives an L/S-ratio of 110, which corresponds to water saturated conditions. The high L/S-ratio was also needed to make the automated equipment work properly. Duplicates of each material were performed. The sample was titrated to pH 7 and after that continued to pH 4.5. Leachate (ARD) from the dry tank with an approximate start pH of 2.25, was used as titrant and was added until the pH had stabilized according to the criterion that the pH should be stable for 300 sec = 5minutes. The duration of the titration for each waste material was approximately 4-6 hours (for both endpoints).

4.4.2. Experiment 2, Leaching of waste rock mixed with alkaline rest products Method for this experiment was modified based on Sartz (2010). Based on the results from the titration in experiment 1, theoretical mixing ratios of the waste rock and alkaline materials were calculated. After mixing a pH value between 7 and 9 is desirable in the leachate after 65 days of leaching. Low pH waters enhance the dissolution of many elements and increases the mobility and bioavailability of elements (Lottermoser 2003). According to the calculation, highest amount of mesa lime was needed (1.4%) to sustain a pH at 7 to 9. The other materials, Iggesund green liquor dregs (IGLD), Iggesund fly ash (IFA) and AOD slag required less material according to the calculation. Calculated values can be found in the Appendix.

A total of 55 samples were prepared for this experiment. The sample preparation method described in 4.4.1. was used for all samples. Based on the same calculation, unweathered waste rock with size fraction <6mm, was mixed with 0.5%, 1%, 1.5% and 2% alkaline material was mixed to 5g mixtures (noted with 1or 2). Five % of alkaline material was mixed to 10g mixtures (noted with 3 or 4). 50ml of milli-Q water were added to the 5g mixtures, and 100ml to the 10g mixtures. All samples were leached for 65 days, and the 5% mixtures for 115 days. The mixture was put on a shaker for 24hours, and after 1, 3, 8, 28 and 65 days, 20% of the leachate was removed and analyzed for its chemical composition, electric conductivity (EC) and pH. The pH and EC was measured with Metrohm 704 pH Meterand WTW inoLab cond level 1 Conductivity meter (WTW Tetracon® 325), respectively. Fresh milli-Q water (20%) was added after each sampling occasion. The samples were filtered through 0.22µm with the same acid washed equipment described in 4.3.1. Elements such as Al, As, Ba, Ca, Cd, Co, Cr, Cu, Fe, Hg, K, Mg, Mn, Mo, Na, Ni, P, Pb, S, Si, Sr and Zn were analyzed at ALS Environmental in Luleå. Instruments used by the accredited lab were AES and ICP-SFMS.

13

5. RESULTS

5.1. Characterization of Waste Rock

5.1.1. Mineralogy

As expected from ore characterizations done by Montelius (2005), the main components of waste rock are sulfides such as pyrite, arsenopyrite (FeAsS) and chalcopyrite (CuFeS2)

(Appendix 9.3 XRD). The gangue is dominated by quartz-feldspar porphyritic pumice breccias-sandstone (QFP pumice unit) (Montelius 2005).

5.1.2. Chemical composition

The mean chemical composition of 8 samples of waste rock (crushed down to <6mm), 2 from each size fraction is shown in (Table 3). Even though samples were representing different size fractions, there was no difference in the chemical composition between these samples. Concentrations of elements in Maurliden waste rock was compared with rock samples from the Maurliden domain, i.e. Maurträsket, Träskmyran and Rörmyrberget (Table 3) (Montelius 2005).

Concentrations of Be, Cd, Cr, Hg and S are not included in analyzes from earlier studies of samples from Maurliden domain. The sulfur concentration is high in the waste rock (307000ppm = 30.7%).This is considered high even though no comparison could be done with other rocks from Maurliden domain. Concentration of Hg is important (12.5ppm in the waste rock), because Hg in the leachate is not desirable for environmental concerns.

Higher concentrations of As, Fe, Pb and Sb were found in the waste rock compared to rock samples from Maurliden domain. Concentrations of As was high 190.5ppm in the waste rock, compared to 13.5ppm in rock samples from the Maurliden domain. Some elements are found in lower concentrations in the mine waste, e.g. Cu, Ni, V and Zr, with largest differences for Zr 152.3ppm compared to 67.4ppm in the rock and waste rock, respectively.

Concentrations of Ca, K, Mg, Mn, Mo, Na, Nb, P, Ti, U, W and Zn are more or less similar between the waste rock and the rock samples. A low concentration of Ca indicates a low calcite content of the rock.

14

Table 3 Results of the chemical composition of 8 samples of Maurliden waste rock, compared with 14 samples of coherent

and volcaniclastic sedimentary facies in Maurliden domain (Montelius 2005)

5.1.3. Neutralization potential

The results of the ABA test performed on 5 waste rock samples from Maurliden, shows that the waste rock is theoretically considered to be potentially acid producing (Table 4). A NNP value of <-20 or <-30 kg CaCO3/t indicates an acid producing potential of the mine waste.

Results of calculated acid potential (AP), neutralizing potential (NP), neutralization potential ratio (NPR) and net neutralization potential (NNP) are shown in (Table 4). Median value of 30.6% was used for the sulfur content WS (Eq. 8) when calculating AP.

Maurliden waste rock

No.8

Maurliden domain

No. 14

Element Unit Mean value Stdv Mean value Stdv

SiO2 % 28.86 4.58 71.08 7.2 Al2O3 % 5.93 1.45 12.06 2.1 CaO % 1.29 0.31 3.65 1.54 Fe2O3 % 17.2 4.39 3.94 3.08 K2O % 0.54 0.15 1.44 0.44 MgO % 0.97 0.09 1.33 1.31 MnO % 0.03 0.01 0.09 0.04 Na2O % 0.42 0.15 2.22 0.84 P2O5 % 0.02 0.01 0.11 0.08 TiO2 % 0.12 0.04 0.34 0.2 Sum % 55.38 4.85 100 LOI 1000°C % 21.83 1.99 3.88 1.31 As ppm 190.5 9.62 13.5 11.51 Ba ppm 91.41 29.27 273.5 109 Be ppm <0.5 Cd ppm 0.13 0.04 Co ppm 2.14 0.67 4.2 6.75 Cr ppm 57.49 20.43 Cu ppm 14.91 1.32 23.03 15.46 Hg ppm 12.35 2.22 Mo ppm <5 2.7 Nb ppm <5 4.54 1.41 Ni ppm 1.25 0.21 21.5 6.36 Pb ppm 20.31 2.08 6.68 1.04 S ppm 307000 33915 Sc ppm 5.89 1.63 13.1 6.87 Sr ppm 41.68 12.2 102.7 32.7 U ppm 1.74 0.31 1.95 0.39 V ppm 7.48 2.15 23.1 20.46 W ppm <50 0.83 0.33 Y ppm 11.24 1.87 39.4 6.01 Zn ppm 75.4 20.95 76.95 32.53 Zr ppm 67.38 13.44 152.3 17.16 Sb ppm 21.05 2.2 1.74 0.85

15

Table 4 Results from ABA test of five mine waste samples. The total concentration of sulfur is based on the results of the

chemical composition, presented in Table 3.

No. 5

Total sulfur % Eq. 8

AP kg /t CaCO3 Eq. 9 NP kg/tCaCO3 Eq. 10 NPR Eq. 11 NNP kg/tCaCO3 Minimum 25.1 784 12.1 0.013 -944 25th percentile 28.6 892 12.5 0.014 -944 Median 30.6 956 13.1 0.014 -943 75th percentile 32.3 1007 14.1 0.015 -942 Maximum 35.9 1122 15.2 0.016 -941 5.2. Characterization of leachate

5.2.1. Leachate from 1m3 tanks

The initial pH of 2.4in the water saturated tank was the lowest pH measured in that tank. The pH showed an increasing trend during the 8 months sampling period (Fig. 4). The electrical conductivity exhibits an opposite trend, starting at 9.2mS/cm and decreasing to 1mS/cm at the end of the 8 months sampling period.

During sampling on the 15th of February an accident happened which resulted in water was gushed out from the water filled tank. After that accident, the tap was sealed, water was refilled into the tank and sampling was resumed.

In the dry tank the pH has continually decreased from 6.7 down to constant pH around 1.4 (Fig. 4). Lowest pH was measured after 6 months at a pH of 0.9. Electric conductivity show an opposite trend and was continually increasing from 2.2ms/cm to almost 30mS/cm.

Fig. 4 pH and electrical conductivity measurements of both 1m3 _{tank, under 8 months sampling period.}

Chemical composition was analyzed every second week for both water saturated and dry tank (Fig. 5, 6 and 7). There is very small differences in concentrations between filtered and unfiltered samples, except for Hg were some differenced can be noted. In the saturated tank the concentration of most elements were kept at constant concentrations during the sampling period, or decreased even more compared the initial concentration. Most elements slightly decreased in concentration after the 15th of February when the accident happened. The opposite trends were observed in the dry tank, where concentrations of most elements

16

increased from their initial concentration. Elements such as Al (>2000mg/L), Fe (>40-50g/L) and S (~40g/L) increased to extremely high concentrations. Concentrations of trace elements such as As, Cr, Hg, Mo and Sb also increased in the dry tank. The concentrations of Ca, Mn, Na and Zn on the other hand showed a different trend, and were more constant in element concentrations. Concentration of bicarbonate (HCO3-) is reported under the detection limit

<1mg/L and <2.4mg/L for all samples, except the one analyzed in the water saturated tank at <3mg/L. Sulfate decreased in the water saturated tank from 12100mg/L to 1540mg/L, in the dry tank the concentration increased from 2010mg/L to 121000mg/L (=121g/L). More specified details of some element concentrations are shown in the figures 5, 6 and 7. All analyzed results can be found in Appendix 9.6.

Concentrations of calcium maintained a constant level between 400-600mg/L in both water saturated and dry tank throughout the 8 months of sampling. The concentration of Fe in the water saturated tank was 3400mg/L in the beginning of the sampling period and in the end decreased to <0.5mg/L. Compared to the dry tank this is an extremely low concentration. In the dry tank the concentration continually increased from 36mg/L to almost 50g/L, and decreased to 40g/L in the end of the sampling period. That is a 100,000 times higher concentration compared to the concentration in the water saturated tank. The amount of sulfur in the water saturated tank decreased from 3120mg/L to 578mg/L by the end of the sampling period. In the dry tank the concentration of sulfur constantly increased from 667mg/L to 40.000mg/L. The initial zinc concentration was 120mg/L in the water saturated tank and 90mg/L in the dry tank. After 2 months the concentration decreased to a constant concentration <10mg/L in the water saturated tank. In the dry tank it increased in the beginning up to 189mg/L, but was then kept at constant concentration at 100mg/L. This is approximately the same concentration as during the first week. The concentration of arsenic started at 1860µg/L but then decreased immediately to a concentration close to zero. In the dry tank though the concentration increased from <0.6mg/L to more than 80.000µg/L (=80mg/L). The concentration of As decreased in the end of the sampling period to 44.5mg/L. The concentration of chromium started at 74.4µg/L in the water saturated tank and then decreased down to <0.05µg/L. In the dry tank the initial concentration was <0.1µg/L and then it increased to <200µg/L. The concentration of copper decreased in the water saturated tank from 6520µg/L to 2µg/L. In the dry tank the concentration instead increased from 119µg/L to 4840µg/L in the end of the sampling period, with a maximum concentration of 14600µg/L after 12 weeks. Concentration of mercury was constantly kept at close to zero concentrations. 4 µg/L was though reported once after 6 months in the water saturated tank. In the dry tank the concentrations increased from 0.182µg/L to maximum 13µg/L after more than 5 months. But after that the concentrations decreased to 0.5µg/L again. The nickel concentration decreased in the water saturated tank from 386µg/L to 3µg/L. In the dry tank the concentration started at 235µg/L and was back on almost the same concentration at 380µg/L after a small increase to 611µg/L after 2 months leaching. The lead concentration decreased in the water saturated tank from 56.3µg/L 24.1µg/L after 4 months and then increased to 24.1µg/L in the end of the 8 month sampling period. In the dry tank the concentration increased from 63.2µg/L to 215µg/L, and in then decreased to 33µg/L.

17

Fig. 5 Concentration of selected main elements (mg/L) in lea chate from sulfidic containing waste rock from Maurliden.

Fig. 6 Concentration of selected trace elements (µg/L) in leachate from sulfidic containing waste rock from Maurliden

18

5.3. Laboratory leaching experiments

The chemical composition of the alkaline rest products used in this study is presented in Table 5. All products contain high concentrations of Ca. SCA mesa lime contains low concentrations of metals such as Cr, Cu, Ni and Zn. AOD slag also contains low concentration of metals, except for Cr, which is high in the slag (1.3%).

Table 5 Chemical composition of selected rest products used in analysis and experiments. SCA mesa lime and fly ash from

both Billerud and Iggesund are analyzed at ALS Environmental in Luleå. Green liquor dregs from Billerud and Iggesund are analyzed by Mäkitalo (2012), and the AOD slag by Ghan (200). Tables can also be found in the Appendix 9.1.

Element Unit SCA ML No. 1 AOD slag Billerud FA No. 1 Iggesund GLD No. 3 Iggesund FA No. 1 Billerud GLD No. 4 Al % 0.0476 2.5 7.52 11.98 ± 0.83 3.09 0.52 ± 0.33 Ca % 31.2 37.5 24.1 29.5 ± 5.31 20.8 26.9 ± 4.53 Fe % 0.203 1.6 2.39 0.85 ± 0.42 3.43 0.54 ± 0.19 K % <0.08 0.1 5.48 0.56 ± 0.33 4.49 0.29 ± 0.07 Mg % 0.277 3.7 3.6 2.20 ± 0.67 2.22 5.10 ± 1.56 Mn % 0.0198 0.7 1.41 0.71 ± 0.22 0.67 1.23 ± 0.39 Na % 0.804 0.09 2.24 1.23 P % 0.231 3.37 0.32 ± 0.07 1.8 0.24 ± 0.16 Ti % 0.29 0.04 ± 0.03 0.10 0.01 ± 0.003 As ppm 0.283 <3 0.77 ± 0.47 <3 0.35 ± 0.06 Ba ppm 171 101 2070 1250 Be ppm <0.5 <0.6 <0.5 Cd ppm 0.187 4.49 5.74 ± 2.08 8.03 5.73 ± 2.20 Co ppm 0.583 12.6 8.21 ± 2.56 7.04 4.45 ± 1.07 Cr ppm 14.4 13,100 195 78.20 ± 29.62 69 97.50 ± 22.66 Cu ppm 2.17 89.3 97.90 ± 37.30 67.7 143.50 ± 46.40 Hg ppm <0.04 0.691 <0.04 0.153 <0.04 Mo ppm 0.277 232 9.29 <5 Nb ppm <5 <6 <5 Ni ppm 6.94 95.1 40.67 ± 10.43 19.8 43.28 ± 10.76 Pb ppm 1.3 19.4 47.5 31.53 ± 13.40 27 3.65 ± 1.68 S ppm 371 0.12 14100 6460 ± 3090 8010 12570 ± 4628 Sc ppm <1 4.31 1.6 Sn ppm 0.201 2.07 0.627 ± 0.343 2.99 0.225 ± 0.096 Sr ppm 271 824 376 V ppm 2.01 205 83.2 22.7 W ppm <0.4 <60 <50 Y ppm 7.25 10.3 4.39 Zn ppm 37.8 251 1360 1450 ± 350 1720 1205 ± 453 Zr ppm <2 150 27.7

19 5.3.1. Titration with ARD

The titrand used in this experiment was leachate from the Maurliden waste rock with an initial pH of 2.25 (Fig. 8). The initial pH values were between 10.2 and 12.4 (Table 6), when using L/S-ratio 110, shown in (Table 19) in the Appendix 9.4.1. L/S 110. Both GLD and FA from Billerud showed great capacity to neutralize acid, 20ml ARD was added to reach pH 7 for Billerud FA and 60ml ARD to reach pH 4.5. The AOD slag and Iggesund FA was able to neutralize 10-15ml ARD, while calcite and Iggesund GLD had the lowest ability to neutralize acid.

Table 6 Initial pH data, dry matter ratios and from which year the alkaline materials are generated, for the materials studied.

Fig. 8 Results of neutralization experiment performed with different alkaline rest products: mesa lime, fly ash, green liquor

dregs, slag and calcite, at a L/S-ratio of 110. Note that the scale on the y-axis differs for different alkaline rest products.

SCA Mesa lime

AOD Slag Iggesund GLD Iggesund FA Iggesund FAS Billerud GLD Billerud FA Smurfit GLD Calcite Origin (year) 2012 2011 2011 2011 2011 - Dry matter ratio 0.70 1.00 0.53 1.00 0.56 0.52 1.00 0.49 - pH (L/S 110) 10-10.5 11.7-11.9 10.2-10.5 11.4-11.5 10.8-11.7 10.5-10.7 11.4-12.4 10.8-11.1 9.7-9.9 pH (L/S 27) 9.4-9.7 11-13.2 - 11.9-12.6 - 9.9-11.3 11.2-12.2 5.8-9.7 8.8

20 5.3.2. Leaching experiment

Results of pH and electrical conductivity (EC) meassurements are shown in Fig. 8. More detailed data of the measurements can be found in the Table 22 in Appendix 9.5.2.

All 14 samples with SCA mesa lime (ML) showed approximately similar pH at each sampling occasion, and decreased at the same rate, independent on the % of ML (Fig. 9). The initial pH ranged between 8.8 and 9.4, and ended at a pH near-neutral. After 65 days of leaching, one sample of the 1% mixtures and one sample of the 2% mixtures reached a pH <7. Those two samples together with one 5% sample were further leached until totally 115 days, when the pH showed 3.4 and 3.7 for 1% and 2% mixture, respectively, but the 5% mixture remained pH 7. The EC increased in all samples, but the values differed between the samples.

The AOD slag mixtures had an initial pH in the range of 11-12 and high pH throughout the 65 days of leaching (Fig. 9). After 115 days of leaching, the pH remained >9. Although, mixtures with 0.5 - 1% of slag mixtures decreased to pH 7-9. EC of the slag samples decreased during the sampling period. One sample with 2% mixture and one sample with 5% mixture were continued leaching for 115 days, and then reached a pH of 9.1 and 10.6, respectively.

All 10 Iggesund green liquor dregs (GLD) samples had similar pH during the 65 days of leaching, with an initial pH around 10 and a circum-neutral end pH (Fig. 9). The opposite trend was observed for the EC.One sample with 2% mixture and one sample with 5% mixture were leached for 115 days, and reached a pH of 7.3 and 7.9, respectively.

In the Iggesund fly ash (FA) samples, the pH decreased rapidly (Fig. 9). The 5% mixtures had initial pH 12 and the remaining samples had a pH in the range of 11.4-11.5. Samples with 1%, 1.5% and 2% decreased to pH<7 after 65 days of leaching. The EC varied between the samples. It decreased for 5% mixtures and increased for the rest of the samples. One sample with 2% mixture and one sample with 5% mixture were continued leaching for 115 days, and reached a pH of 7.4 and 8.6, respectively.

Only 2 samples of 5% Billerud GLD was prepared (Fig. 9). The pH decreased from 9.4 to 8.4, and EC showed the opposite trend and slightly increased, after 65 days of leaching.

The pH of Billerud FA mixtures decreased the most rapidly of all 55 samples compared (Fig. 9). Initial pH varied from 9.5 to 12, and decreased to pH 7 to 10. The 0.5% mixtures had a pH <7 after 65 days of leaching, and EC showed similar trend and decreased with time in all samples. One sample with 2% mixture and one sample with 5% mixture were continued leaching for 115 days, and reached a pH of 7.1 and 9.5, respectively.

21

Fig. 9 Results from pH and electric conductivity measurements after 65 days of leaching. Some samples are leached for 115 days.

Totally 55 samples.

A total of 55 samples were prepared for the experiment, but only 2% and 5% mixtures were stored from each sampling (duplicates). Of those stored samples, only 1 sample from 2% and 1 sample from 5% mixtures of SCA mesa lime, AOD slag Iggesund GLD and Iggesund FA were analyzed (totally 8 samples from each sampling). Results of some selected elements are shown in (Fig. 9). Concentrations of all analyzed elements can be found in Appendix 9.5.3. Ca. The concentration of calcium increases in 5 of 8 samples (Fig. 10). In the 2% ML mixture the concentration increases from 47mg/L to 244mg/L, and in the 5% ML mixture from 33mg/L to 169mg/L. The Ca concentration in the 2% slag mixture decreased from 181mg/L to 170mg/L, and the 5% slag mixture from 197mg/L to 124mg/L. The Ca concentration in the 2% Iggesund GLD mixture increased from 20mg/L to 211mg/L, and the 5% mixture from 5.6mg/L to 62mg/L. The Ca concentration in the 2% Iggesund FA mixture increased from 83mg/L to 627mg/L, while the 5% mixture decreased from 131mg/L to 112mg/L. Highest concentration of Ca was found in the 2% Iggesund FA mixture, and lowest concentration in the 5% Iggesund GLD mixture.

22

Fe. The concentration of iron increased in all 8 samples (Fig. 10). The Fe concentration in the 2% ML mixture increased from 0.002µg/L to 0.02µg/L, while the 5% ML mixture increased from 0.003µg/L to 0.02µg/L and in the end decreased again to 0.004µg/L. The Fe concentration in the 2% slag mixture slightly increased from 0.01µg/L to 0.04µg/L, and the 5% slag mixture increased from 0.01µg/L to 0.02µg/L. The Fe concentration in the 2% Iggesund GLD mixture was constantly 0.012µg/L, while the 5% mixture slightly increased from 0.005µg/L to 0.01µg/L. The Fe concentration in the 2% Iggesund FA mixture increased from 0.01µg/L to 206µg/L at day 65´s analyze, while the 5% mixture decreased from 0.05µg/L to 0.01µg/L. Fe concentration was higher in the 2% mixture compared to the 5% mixture. Highest concentration of Fe was found in the 2% Iggesund FA mixture after 65days of leaching.

S. The concentration of sulfur increased in all 8 samples (Fig. 10). The S concentration in the 2% ML mixture increased from 39µg/L to 170µg/L, and the 5% mixture from 31.5µg/L to 115µg/L. The S concentration in the 2% slag mixture increased from 25µg/L to 295µg/L and then decreased again to 191µg/L, while the 5% slag mixture increased from 16µg/L to 30µg/L. The S concentration in the 2% Iggesund GLD mixture increased from 57µg/L to 179µg/L, and the 5% mixture from 56µg/L to 117µg/L. The S concentration in the 2% Iggesund FA mixture increased from 94 µg/L to 843 µg/L, and the 5% mixture from 64µg/L to 136µg/L. Sulfur concentrations were higher in the 2% mixtures compared to the 5% mixtures, and the highest concentration was found in Iggesund FA.

Fig. 10 Concentration of selected main elements in leachate, after mixing experiment of alkaline rest products and waste rock

after 65 days of leaching. 5% samples are leached for 115 days.

Al. The concentration of aluminum decreased in 7 of 8 samples (Fig. 11). The Al concentration in the 2% ML mixture decreased from 24µg/L to 9µg/L, and the 5% ML mixture from 67µg/L to 13µg/L. The Al concentration in the 2% slag mixture decreased from

23

1210µg/L 16µg/L, while the 5% mixture decreased from 4060µg/L to774 µg/L. The Al concentration in the 2% Iggesund GLD mixture decreased from 67µg/L to 17µg/L, while the 5% mixture decreased from 227µg/L to 38µg/L. The Al concentration in the 2% Iggesund FA mixture decreased from 3120 µg/L to 18.5 µg/L, but increased again to 18,000µg/L after 65 days of leaching. Al concentrations was higher in the 5% mixtures compared to the 2% mixtures, except for Iggesund FA mixtures where the concentrations in the 2% mixture after 65 days of leaching. Highest concentration of Al was found in the 2% Iggesund FA mixture, while the lowest concentration was found in the ML mixtures.

As. The arsenic concentration decreased in the 2% ML mixture from 0.6µg/L to <0.5µg/L, and in the 5% ML mixture from 1.6µg/L to 0.9µg/L (Fig. 11). The As concentration in the 2% slag mixture decreased from 4.8µg/L to 1.2µg/L, and the 5% mixture from 4.2µg/L to 2.4µg/L. The As concentration in the 2% Iggesund GLD mixture slightly increased from 0.5µg/L to 0.7µg/L, and the 5% mixture slightly decreased from 1.7µg/L to 1.4µg/L. The As concentration in the 2% Iggesund FA mixture decreased from 3.8µg/L to <1µg/L, and the 5% mixture increased from 1.5µg/L to 3.5µg/L. Highest concentration of As is found in the 5% Iggesund FA mixture, and lowest concentration in the 2% ML mixture.

Cr. The chromium concentration was the same in both the 2% and 5% ML mixtures, varying between 0.05µg/L and 0.09µg/L (Fig. 11). The Cr concentration in the 2% slag mixture decreased from 41µg/L to 4.µg/L, while the 5% slag mixture increased from 47.5µg/L to 116µg/L. The Cr concentration in the 2% Iggesund GLD mixture, decreased from 0.11µg/L and the rest of the analyzes were reported under detection limit <0.1µg/L. The Cr concentration in the 5% Iggesund GLD mixture decreased from 0.26µg/L to 0.13µg/L. The Cr concentration in the 2% Iggesund FA mixture increased from 1.02µg/L to 0.47µg/L, and the 5% mixture from 5µg/L to 1.5µg/L. Highest Cr concentration was found in the 2% Iggesund FA mixture, and lowest concentration in the 5% Iggesund GLD mixture.

Cu. Many samples analyzed for concentration of Cu were reported under the detection limit <0.5µg/L or <1µg/L, which result in a straight line (Fig. 11). That was the case for ML and Iggesund GLD, both the 2% and 5% mixtures. The Cu concentration in the 2% slag mixture had an initial concentration of 0.7µg/L, increased to 1.7µg/L and ended at <1µg/L after 65 days of leaching. The Cu concentration in the 2% Iggesund FA mixture ended at a concentration of <2µg/L, while the 5% mixture had a concentration of 0.12 after 65 days of leaching. Highest Cu concentration was found in the 2% Iggesund FA mixture and lowest concentration in the 5% Iggesund FA mixture.

Hg. The mercury concentration decreased in the 2% ML mixture from 0.018µg/L to 0.007µg/L, and in the 5% mixture from 0.014µg/L to <0.002 µg/L (Fig. 11). The Hg concentration in the 2% slag mixture decreased from 0.036µg/L to 0.013µg/L, and the 5% mixture from 0.088µg/L to 0.014µg/L. Analyzes for the Cr concentration in the 2% Iggesund GLD mixture was reported under detection limits <0.005 µg/L to <0.01µg/L, and the 5% mixture under detection limits <0.004µg/L to <0.002µg/L. Highest concentration of Hg was found in the beginning of leaching the 5% slag mixture. After 65 days leaching, the highest concentration was still found in the same slag.

Pb. The lead concentration was reported under the detection limit <0.05µg/L for both the 2% and 5% ML mixtures (Fig. 11). Day 65´s concentration was <1µg/L for the 2% mixture and 0.017µg/L for the 5% mixture. The Pb concentration in the 2% slag mixture decreased from 0.66µg/L to <0.1µg/L and in the 5% slag mixture the concentration decreased from 0.2µg/L

24

to 0.1µg/L. Lead concentrations in the 2% and 5% Iggesund GLD mixtures, were reported under the detection limit <0.05µg/L and <0.1 µg/L, until after 65 days when the 2% mixture had a concentration of <1µg/L and the 5% mixture 0.025µg/L. The Pb concentration in the 2% Iggesund FA mixture increased from <0.05µg/L to 12.8µg/L, while the 5% mixture decreased from 0.3 µg/L to 0.01µg/L. Highest lead concentration was found in the 2% Iggesund FA mixture and lowest concentration was found in the ML mixtures.

Zn. The zinc concentration in the 2% ML mixture varied between 2.4µg/L, 12.7µg/L and 2.5µg/L, while the 5% ML mixture varied from 1.6µg/L to 2.5µg/L (Fig. 11). The Zn concentration in the 2% slag mixture increased from 3.8µg/L to 11.7µg/L, and the 5% slag mixture decreased from 2.7to 1.5µg/L. The Zn concentration in the 2% Iggesund GLD mixture increased from 3.4µg/L to 5.7µg/L after a highest concentration at 9µg/L. In the 5% Iggesund GLD mixture the concentration varied more, from 1.6µg/L to 13.2µg/L, down to 3.7µg/L and then increased again to 13µg/L. The Zn concentration in the 2% Iggesund FA mixture increased from 8.85µg/L to 3260µg/L, while the 5% mixture decreased from 71.6µg/L to 2.4µg/L. Highest Zn concentration was found in Iggesund FA mixtures and lowest Zn concentration in the slag mixtures.

Concentrations of element such as Cd, Co, Mo, Ni and Sr after 65 days of leaching can be found in Appendix 8.5.3. ALS Analyzes.

25

Fig. 11 Concentration of selected trace elements in leachate, after mixing experiment of alkaline rest products and waste rock