bulletin 406 research report · rasilainen kalevi, eilu pasi, ahtola timo, halkoaho tapio,...

Geological Survey of Finland

Geological Survey of Finland

Bulletin 406 • Research Report

Quantitative assessment of undiscovered resources in lithium–caesium–tantalum pegmatite-hosted deposits in Finland

Rasilainen Kalevi, Eilu Pasi, Ahtola Timo, Halkoaho Tapio, Kärkkäinen Niilo, Kuusela Janne, Lintinen Panu, Törmänen Tuomo

2018

Bulletin 406 • Research Report • Kalevi Rasilainen et al.

www.gtk.fi

Geological Survey of Finland, Bulletin

The Bulletin of the Geological Survey of Finland publishes the results of scientific research that is thematically or geographically connected to Finnish or Fennoscandian geology, or otherwise related to research and innovation at GTK. Articles by researchers outside GTK are also welcome. All manuscripts are peer reviewed.

Editorial BoardProf. Pekka Nurmi, GTK, ChairDr Stefan Bergman, SGUDr Asko Käpyaho, GTKDr Antti Ojala, GTKDr Timo Tarvainen, GTK, Scientific Editor

Instructions for authors available from the Scientific Editor.

GEOLOGICAL SURVEY OF FINLAND

Bulletin 406

Quantitative assessment of undiscovered resources in lithium–caesium–tantalum pegmatite-hosted deposits in Finland

by

Rasilainen Kalevi, Eilu Pasi, Ahtola Timo, Halkoaho Tapio, Kärkkäinen Niilo, Kuusela Janne, Lintinen Panu and Törmänen Tuomo

Unless otherwise indicated, the figures have been prepared by the authors of the publication.

https://doi.org/10.30440/bt406

Received 1 December 2017; Accepted 23 April 2018

Layout: Elvi Turtiainen Oy

Espoo 2018

https://www.tsv.fi/tunnushttps://doi.org/10.30440/bt406

Rasilainen, K., Eilu, P., Ahtola, T., Halkoaho, T., Kärkkäinen, N., Kuusela, J., Lin-tinen, P. & Törmänen, T. 2018. Quantitative assessment of undiscovered resources in lithium–caesium–tantalum pegmatite-hosted deposits in Finland. Geological Survey of Finland, Bulletin 406, 31 pages, 9 figures, 6 tables and 3 appendices.

Lithium resources in undiscovered lithium–caesium–tantalum (LCT) pegmatite deposits were estimated down to the depth of one kilometre in the bedrock of Finland using a three-part quantitative assessment method. There are six well-known LCT pegmatite lithium deposits in Finland, and they are all located within or near the border of the municipality of Kaustinen in central Ostrobothnia. The identified resources in these deposits are 45,520 t of lithium. Well-known global LCT pegmatite deposits from outside Finland have an order of magnitude larger median ore tonnage than the Finnish deposits. The global and Finnish deposits do not differ in their median lithium grade. Due to the small number of well-known Finnish deposits and the probably incomplete delineation of most of the known LCT pegmatite deposits both in Finland and abroad, a grade-tonnage model was constructed using data for 29 deposits in seven countries. Nineteen permissive tracts were delineated for LCT pegmatite lithium deposits in different parts of Finland, covering altogether 22,404 km2. The expected number of undiscovered deposits within the delineated permissive tracts is 6.8 deposits, and the undiscov-ered deposits are estimated to contain, with a 50% probability, at least 510,000 t of lithium. Over 90% of the estimated undiscovered lithium resources in Finland are located within either the Kaustinen permissive tract or the surrounding larger Järvi-Pohjanmaa tract. The assessment results indicate that at least 90% of the remaining lithium endowment within the uppermost one kilometre of the Finnish bedrock is in poorly explored or entirely unknown deposits. The identified lithium resources in Finland constitute 0.1% of the known world resources, 1.5% of the European and 2.6% of the EU lithium resources. There have been no global assess-ments of undiscovered lithium resources to compare with the results from Finland.

Keywords: Lithium, LCT pegmatite, undiscovered resources, evaluation, quantita-tive analysis, Proterozoic, Archaean, Finland

Kalevi RasilainenGeological Survey of FinlandP.O. Box 96FI-02151 ESPOOFINLAND

E-mail: [email protected]

ISBN 978-952-217-396-6 (PDF) ISSN 0367-522X (print) ISSN 2489-639X (online)

2

CONTENTS

1 INTRODUCTION ............................................................................................................................................ 51.1 Assessment project of the Geological Survey of Finland .................................................................51.2 Terminology ...........................................................................................................................................7

2 LITHIUM DEPOSITS IN FINLAND ................................................................................................................72.1 LCT pegmatites ......................................................................................................................................82.2 The LCT pegmatite-hosted Li deposits in Finland ............................................................................9

2.2.1 The Rapasaari lithium deposit ................................................................................................ 12

3 THE THREE-PART QUANTITATIVE RESOURCE ASSESSMENT METHOD .............................................. 123.1 Deposit models .................................................................................................................................... 14

3.1.1 Descriptive models ...................................................................................................................143.1.2 Grade-tonnage models ............................................................................................................14

3.2 Permissive tracts ................................................................................................................................. 143.3 Estimation of the number of undiscovered deposits ....................................................................... 153.4 Statistical evaluation ........................................................................................................................... 15

4 ASSESSMENT OF LITHIUM RESOURCES IN FINLAND .............................................................................164.1 Resources covered by the assessment ............................................................................................... 164.2 GTK assessment process .................................................................................................................... 164.3 Data used .............................................................................................................................................. 16

4.3.1 Geology ......................................................................................................................................164.3.2 Known mineral deposits and occurrences .............................................................................164.3.3 Geophysical and geochemical data ......................................................................................... 174.3.4 Exploration history ................................................................................................................... 17

4.4 Deposit model ...................................................................................................................................... 174.5 Tract delineation ................................................................................................................................. 174.6 Estimation of the number of undiscovered deposits ....................................................................... 184.7 Assessment of metal tonnages .......................................................................................................... 19

5 RESULTS AND DISCUSSION ........................................................................................................................195.1 Permissive tracts delineated ..............................................................................................................205.2 Undiscovered resources in LCT pegmatite lithium deposits .......................................................... 215.3 Finnish lithium endowment in national and global contexts ........................................................255.4 Reliability and usability of the estimates .........................................................................................26

6 SUMMARY .................................................................................................................................................... 26

ACKNOWLEDGMENTS ...................................................................................................................................... 27

REFERENCES ...................................................................................................................................................... 27

APPENDICESAppendix 1: Descriptive model for LCT pegmatite Lithium deposits .....................................................32Appendix 2: Grade-tonnage models for LCT pegmatite lithium ............................................................ 33Appendix 3: Assessment results for the LCT pegmatite LIthium permissive tracts .............................42

3

Geological Survey of Finland, Bulletin 406Rasilainen Kalevi, Eilu Pasi, Ahtola Timo, Halkoaho Tapio, Kärkkäinen Niilo, Kuusela Janne, Lintinen Panu and Törmänen Tuomo

44

1 INTRODUCTION

Throughout the history of humankind, the demand for mineral resources has increased with the contin-uing growth of the world population and the rise in the average material standard of living of an ever-increasing number of individuals. In tandem with this trend, exploration for and the development of new mineral resources all over the world are fac-ing increasing competition from other land uses (e.g., Briskey et al. 2007, Cunningham et al. 2007, Idman et al. 2007, Rasmussen 2011, World Economic Forum 2014, 2015). Recycling cannot cover all the demand for practically any metal, even where very little of the commodity, such as gold, is lost dur-ing production, manufacturing and recycling (e.g., Buchert et al. 2009, Wellmer & Dalheimer 2012). Concerns about the environmental effects of min-ing are also having a growing influence on the cost and pace of development of new natural resources. In the modern world, Finland can no longer rely on the ready availability of imported raw materi-als for manufacturing and other industries. This applies to the entire European Union, which is glob-ally a major net importer of nearly all metallic ores and concentrates (Kauppa- ja teollisuusministeriö 2006, Commission of the European Communities 2008, European Commission 2011, 2014). We need to know our mineral resources and how they might be expanded and managed. The essential information includes the location of the known resources, the location and amount of the possibly existing, yet

undiscovered resources, and the uncertainty related to their existence. Furthermore, it is important to know how the development of mineral deposits will affect local people and other surrounding resources, such as biological diversity, forests, arable land, air and water.

This report aims to answer the questions of ‘where’ and ‘how much’ we could still expect to find, given all that we have already discovered and used. We describe the process and results of a quantitative assessment of lithium (Li) resources in undiscovered lithium–caesium–tantalum (LCT) pegmatite-hosted deposits in Finland. The report contains two parts. The first part reviews the LCT-type pegmatite-hosted deposits in Finland and their geological environments, the assessment method, the data used and the assessment process itself. A summary of the assessment results is given and the results are discussed. The second part comprises the Appendices, which include the deposit model employed and detailed information on each permis-sive tract delineated.

The information provided here on the location and amount of undiscovered mineral resources is expected to be valuable for effective land-use plan-ning and the sustainable development of mineral resources, and also in evaluating the long-term productivity of investments in exploration and related research and education.

1.1 Assessment project of the Geological Survey of Finland

The demands defined above, and requests from various stakeholders (including the National Audit Office of Finland) to produce exact information on potential resources, resulted in the initiation of an assessment project at the Geological Survey of Finland (GTK) in 2008. The project was established to produce unbiased information on undiscovered minerals resources for national and regional plan-ning of land use, natural resources management

and environmental actions, as well as to develop assessment tools and enhance their proper appli-cation in the conditions of Finnish bedrock. The results of the project aim to enable accounting of metallic natural resources according to the princi-ples of sustainable development. The project has also provided new information for metallogenic and lithological research and for national-level planning of mineral exploration.

55

Geological Survey of Finland, Bulletin 406Quantitative assessment of undiscovered resources in lithium-caesium-tantalum pegmatite-hosted deposits in Finland


The project started in 2008 with the selection of the working methods. By the end of 2017, undis-covered resources in the following deposit classes in Finland had been assessed: Platinum-group element deposits in mafic–ultramafic layered intrusions (Rasilainen et al. 2010a, b); nickel±copper deposits related to synorogenic mafic–ultramafic intrusions and komatiitic rocks (Rasilainen et al. 2012); vol-canogenic massive sulphide, porphyry copper and Outokumpu-type deposits (Rasilainen et al. 2014); orogenic gold deposits (Eilu et al. 2015); stratiform and podiform chromite deposits (Rasilainen et al. 2016) and orthomagmatic mafic intrusion-related iron–titanium–vanadium deposits (Rasilainen et al., in preparation). The results of the assessment of lithium resources in LCT-type pegmatite-hosted deposits in Finland are provided in this report.

The procedure selected for the GTK assessments is based on a three-part quantitative assessment method developed by the U.S. Geological Survey (USGS) starting from the mid-1970s (Singer 1993, Singer & Menzie 2010). It must be emphasized that the method does not provide mineral resource or reserve estimates consistent with the present industrial standards such as the JORC, CRIRSCO,

NI 43-101, PERC and UNFC codes (Joint Ore Reserves Committee of the Australasian Institute of Mining and Metallurgy, Australian Institute of Geosciences and Mineral Council of Australia 2012, Committee for Mineral Reserves International Reporting Standards 2013, National Instrument 43-101 2011, Pan-European Reserves and Resources Reporting Committee 2013, United Nations Economic Commission for Europe 2009). The results of undiscovered resource assessments should never be confused with proper reserve or resource esti-mates based on international standards. Rather, the assessment process produces probabilistic estimates of the total amount of metals in situ in undiscovered deposits of selected types, down to the depth of one kilometre. The modification of the process used in the GTK assessments does not take into account the economic, technical, social or environmental factors that might in the future affect the potential for eco-nomic utilisation of a resource. Hence, part of the estimated undiscovered resources may be located in sub-economic occurrences (Fig. 1), and it might be more appropriate to use the term ‘metal endow-ment’, which is not directly dependent on economic or technological factors (Harris 1984).

UNDISCOVERED RESOURCES

Demonstrated

Probability range

Measured Indicated InferredHypothetical Speculative

Economic

Marginallyeconomic

Subeconomic

Three-partassessment

CRIRSCO, JORC, NI 43-101,PERC, UNFC

Eco

no

mic

fe

asib

ility

Geologic certainty

Identified resources

DISCOVERED RESOURCES

Pastproduction

Fig. 1. Classification of mineral resources used in GTK assessments (modified from U.S. Geological Survey National Mineral Resource Assessment Team 2000). Economic feasibility increases upwards and geological uncertainty increases to the right.

66

1.2 Terminology

Some terms essential to the proper understand-ing of this report are briefly described below. The definitions follow the usage by the minerals indus-try and the resource assessment community (U.S. Bureau of Mines and U.S. Geological Survey 1980, U.S. Geological Survey National Mineral Resource Assessment Team 2000, Committee for Mineral Reserves International Reporting Standards 2013).

Mineral depositA mineral occurrence of sufficient size and grade that it might, under the most favourable circum-stances, be considered to have economic potential.

Well-known mineral depositA completely delineated mineral deposit, for which the identified resources and past production are known.

Undiscovered mineral depositA mineral deposit believed to exist less than 1 km below the surface of the ground, or an incompletely explored mineral occurrence within that depth range that could have sufficient size and grade to be classified as a deposit.

Mineral occurrenceA concentration of any useful mineral found in bedrock in sufficient quantity to suggest further exploration.

Mineral resourceA concentration or occurrence of material of eco-nomic interest in or on the Earth’s crust in such a form, quality and quantity that there are reason-able prospects for eventual economic extraction.

The location, quantity, grade, continuity and other geological characteristics of a mineral resource are known, estimated or interpreted from specific geo-logical evidence, sampling and knowledge.

Identified resourcesResources whose location, grade, quality and quan-tity are known or can be estimated from specific geological evidence.

Well-known resourcesIdentified resources that occur in completely delin-eated deposits included in grade-tonnage models.

Discovered resourcesThe total amount of identified resources and cumu-lative past production.

Undiscovered resourcesResources in undiscovered mineral deposits whose existence is postulated based on indirect geological evidence.

Hypothetical resourcesUndiscovered resources in known types of mineral deposits postulated to exist in favourable geological settings where other well-explored deposits of the same types are known.

Speculative resourcesUndiscovered resources that may occur either in known types of deposits in favourable geological settings where mineral discoveries have not been made, or in types of deposits as yet unrecognized for their economic potential.

2 LITHIUM DEPOSITS IN FINLAND

Presently, the world’s dominant sources of lith-ium are continental brines and pegmatites, but hectorite clay deposits and jadarite deposits may become important in the future (British Geological Survey 2016, Jaskula 2017). All the known lithium

deposits and occurrences in Finland are hosted by pegmatites, and at least all the well-known lithium deposits are hosted by LCT pegmatites (Alviola 2012, Ahtola et al. 2015). No indications of other types of lithium deposits are known in Finland.

77



2.1 LCT pegmatites

The following summary is based on the U.S. Geological Survey mineral deposit model for LCT pegmatites (Bradley et al. 2017, and references therein). LCT pegmatites are granitic rocks that form small igneous bodies characterised by large crystals and distinctive textures. They are highly enriched in lithium, caesium and tantalum, and this diagnostic suite of elements gives LCT pegmatites their name and distinguishes them from other rare-element pegmatites. LCT pegmatites occur in Cainozoic to Mesoarchaean orogenic belts on all continents. Most of them are differentiated end members of peraluminous, S-type granitic melts, whereas some are related to metaluminous granites or I-type granites.

LCT pegmatites are most commonly hosted by metasedimentary or metavolcanic rocks, and less commonly by plutonic rocks (granites, gabbros). The country rocks are typically metamorphosed

in upper greenschist to amphibolite facies condi-tions. LCT pegmatites can display a mineralogical and geochemical zoning pattern, with pegmatites most strongly enriched in incompatible elements being located furthest away from the known (or inferred) granitic pluton (Fig. 2).

LCT pegmatite bodies occur in various forms, including tabular dykes or sills, lenticular bod-ies and irregular masses. Some pegmatites can be spatially and genetically linked to exposed granitic bodies, but in other cases, no parent plutons are exposed. Most LCT pegmatite bodies show a con-centric zonation with an outer border zone, a wall zone, one or more intermediate zones and a core (Fig. 3). Lithium, caesium and tantalum are gener-ally concentrated in the intermediate zones. Narrow LCT pegmatite bodies can show layering instead of concentric zoning. In more uncommon cases, LCT pegmatites can be unzoned.

Parental granite

Lim

it of

peg

mat

ite h

alo

arou

nd g

ranite

Ber

ylliu

m

Lith

ium

, ber

ylliu

m, n

iobi

um >

tant

alum

Lith

ium

, ber

ylliu

m, t

anta

lum

,

tin, (

caes

ium

, rub

idiu

m)

Metasedimentary or meta-igneous country rocks

Granite

Pegmatitic granite

Pegmatite

Boundary between zones ofrare-element enrichment

Fault

Fig. 2. Idealized representation of a regional zoning pattern in a pegmatite field (modified from Bradley et al. 2017). Characteristic rare-element suites of the most enriched pegmatites in each zone are indicated.

88

Approximately 25% of the lithium production of the world comes from LCT pegmatites (Bradley et al. 2017), as well as about 10% of the beryllium production (Lederer et al. 2016), most of the tan-talum production and all of the caesium production (Schulz & Papp 2014, Tuck 2017). LCT pegmatites are also mined for tin, high purity quartz, potassium feldspar, albite, kaolinite, white mica, gem beryl, gem tourmaline, and museum-quality specimens of many rare minerals.

Several mineral deposit types can occur associ-ated with LCT pegmatites. The least fractionated and mineralogically simplest pegmatites within an LCT pegmatite field can be considered as common granitic pegmatites. They are not enriched in rare elements but can be important sources of ceramic-grade feldspar, ultra-pure quartz and muscovite.

Niobium–yttrium–fluorine (NYF) pegmatites rep-resent the other endmember of rare-element peg-matites. These pegmatites are enriched in several rare elements, have Nb > Ta, and are impoverished in Li, Rb and Cs (Ercit 2005). Some pegmatites have mixed characteristics of LCT and NYF peg-matites, interpreted to represent contamination of an NYF melt with local sources (Martin & De Vito 2005). Certain Phanerozoic granitic plutons show enrichment of rare elements (Li, Ta, Sn, F), typi-cally within zones of a larger granite body (Schwartz 1992, Raimbault 1998). Hydrothermal quartz veins containing Sn-, Ta-, Nb- and W-bearing miner-als occur within certain pegmatite districts. In the Pilbara Craton, Western Australia, they are inter-preted to be offshoots of mineralised pegmatites (Sweetapple 2000).

2.2 The LCT pegmatite-hosted Li deposits in Finland

There are six well-known lithium deposits and over 30 partially explored occurrences in Finland (Fig. 4), all located in the southern part of the coun-try. The six well-known deposits are all hosted by LCT pegmatites and occur in a restricted area at Kaustinen (Fig. 5).

The Kaustinen lithium pegmatite prov-ince is located in western Finland within the Palaeoproterozoic supracrustal rocks of the Pohjanmaa belt, which is bordered by Vaasa grani-toid complex in the west and the Central Finland granitoid complex in the east (Vaasjoki et al. 2005). The most common rock types within the Pohjanmaa belt are mica schists and mica gneisses, which are

intercalated with metavolcanic rocks. The suprac-rustal rocks have been divided into the Evijärvi and Ylivieska fields (Kähkönen 2005), and the Kaustinen lithium pegmatite area is located in the northern continuation of the Evijärvi field.

The lithium potential of the Palaeoproterozoic Pohjanmaa belt in western Finland has been known since the late 1950s. The lithium mineral spodumene was first identified from boulders in Kaustinen in 1959 (Boström 1988), and since then, lithium-bear-ing pegmatites have been explored in the area by several enterprises as well as the Geological Survey of Finland. The Kaustinen area is known to contain 10 drilled spodumene pegmatite targets (Fig. 5). Six

Border zone

Wall zone

Albite zone

Intermediate zone(s)

Core zone

Core margin: Large crystals oftourmaline, beryl or spodumene

Fig. 3. Idealized deposit-scale zoning of a pegmatite body (modified from Bradley et al. 2017). The thickness of the border zone is exaggerated.

99



!(#*#*

#*

#*

#*

!(!(!(

#*

!(!(

#*

#*

#*#*

#*#*#*

#*#*#*

#*#*#*#*#*

#*#*#*

#*#*#*#*#*#*#*

#*#*#*

"

"

"

"

"

"

"

"

"

Oulu

Turku

Vaasa

Kuopio

Tampere

Joensuu

Helsinki

Jyväskylä

Rovaniemi

3 000 000 3 200 000 3 400 000 3 600 000 3 800 000

6 8

00

00

07

00

0 0

00

7 2

00

00

07

40

0 0

00

7 6

00

00

0

0 100 20050 km

Kaustinen

Seinäjoki

Somero-Tammela

Eräjärvi

Kisko-Orijärvi-Pohja

Kemiö

Dolerites

Rapakivi granites

Caledonian orogenic belt

Palaeo - Mesoproterozoic rocks

Sedimentary rocks

Palaeoproterozoic rocks

Intrusive rocks

Supracrustal rocks

Granulitic rocks

Archaean rocks

Intrusive rocks and gneisses

Supracrustal rocks

!( Li deposit

#* Li occurrence

LCT pegmatite Li deposits

Fig. 4. Lithium deposits and occurrences in Finland. The names of the most important districts are shown in italics. The geological map is based on the GTK in-house digital bedrock database (Bedrock of Finland − DigiKP).

1010

of these targets have been developed into lithium deposits (Table 1), for which a pre-feasibility study including a mineral resource estimate has recently been published (Sweco Industry 2016).

In the classification of granitic pegmatites (Černý & Ercit 2005), the lithium pegmatites of the Kaustinen province belong to the rare-element class of the LCT family, and more specifically to the albite-spodumene type of the lithium subclass (Ahtola et al. 2015). Rocks of the Kaustinen area were metamorphosed under low-pressure lower amphibolite conditions, and the lithium pegma-

tites intruded after the metamorphic peak at ca 1.79 Ga (Alviola et al. 2001). The pegmatites are mostly covered by Quaternary overburden, and the contact relations with their metavolcanic and meta-sedimentary wall rocks can often only be observed in erratic boulders and diamond drill cores. The pegmatite granites in the Kaustinen area are the possible source granites for the lithium pegmatites (Martikainen 2012), but this has not been confirmed by age determinations. The Rapasaari deposit is briefly described below as an example of the Finnish lithium pegmatite deposits.

Table 1. Well-known lithium deposits and resources in Finland.

Deposit Resource (t) Li2O% Li (t) Reference

Rapasaari 3,456,000 1.15 18,464 Keliber Oy (2017)

Syväjärvi 1,970,000 1.24 11,349 Keliber Oy (2017)

Länttä 1,347,000 1.06 6,633 Sweco Industry (2016)

Emmes 820,000 1.40 5,333 Sweco Industry (2016)Leviäkangas 400,000 1.01 1,877 Sweco Industry (2016)Outovesi 282,000 1.43 1,873 Keliber Oy (2017)Total 8,275,000 45,529

Granite, pegmatite (1.85-1.79 Ga)

Granite (1.89-1.87 Ga)

Granodiorite (1.89-1.88 Ga)

Tonalite, quartz diorite (1.89-1.88 Ga)

Gabbro, diorite, peridotite (1.89-1.87 Ga)

Intermediate-felsic volcaniclastic metasedimentary rocks (≤ 1.88 Ga)

Intermediate-felsic metavolcanic rocks (1.90-1.88 Ga)

Amphibolite, mafic metavolcanic rock (1.90-1.88 Ga)

Ultramafic-mafic metavolcanic rocks, chert interlayers (1.92-1.90 Ga)

Biotite paragneiss, metagreywacke (1.95-1.87 Ga)

Li deposit Li occurrence

!(

#*

#*

!(

!(

!(

!(

!(

#*

#*

Emmes

Länttä

Matoneva

Outovesi

Syväjärvi

Päiväneva

Rapasaari

Jänislampi

Leviäkangas

Heikinkangas

3 330 000 3 340 000 3 350 000

7 0

60

00

0

0 52.5 km

YKJ coordinates

Fig. 5. Geology of the Kaustinen area and location of the known LCT pegmatite lithium deposits and occurrences. The geological map is based on the GTK in-house digital bedrock database (Bedrock of Finland − DigiKP).

1111



2.2.1 The Rapasaari lithium deposit

The Rapasaari (formerly called Rapasaaret) deposit is the largest of the well-known lithium deposits in the Kaustinen province (Table 1), but compared with well-known lithium pegmatite deposits abroad, it is among the smallest (Appendix 2). The Rapasaari deposit is partly covered by a bog area with active peat production, and partly by shallow glacial till. No outcrops exist in the area. The deposit was dis-covered in 2009 by the Geological Survey of Finland, by diamond drilling guided by a spodumene peg-matite boulder fan (Kuusela et al. 2011)

The Rapasaari deposit consists of at least two spodumene pegmatite dyke swarms (Fig. 6). The eastern swarm strikes approximately 700 m to the northwest and dips southwest 40–50 degrees. The western dyke swarm has a strike length of 275 m, and according to diamond drilling infor-mation, its strike direction changes from north to northeast, and the dip is 60–75 degrees to the northwest. The dykes vary in thickness from 1 to 24 m. Available diamond drilling and till sampling information suggests that the eastern and western dykes belong to the same swarm system, which has intruded into a fold or fracture.

Typical host rocks for the Rapasaari pegmatites are mica schists, greywackes and intermediate vol-canic rocks (Kuusela et al. 2011). The mica schists are mostly greywackes that locally contain staurolite; in places, tremolite and garnet are present. The inter-mediate volcanic rock has plagioclase phenocrysts, biotite aggregates and locally green amphibole. The spodumene pegmatites have intruded both the mica schists and the intermediate volcanic rocks, mainly parallel to primary bedding (Sweco Industry 2016). Most commonly, they occur in mica schist in or

close to the contact zone with intermediate volcanic rock. Muscovite pegmatite veins with or without spodumene commonly occur associated with the spodumene pegmatite veins.

The main minerals in the spodumene pegma-tites are spodumene, albite, quartz, K-feldspar and muscovite. Apatite, zinnwaldite, Nb–Ta oxides (Mn and Fe tantalite), beryl, tourmaline, fluorite, gar-net (grossular), andalusite, calcite, chlorite, Fe–Mn phosphate, arsenopyrite, pyrite, pyrrhotite and sphalerite occur as accessory minerals. Spodumene occurs as elongated light green to light greyish, 0.5–10 cm long, lath-shaped crystals, which are oriented perpendicular to the wall-rock contact. Spodumene grains are unevenly distributed in the pegmatite dykes, but in most cases, the spodumene content increases from the wall-rock contact towards the core of the dyke. At the wall-rock contacts, spo-dumene has usually altered to muscovite (Sweco Industry 2016). The average content of spodumene in the Rapasaari pegmatite dykes is 14.7 wt%, and based on microprobe analyses, the average lithium concentration of spodumene in Rapasaari is 7.21% Li2O (Kuusela et al. 2011).

The average niobium and tantalum contents of the Rapasaari spodumene pegmatites are 58 ppm Nb2O5 and 53 ppm Ta2O5 (Kuusela et al. 2011). Both elements are carried by Fe columbite and Mn colum-bite. The grain size of these minerals in Rapasaaret is small, about 0.2–2 mm. The highest Nb and Ta contents occur within the most albite-rich zones.

The Rapasaari deposit is being developed by Keliber Oy. In 2017, the deposit was in the pre-feasibility stage. According to a recent estimate, the deposit has an indicated mineral resource of 3.456 Mt grading 1.15% Li2O (Keliber Oy 2017).

3 THE THREE-PART QUANTITATIVE RESOURCE ASSESSMENT METHOD

Numerous methods have been developed and applied to the estimation of undiscovered mineral resources during the past decades, but the task still remains challenging and there are no univer-sally accepted, definitive procedures (e.g., Lisitsin et al. 2007 and references therein). The procedure we selected is based on a three-part quantitative assessment method developed at the USGS start-ing from the mid-1970s (Singer 1975, Cox & Singer 1986, Root et al. 1992, Harris et al. 1993, Barton et

al. 1995, Singer 1993, Drew 1997, Singer & Menzie 2010) and increasingly used by the USGS and oth-ers since 1975 (e.g., Richter et al. 1975, Singer & Overshine 1979, Drew et al. 1984, Bliss 1989, Brew et al. 1992, Box et al. 1996, U.S. Geological Survey National Mineral Resource Assessment Team 2000, Kilby 2004, Lisitsin et al. 2007, 2014, Cunningham et al. 2008, Hammarstrom et al. 2010, 2013, 2014, Rasilainen et al. 2010a, 2012, 2014, Mihalasky et al. 2011, 2015a,b, Box et al. 2012, Ludington et al.

1212

!.

!.

!.

!.

!.

!.

!.

!.

!.!. !.

!.

!.

!.!.

!. !.

!. !.

!.

!.

!.

!.

!.

#*

#*

#*#*#*#*#*#*#*#*#*#*

#*#*#*#*#*#*#*#*#*#*#*#*#*#*#*#*#*#*#*#*

#*#*

#*

#* #*#*

#*

#*#*

#*#*#*#*

#*

#*

343 000 343 500 344 000

7 0

61

50

07

06

2 0

00

7 0

62

50

0

0 0.50.25 km

Finnish national KKJ zone 2 coordinates

Mica gneiss and mica schist,black schist intercalations

Intermediate volcanic rock

Spodumene pegmatite

Spodumene pegmatite boulder

Diamond drill hole

!.

!.

Fig. 6. Surface geology of the Rapasaari deposit at Kaustinen.

2012a,b, Sutphin et al. 2013, Cossette et al. 2014, Gray et al. 2014, Zientek et al. 2014a,b,c, 2015a,b, Zürcher et al. 2015, Wynn et al. 2016, Cocker et al. 2017). The method was considered well suited to accomplishing the goal of the GTK assessment pro-ject to estimate the undiscovered mineral endow-ment in Finland. The assessment is based on the statistical methods of data analysis and integration and it treats and expresses uncertainty. The method

enables the use of varying amounts of objective geological data and subjective expert knowledge and it generates reproducible assessment results.

The three-part method consists of the following components: (1) evaluation and selection or con-struction of descriptive models and grade-tonnage models for the deposit types under consideration; (2) delineation of areas according to the types of deposits permitted by the geology (permissive

1313



tracts); and (3) estimation of the number of undis-covered deposits of each deposit type within the permissive tracts. The estimated number of deposits

is combined with the grade and tonnage distribu-tions from the deposit models to assess the total undiscovered metal endowment.

3.1 Deposit models

Deposit models designed for quantitative assess-ments are the cornerstone of the method. They are used to classify mineralised and barren environ-ments, as well as types of known deposits, and to discriminate mineral deposits from mineral occur-rences (Singer & Berger 2007). Deposit models that can be used in the three-part assessment method include descriptive models, grade-tonnage mod-els, deposit density models, economic models and quantitative descriptive models. Descriptive models and grade-tonnage models are an essential compo-nent of the three-part method and they are used in all GTK assessments. Deposit density models, when available, can be used in the estimation of the num-ber of undiscovered deposits for an area. Economic models and quantitative descriptive models have not been used in the GTK assessment project.

3.1.1 Descriptive models

A descriptive model consists of systematically arranged information describing all of the essen-tial characteristics of a class of mineral deposits (Barton 1993). A descriptive model usually consists of two parts. The first part describes the geological environments in which the deposits occur. It con-tains information on favourable host rocks, possi-ble source rocks, age ranges of mineralisation, the depositional environment, tectonic setting, and associated deposit types. This part of the descrip-tive model plays a crucial role in the delineation of permissive tracts, i.e., areas where the geology permits the occurrence of deposits of the type under consideration.

The second part of a descriptive model lists the essential identifying characteristics by which a given deposit type might be recognised. These include ore textures and structures, mineralogy, alteration, and geochemical and geophysical signa-

tures. The second part of the model is used to clas-sify known deposits and occurrences. Identifying the types of known deposits is important for the tract delineation process, and it can sometimes help to delineate geological environments not indicated on geological maps.

3.1.2 Grade-tonnage models

A grade-tonnage model consists of data on average metal grades and the associated total tonnage of well-studied and completely delineated deposits of a certain type (Singer 1993, Singer & Menzie 2010). The total tonnage combines total past production and current resources (including reserves) at the lowest possible cut-off grade. Grade-tonnage mod-els are usually presented as frequency distributions of tonnage and average metal grades. These dis-tributions are used as models for grades and ton-nages of undiscovered deposits of the same type in geologically similar settings. They also help in differentiating between a deposit and a mineral occurrence, and in judging whether a deposit or group of deposits belongs to the type represented by the model.

It is very important to use the same sampling unit criteria for all deposits in the grade-tonnage model. Mixing old production data from some deposits with resource data from other deposits is among the most common errors in the construction of grade-tonnage models and will produce biased models (Singer & Berger 2007). Spatial aspects of the sampling unit must also be considered. A spatial rule identifying the minimum distance between two separate deposits of a given type should be defined, and deposits closer to each other than the minimum distance should be combined in the grade-tonnage model.

3.2 Permissive tracts

A permissive tract is an area within which the geol-ogy permits the existence of mineral deposits of the type under consideration (Singer 1993, Singer & Menzie 2010). It is important to distinguish between

areas favourable for the existence of deposits and permissive tracts: the former are a subset of the latter. The presence of a permissive tract in an area does not specify the level of favourability for the

1414

occurrence of deposits within the area; it only indi-cates the possibility for the existence of deposits. Furthermore, the existence of a permissive tract does not specify the likelihood of discovery of exist-ing undiscovered deposits in the area.

In the three-part assessment method, permis-sive tracts should be based on criteria derived from descriptive models. Tract boundaries should be defined so that the likelihood of deposits occurring outside of the tract is negligible. The boundaries of the tracts are first defined based on mapped or

inferred geology. Tracts may or may not contain known deposits. The existence of deposits is used to confirm and extend the tracts, but the lack of known deposits is not a reason to exclude any part of a permissive area from the tract. Original tract boundaries should only be reduced where it can be firmly demonstrated that a deposit type could not exist. This evidence could be based on geol-ogy, knowledge of unsuccessful exploration, or the presence of barren overburden exceeding the pre-determined delineation depth limit.

3.3 Estimation of the number of undiscovered deposits

The third part of the three-part assessment method is the estimation of the number of undiscovered deposits of the type(s) that may exist in the deline-ated tracts (Singer 1993, Singer & Menzie 2010). The estimates represent the probability that a certain fixed but unknown number of undiscovered depos-its exist in the delineated tracts. The estimates are carried out according to the deposit type and they must be consistent with the grade-tonnage models. This means that, for example, about half of the esti-mated undiscovered deposits should be larger than the median tonnage given by the grade-tonnage model, and about 10% of the estimated deposits should be larger than the upper 10th quantile of the model. The spatial rule used to define a deposit in the grade-tonnage model must be respected in the estimates. Well-explored and completely delineated deposits, for which published grade and tonnage values exist, are considered as discovered deposits, whereas deposits without publicly available grade and tonnage information, partly delineated depos-its, and known occurrences without reliable grade-tonnage estimates are counted as undiscovered.

Several methods can be used either directly or as guidelines to make the estimates. These include the frequency of deposits in well-explored geo-

logically analogous areas (deposit density mod-els), local deposit extrapolations, counting and assigning probabilities to geophysical and/or geo-chemical anomalies, process constraints, relative frequencies of associated deposit types, and limits set by the total available area or total known metal (Singer 2007). Some of these methods produce a single estimate of the expected number of depos-its; others produce a probability distribution of the expected number of deposits. In the latter case, the spread of the estimates for the number of deposits associated with high and low quantiles of the prob-ability distribution (for example, the 90% and 10% quantiles) indicates the uncertainty of the estimate. The expected number of deposits, or the estimated number of deposits associated with a given prob-ability level, measures the likelihood of the exist-ence of a deposit type.

The estimates are typically made by a team of experts knowledgeable about the deposit type and the geology of the region. The process follows the Delphi technique (Dalkey 1967, Sackman 1974, Rowe & Wright 1999), in which each expert makes an estimate independently and all the estimates are then discussed to possibly reach a final consensus estimate.

3.4 Statistical evaluation

The three parts of the assessment method described above produce consistent estimates of the number of undiscovered deposits for the delineated areas and of the probability distribution of grades and tonnages of the deposit type (Singer & Menzie 2010). As the final step of the assessment, these

estimates are combined using statistical methods to achieve probability distributions of the quantities of contained metals and ore tonnages in the undiscov-ered deposits. Software using Monte Carlo simula-tion has been developed for this purpose (Root et al. 1992, Duval 2012).

1515



4 ASSESSMENT OF LITHIUM RESOURCES IN FINLAND

4.1 Resources covered by the assessment

Here, we report the results of the assessment of undiscovered resources of lithium in LCT pegma-tite-hosted deposits. All the known lithium deposits and occurrences in Finland are hosted by pegmatites

(Alviola 2012, Ahtola et al. 2015), and no indica-tions of other types of lithium deposits are known in Finland.

4.2 GTK assessment process

The assessment process started with the selection of experts for the assessment team. As the work was conducted as a GTK internal project, only per-sons employed by GTK were assigned. To ensure good local knowledge of all covered areas, scien-tists based in GTK offices in northern (2), central (1) and southern (5) Finland were included in the assessment team. The work started with a work-shop on 9th February 2017, at which the partici-pants were introduced to the three-part assessment method. The types of known lithium deposits in Finland and their characteristics were reviewed and deposit types to be included in the assessment were selected. Responsibilities for the delineation and documentation of initial permissive tracts and for checking the existing grade-tonnage data for the selected deposit types were assigned to the assess-ment team members.

After the first workshop, the work continued with the delineation of preliminary permissive tracts, the preparation of tract description documents and the evaluation and development of descriptive and grade-tonnage models for the LCT pegmatite lithium deposits. Another workshop was arranged on 4th April 2017, in which the delineation of the permissive tracts was further refined. Assessment of the number of undiscovered deposits within the delineated permissive tracts was carried out in four successive workshops on 2nd June, 19th June, 28th August and 30th August 2017.

After the assessment workshops, Monte Carlo simulations were run to estimate the metal abun-dances in undiscovered deposits for each tract. The tract reports were finalised and all assessment documents were combined in this report.

4.3 Data used

The assessment team used geological maps in digi-tal and paper format, databases of mineral deposits and occurrences, technical reports on deposits and occurrences, mining company websites, and pub-lished geological literature. The personal experience and knowledge of the assessment team members concerning many of the areas assessed was a valu-able addition to the publicly available informa-tion. All data used in the assessment work for any permissive tract are listed in the respective tract description document in Appendix 3.

4.3.1 Geology

The GTK in-house GIS map database of Finnish bedrock (Bedrock of Finland – DigiKP) formed the main source of lithological data for this work. The multi-scale database covers the whole of Finland and is regularly updated. A version of the database

at the 1:200,000 scale can be viewed online at http://gtkdata.gtk.fi/mdae/index.html. Detailed maps pro-duced by exploration and research campaigns by various parties were also available for many areas assessed.

4.3.2 Known mineral deposits and occurrences

The compilations of lithium pegmatite deposits by Alviola (2003, 2012) served as a starting point in the collection of information on known lithium depos-its and occurrences in Finland. The GTK mineral deposits database, the Fennoscandian Ore Deposit Database, and compilations of metallogenic belts in the Fennoscandian Shield (Eilu 2012) were the main sources of information for deposits and occur-rences in Finland and the Fennoscandian Shield. These databases are available online via GTK web pages (Hakku search service). Where possible, the

1616

grade and tonnage data were checked and updated from company reports and publications. Reports in the GTK report database (Hakku search service), published literature on known deposits, prospects and occurrences, and mining company websites were used as additional sources of information on Finnish deposits and occurrences. For global data on LCT pegmatite lithium deposits, U.S. Geological Survey deposit models were used (Orris & Bliss 1992, Bradley et al. 2017), as well as the SNL data-base (SNL 2017).

4.3.3 Geophysical and geochemical data

Low-altitude airborne magnetic survey data of GTK (Hautaniemi et al. 2005) covering the whole of Finland were used to support the delineation of permissive tracts. Gravimetric maps, based on data provided by the Finnish Geodetic Institute (Kääriäinen & Mäkinen 1997) or on regional grav-ity measurements by GTK where available (Elo 2003), were also occasionally utilised. Regional till

geochemical data (Salminen 1995) and rock geo-chemical data (Rasilainen et al. 2007) were used in the delineation of permissive tracts. All the geo-physical and geochemical data mentioned in this paragraph are available via GTK web pages (Hakku search service).

4.3.4 Exploration history

Spatial data on the location of effective and recently expired exploration permits and the coverage of geophysical measurements, geochemical sampling and diamond drilling by GTK, Outokumpu Oy and some other exploration companies, combined with information in exploration reports (Hakku search service), were used to estimate the coverage and intensity of exploration activities in various areas in Finland. This information was supplemented with further information from exploration and mining company web sites, and with the personal knowl-edge of the assessment team members.

4.4 Deposit model

Grade and tonnage information was available for six LCT pegmatite lithium deposits in Finland and for 23 global deposits from outside Finland (Table 2 and Appendix 2). The median tonnage of the global deposits (15.7 Mt) is almost 15 times larger than the median tonnage of the Finnish deposits (1.08 Mt), but there is no significant difference in the median lithium grades between the global (1.19% Li2O) and Finnish (1.20% Li2O) deposits. At least four of the six Finnish deposits with a published resource estimate are not totally delineated but remain open either at depth or along strike, or both (Sweco Industry 2016). This, together with the similarity of average lithium grades in the Finnish and global deposits,

suggests that there is no separate Finnish lithium pegmatite population. On the other hand, several of the global deposits also are not entirely deline-ated. In any case, the number of Finnish deposits is too small for a stable deposit model, and hence, the global data were used to construct a grade-tonnage model for LCT lithium pegmatite deposits (Table 2). The distributions of lithium grades and ore tonnages in the global model do not signifi-cantly deviate from lognormality, and there is no significant correlation between grades and tonnages (Fig. 3 in App. 2). The model is described in detail in Appendix 2.

4.5 Tract delineation

Permissive tracts were delineated for the LCT peg-matite lithium deposits based on the information described in section 4.3. For each area, the member or members of the assessment team most familiar with the geology and mineralisation of the region delineated the tracts. The criteria for the deline-

ation of each tract are given in the tract reports in Appendix 3. On a map, a permissive tract is a projection to the surface of the domain where geol-ogy allows the existence of the deposit type being assessed. The permissive volumes of rock were delineated down to the depth of one kilometre.

1717



Table 2. Summary statistics for the global and local (Finnish) models, and for deposits outside Finland.

Global model (Finnish and other deposits)Tonnage (t) Li2O (%) Li (%) Ta2O5 (%)

Number of deposits 29 29 29 9Minimum 282,000 0.87 0.40 0.007Maximum 174,000,000 1.57 0.73 0.020Arithmetic Mean 30,279,655 1.20 0.56 0.013Standard Deviation 43,685,771 0.19 0.09 0.00410th percentile 1,030,800 1.00 0.47 0.00750th percentile 10,320,000 1.19 0.55 0.01390th percentile 102,400,000 1.44 0.67 0.018Shapiro-Wilk p-value* 0.725 0.371 0.371 0.763

Local model (Finnish deposits)Tonnage (t) Li2O (%) Li (%) Ta2O5 (%)

Number of deposits 6 6 6 0Minimum 282,000 1.01 0.47 -Maximum 3,456,000 1.43 0.66 -Arithmetic Mean 1,379,167 1.22 0.56 -Standard Deviation 1,194,869 0.17 0.08 -10th percentile 293,800 1.02 0.47 -50th percentile 1,083,500 1.20 0.56 -90th percentile 3,307,400 1.43 0.66 -Shapiro-Wilk p-value* 0.885 0.567 0.567 -

Deposits outside FinlandTonnage (t) Li2O (%) Li (%) Ta2O5 (%)

Number of deposits 23 23 23 9Minimum 1,353,000 0.87 0.40 0.007Maximum 174,000,000 1.57 0.73 0.020Arithmetic Mean 37,818,913 1.20 0.56 0.013Standard Deviation 46,275,169 0.19 0.09 0.00410th percentile 2,976,000 0.97 0.45 0.00750th percentile 15,700,000 1.19 0.55 0.01390th percentile 122,200,000 1.44 0.67 0.018Shapiro-Wilk p-value* 0.905 0.609 0.609 0.763* Shapiro-Wilk normality test p-value was calculated for logarithmic tonnage and grade values.

4.6 Estimation of the number of undiscovered deposits

The number of undiscovered LCT pegmatite lithium deposits consistent with the global grade-tonnage model (Chapter 4.4) was estimated separately for each permissive tract in a series of workshops by the members of the assessment team. The names of the estimators for each permissive tract are given in the tract reports in Appendix 3. Each estimator independently assessed the number of undiscovered deposits at the 90%, 50%, 10%, 5% and 1% proba-bility levels. Only three probability levels (90%, 50%

and 10%) are normally used in GTK assessments. In the case of LCT pegmatite lithium deposits, the average deposit size in the grade-tonnage model is so much larger than the known deposits in Finland that it was considered necessary to also use the 5% and 1% probability levels. The initial estimated numbers of undiscovered deposits were provided for discussion, during which the participants explained and sometimes adjusted their estimates. The pur-pose of the discussion was to determine whether a

1818

consensus estimate could be reached. When a con-sensus was not reached, the averages of the expert estimates at each probability level were used as the final estimates at the 90%, 50%, 10%, 5% and 1% probability levels.

According to the three-part method, incompletely known deposits and occurrences are considered as undiscovered. The estimators assess the prob-ability that an incompletely known deposit might

with further exploration become a deposit with a potential for economically viable mining. However, the well-known deposits that are included in the grade-tonnage model are by definition assumed to be totally delineated and without any remain-ing unknown resources. This means that the pos-sibly existing hidden resources in the well-known deposits are not considered and remain outside the assessment.

4.7 Assessment of metal tonnages

The assessment of metal tonnages in the undis-covered deposits was performed using Eminers software (Root et al. 1992, Duval 2012). As input, the software uses data from the grade-tonnage model and the estimated numbers of undiscov-ered deposits at the 90%, 50%, 10%, 5% and 1% probability levels. The software estimates an aver-age non-parametric frequency distribution for the number of undiscovered deposits within a tract. It also estimates empirical and lognormal frequency distributions for the ore tonnage and metal grades in the grade-tonnage model. The software then runs Monte Carlo simulations using the estimated frequency distributions to produce estimates of ore and metal tonnages in the undiscovered deposits.

For each simulation round, metal tonnages are cal-culated by multiplying ore tonnage and metal grade values selected randomly from the constructed empirical or lognormal frequency distributions. As the final result of the simulation, the software produces probability distributions of the amount of metals and ore in the undiscovered deposits.

The assessment of metal tonnages in the undis-covered deposits was performed separately for each permissive tract. It is not statistically correct to add together the frequency distributions of ore and metal tonnages produced for the tracts. Hence, the total metal endowment for all undiscovered LCT pegmatite lithium deposits was estimated in a sepa-rate simulation run.

5 RESULTS AND DISCUSSION

The results of the Monte Carlo runs are simulated frequency distributions of ore and metal tonnages in the undiscovered deposits. These distribu-tions combine the amount of undiscovered metal and the probability that this amount exists. The results of the LCT pegmatite lithium assessments are summarised in Tables 4 and 5 and Figures 8 and 9. Detailed assessment information for each permissive tract is presented in Appendix 3, where cumulative frequency distributions of undiscovered metal and ore tonnages are plotted and metal ton-nages corresponding to several probability values are tabulated.

The arithmetic mean value of the frequency dis-tribution for metal tonnage can be considered as the expected amount of undiscovered metal. Since the frequency distributions of lithium grades and ore tonnages in the LCT pegmatite lithium deposit grade-tonnage model are skewed and closer to a lognormal than a normal distribution, the probabil-

ities associated with the expected (mean) amounts of undiscovered lithium metal and ore are less than 50%, around 30% for aggregated results (Table 5) and generally less than 5% for individual permis-sive tracts (Appendix 3). On the other hand, there is a 50% probability that at least the amount of lithium given by the median of the simulated fre-quency distribution (value associated with the 50% quantile) exists. This is why we prefer to report the median estimate of undiscovered lithium metal and ore in Table 4, in which the results for individual permissive tracts are summarised. Table 5, which aggregates the total undiscovered lithium endow-ment over a number of permissive tracts, gives several quantile values to better characterise the frequency distributions of metal and ore tonnages. In the following, median estimates, that is, values corresponding to the 50% quantile of the cumula-tive frequency distribution, are used when discuss-ing the amounts of undiscovered lithium metal.

1919



5.1 Permissive tracts delineated

In total, 19 permissive tracts were delineated for LCT pegmatite lithium deposits (Fig. 7, Table 3, Appendix 3). These tracts contain all the known lithium deposits and significant occurrences in Finland. Altogether, the tracts cover an area of 22,404 km2, which is approximately 7% of the

total land area of Finland. The size of the permis-sive tracts varies from 16 km2 to 6624 km2 (Table 3), and the median area is 417 km2. Maps of the individual permissive tracts are included in the tract reports in Appendix 3.

"

"

"

"

"

"

"

"

"

Oulu

Turku

VaasaKuopio

Tampere

Joensuu

Helsinki

Jyväskylä

Rovaniemi

3 000 000 3 200 000 3 400 000 3 600 000 3 800 000

6 8

00

00

07

00

0 0

00

7 2

00

00

07

40

0 0

00

7 6

00

00

0

Caledonian orogenic belt

Palaeo - Mesoproterozoic rocks

Rapakivi granites

Sedimentary rocks

Dolerites

Palaeoproterozoic rocks

Intrusive rocks

Supracrustal rocks

Granulitic rocks

Archaean rocks

Intrusive rocks and gneisses

Supracrustal rocks

0 100 20050 km

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

19

18

Permissive tracts forLCT pegmatite lithium

Fig. 7. Location of the LCT pegmatite lithium permissive tracts in Finland. Numbers on the map refer to Table 3.

2020

Table 3. Permissive tracts for LCT pegmatite lithium deposits in Finland.

Map number Tract name Tract area (km2)

N of known deposits

1 Eräjärvi Li 124 02 Heinola Li 16 03 Järvi-Pohjanmaa Li 3,672 04 Kalajoki Li 417 05 Kaustinen Li 255 66 Kemiö Li 51 07 Kevo Li 392 08 Kisko-Orijärvi-Pohja Li 168 09 Kitee-Tohmajärvi Li 429 010 Muhos Li 2,727 011 Parikkala Li 231 012 Pello Li 1,388 013 Rantasalmi Li 84 014 Seinäjoki Li 800 015 Somero-Tammela Li 466 016 Sotkamo Li 32 017 Tuntsa Li 6,624 018 Vainospää Li 2,927 019 Vanttaus Li 1,601 0

5.2 Undiscovered resources in LCT pegmatite lithium deposits

The expected (mean) number of undiscovered LCT pegmatite lithium deposits for a permissive tract varies from 0.03 to 2.6 deposits, and the sum of the mean estimates across all permissive tracts is 6.8 deposits (Table 4). Over 80% of the undiscov-ered deposits are estimated to be located within three permissive tracts: Kaustinen (38%), Järvi-Pohjanmaa (29%) and Somero-Tammela (15%).

The median estimate of the total in situ lithium metal content in undiscovered LCT pegmatite lith-ium deposits in Finland is at least 510,000 t of Li (Table 5), and 61% of this is estimated to be located in undiscovered deposits in the Kaustinen permis-sive tract (Table 4, Fig. 9).

The assessment results indicate that the most important lithium resources in Finland are con-

centrated in the Kaustinen tract and in the sur-rounding larger Järvi-Pohjanmaa tract. The Somero-Tammela tract in southern Finland also has a non-zero median estimate of undiscovered lithium resources (46,000 t Li). In addition to the above mentioned permissive tracts, only the Kisko-Orijärvi-Pohja tract in the southernmost part of Finland has an average estimated undiscovered lithium content associated with a probability larger than 10%. For all the remaining permissive tracts in Finland, the probability of containing at least the amount indicated by the mean estimate is less than 10% (Fig. 9), and for more than half of these, it is less than 5% (Appendix 3).

2121



Tab

le 4

. Su

mm

ary

of p

rin

cip

al r

esu

lts

for

LC

T p

egm

atit

e li

thiu

m p

erm

issi

ve t

ract

s in

Fin

lan

d.

Trac

t na

me

Esti

mat

ed n

umbe

r of

un

disc

over

ed d

epos

its

Dis

cove

red

reso

urce

s (t

)M

edia

n un

disc

over

ed

reso

urce

s (t

)

Mea

n un

disc

over

ed

reso

urce

s (t

)

Tota

l de

pos-

its

Trac

t ar

ea

(km

2 )

Dep

osit

de

nsit

y (N

/ km

2 )

9050

105

1E

Std

CvN

LiO

reLi

Ore

LiO

reEr

äjär

vi L

i0

00

23

0.18

0.64

350

00

00

031

,000

5,40

0,00

00.

1812

40.

0015

Hei

nola

Li

00

00

10.

030.

2481

00

00

00

4,10

070

0,00

00.

0316

0.00

19

Järv

i-Poh

janm

aa L

i0

24

2.0

1.5

730

00

150,

000

27,0

00,0

0034

0,00

060

,000

,000

2.0

3,67

20.

0005

4Ka

lajo

ki L

i0

00

01

0.03

0.24

810

00

00

05,

700

970,

000

0.03

417

0.00

0072

Kaus

tinen

Li

13

42.

61.

143

645

,529

8,27

5,00

031

0,00

057

,000

,000

470,

000

83,0

00,0

008.

625

50.

034

Kem

iö L

i0

00

12

0.11

0.44

420

00

00

019

,000

3,30

0,00

00.

1151

0.00

22Ke

vo L

i0

00

01

0.03

0.24

810

00

00

04,

800

870,

000

0.03

392

0.00

0077

Kisk

o-O

rijär

vi-

Pohj

a Li

00

12

30.

410.

8220

00

00

00

72,0

0013

,000

,000

0.41

168

0.00

24

Kite

e-To

hmaj

ärvi

Li

00

01

10.

080.

3242

00

00

00

12,0

002,

100,

000

0.08

429

0.00

019

Muh

os L

i0

00

01

0.03

0.24

810

00

00

04,

800

840,

000

0.03

2,72

70.

0000

11Pa

rikka

la L

i0

00

01

0.03

0.24

810

00

00

05,

700

1,00

0,00

00.

0323

10.

0001

3Pe

llo L

i0

00

02

0.06

0.37

610

00

00

011

,000

1,90

0,00

00.

061,

388

0.00

0043

Rant

asal

mi L

i0

00

01

0.03

0.24

810

00

00

04,

600

790,

000

0.03

840.

0003

6Se

inäj

oki L

i0

00

01

0.03

0.24

810

00

00

05,

400

920,

000

0.03

800

0.00

0038

Som

ero-

Tam

mel

a Li

01

21.

00.

7979

00

046

,000

8,50

0,00

017

0,00

030

,000

,000

1.0

466

0.00

21

Sotk

amo

Li0

00

01

0.03

0.24

810

00

00

04,

900

920,

000

0.03

320.

0009

4Tu

ntsa

Li

00

00

20.

060.

3761

00

00

00

10,0

001,

800,

000

0.06

6,62

40.

0000

091

Vain

ospä

ä Li

00

00

10.

030.

2481

00

00

00

4,20

072

0,00

00.

032,

927

0.00

001

Vant

taus

Li

00

00

10.

030.

2481

00

00

00

5,90

01,

000,

000

0.03

1,60

10.

0000

19G

rand

tota

l6.

86

45,5

298,

275,

000

506,

000

92,5

00,0

001,

185,

100

209,

230,

000

12.8

22,4

04

90, 5

0, 1

0, 5

, 1: E

stim

ated

num

ber o

f und

isco

vere

d de

posi

ts a

ssoc

iate

d w

ith th

e 90

%, 5

0%, 1

0%, 5

% a

nd 1

% q

uant

iles;

E: E

xpec

ted

num

ber (

mea

n es

timat

e) o

f und

is-

cove

red

depo

sits

; Std

: Sta

ndar

d de

viat

ion;

Cv:

Coe

ffici

ent o

f var

iatio

n (%

); N

: Num

ber o

f wel

l-kno

wn

depo

sits

with

in th

e pe

rmis

sive

trac

t; O

re: M

iner

aliz

ed ro

ck c

onta

i-ni

ng th

e m

etal

s; M

edia

n es

timat

ed u

ndis

cove

red

reso

urce

s: T

he m

inim

um a

mou

nt o

f met

als

pres

ent w

ithin

the

perm

issi

ve tr

act a

t the

pro

babi

lity

of 5

0%, r

ound

ed to

tw

o si

gnifi

cant

dig

its; M

ean

undi

scov

ered

reso

urce

s: A

rithm

etic

mea

n of

the

estim

ated

am

ount

of m

etal

s pr

esen

t with

in th

e pe

rmis

sive

trac

t, ro

unde

d to

two

sign

ifi-

cant

dig

its; T

otal

dep

osits

: Sum

of t

he n

umbe

r of w

ell-k

now

n de

posi

ts a

nd e

xpec

ted

num

ber o

f und

isco

vere

d de

posi

ts w

ithin

the

trac

t; D

epos

it de

nsity

: Tot

al n

um-

ber o

f dep

osits

with

in th

e tr

act d

ivid

ed b

y tr

act a

rea,

roun

ded

to tw

o si

gnifi

cant

dig

its; G

rand

tota

l: Su

m a

cros

s al

l tra

cts.

E a

nd S

td a

re c

alcu

late

d us

ing

a re

gres

sion

eq

uatio

n (S

inge

r & M

enzi

e 20

05).

Alth

ough

it is

not

sta

tistic

ally

str

ictly

cor

rect

to s

um th

e m

edia

n es

timat

es fo

r ind

ivid

ual t

ract

s, th

e re

sulti

ng to

tals

are

ver

y cl

ose

to

the

valu

es o

btai

ned

by in

clud

ing

all t

he tr

acts

in o

ne s

imul

atio

n (T

able

5).

2222

Table 5. Summary of the well-known and estimated undiscovered resources in LCT pegmatite lithium deposits in Finland.

Well known

At least the indicated amount at the probability of Mean Probability of0.95 0.90 0.50 0.10 0.05 Mean or

greaterNone

Li (t) 45,529 0 12,000 510,000 3,000,000 4,200,000 1,200,000 0.30 0.07

Ore (Mt) 8.275 0 2.1 93 520 740 210 0.30 0.07Ore: Mineralised rock containing the metals. Well-known resources as of 20th June 2017. Data sources are listed in Appendix 2. The undiscovered resources are rounded to two significant digits.

0.40

0.50

0.60

0.70

0.80

0.90

1.00

0.00

0.10

0.20

0.30

Pro

babili

tyP

robabili

tyP

robabili

ty

Material

1E-02 1E-01 1E+00 1E+01 1E+02 1E+03 1E+04 1E+05 1E+06 1E+07 1E+08 1E+09 1E+10 1E+11 1E+12

Li (t) Ore (t)

Fig. 8. Cumulative frequency distributions of simulated undiscovered resources in LCT pegmatite lithium deposits in Finland. Labelled dots indicate mean values.

2323



LCT_pegmatite_resources

Kevo Li

Vainospää Li

Tuntsa Li

Vanttaus LiPello Li

Muhos Li

Sotkamo LiKalajoki Li

Kaustinen Li

Järvi-Pohjanmaa Li

Seinäjoki Li

Rantasalmi Li

Parikkala LiEräjärvi Li

Heinola Li

Somero-Tammela Li

Kemiö Li

Kitee-Tohmajärvi Li

Kisko-Orijärvi-Pohja Li

< 5,000

5,000 - 10,000

10,000 - 50,000

50,000 - 100,000

100,000 - 200,000

200,000 - 300,000

300,000 - 400,000

400,000 - 500,000

Undiscovered resourcesLi (t)

Median undiscovered resource

Mean undiscovered resource,probability of mean or more > 10%

Mean undiscovered resource, probability of mean or more < 10%

Fig. 9. Estimated undiscovered LCT pegmatite lithium resources in Finland plotted on each permissive tract.

2424

5.3 Finnish lithium endowment in national and global contexts

The identified LCT pegmatite lithium resources in Finland are 45,529 t of lithium (Table 1). Comparison with the estimated undiscovered lithium resources in Finland (Table 5) indicates that more than 90% of the LCT pegmatite lithium endowment within the uppermost one kilometre of the Finnish bedrock is in poorly explored or entirely unknown deposits. The Kaustinen permissive tract and the surrounding larger Järvi-Pohjanmaa tract host the most impor-tant known and estimated undiscovered lithium resources in Finland. The Kaustinen deposits are under active development, and it can be expected that exploration within the area surrounding an operating lithium mine will in the future turn at least a part of the presently undiscovered resources into identified resources.

Identified world reserves are 14 Mt of lithium, and world resources are approximately 47 Mt of lithium (Jaskula 2017). Due to the increased lithium exploration in the 2010s, the global resource esti-mate has increased markedly from the 13.8 Mt in 2009 (Jaskula 2009). The proportion of pegmatite lithium resources of the total global resources is approximately 25%, based on the breakdown of Evans (2014). This equals 11.8 Mt of lithium. The remaining global resources are in continental brines (63%) and in other kinds (hectorite, jadarite, geo-thermal brines, oilfield brines) of deposits (12%). Identified European lithium resources are approxi-mately 3 Mt of lithium (Table 6), and only eight per cent of these (0.249 Mt Li) are pegmatite-hosted.

Of the lithium resources in Europe, 59% (1.77 Mt Li) are within the European Union (EU).

Global mine production in 2016 was approxi-mately 35,000 t of contained lithium (Jaskula 2017). Most of the production came from Australia (14,300 t Li, pegmatites), Chile (12,000 t Li, brines), Argentina (5700 t Li, brines) and China (2000 t Li, pegmatites and brines). European mine production of lithium came from Portugal, which produced 200 t of lithium from pegmatite deposits in 2016.

According to an analysis by Deutsche Bank (2016), the global lithium demand is expected to increase from 181,000 t lithium carbonate equiva-lent (LCE) in 2015 to 535,000 t LCE by 2025. The lithium supply is also expected to increase, and global lithium resources are expected to be suffi-cient to support demand for the remainder of the 21st century (Gruber & Medina 2010, Kesler et al. 2012, Deutsche Bank 2016).

In a global or European-wide comparison, the Finnish identified and estimated undiscovered lith-ium resources have little significance. The identified Finnish resources of 45,529 t of lithium constitute 0.1% of the known world resources, 1.5% of the European and 2.6% of the EU lithium resources. If only lithium pegmatite resources are considered, the Finnish identified resources are 18% of the European and 26% of the EU resources. There have been no global assessments of undiscovered lithium resources to compare with the results from Finland.

Table 6. Lithium resources in Europe. Only deposits with published resource estimates are included.

Deposit Country Type Tonnage (t) Li (%) Li (t) Reference

Cinovec Czech Republic Greisen 656,500,000 0.19 1,219,961 European Metals Holdings Ltd (2017)

Jadar Serbia Hydrothermal 135,700,000 0.86 1,172,584 Rio Tinto (2017)San Jose Spain Replacement 92,300,000 0.27 249,210 Plymouth Minerals Ltd (2017)Zinnwald Germany Greisen 36,437,000 0.36 132,739 Bacanora Minerals Ltd (2017)Shavazsai Uzbekistan Pegmatite 31,000,000 0.25 76,905 Levine & Steblez (1997)Alberta-1 Spain Pegmatite 30,990,000 0.14 44,631 Barlett (2014)Sepeda Portugal Pegmatite 10,300,000 0.46 47,851 Dakota Minerals (2017)Kaustinen* Finland Pegmatite 8,275,000 0.55 45,529 Sweco Industry (2016), Keli-

ber Oy (2017)Wolfsberg Austria Pegmatite 6,303,300 0.54 34,261 European Lithium (2016)Total 1,007,805,300 3,023,670

Li (%) was reported for Zinnwald and San Jose, and it has been calculated for the other deposits, based on the reported Li2O grade. * The Kaustinen tonnage figure is a total for six separate deposits, and the Li grade is the average for these deposits (Table 1).

2525



Contrary to the global context, the discovery of at least some proportion of the undiscovered lithium resources in Finland would have a notable effect. According to the published pre-feasibility study for the Kaustinen deposits (Sweco Industries 2016), based on an annual production of 6000–9000 t of lithium carbonate, the annual output of the planned open pit mines could be 275,000–400,000 t of ore. This would facilitate an operating time of 16–11 years for the planned lithium carbonate plant. According to this assessment, there is a 90% probability that

at least 2.1 Mt of undiscovered lithium pegmatite ore exist within the Kaustinen tract (Table 4). If this resource was discovered, and assuming a similar reserve to resource ratio and concentrator through-put as in the published pre-feasibility study, the operating time of the lithium plant at Kaustinen could be increased by 40%. The discovery of new resources corresponding to the median estimate of undiscovered resources in the Kaustinen tract could increase the plant lifetime tenfold.

5.4 Reliability and usability of the estimates

Considering the reliability of the assessment results, sensitivity analysis indicates that changes in grade and tonnage estimates have a much larger influence on the expected metal content in an assessment than changes in the expected number of deposits (Singer & Kouda 1999). Consequently, the greatest sources of possible error in the present assessments are associated with the grade-tonnage models used.

It is of utmost importance that the grade and tonnage information included in the grade-tonnage model represents as accurately as possible total deposits of the correct deposit type. However, even deposits that are considered to be well known may contain undiscovered resources. This is certainly the case with pegmatite-hosted lithium depos-its; it is possible that most of the deposits in the grade-tonnage model (Appendix 2) are not entirely delineated, and during the assessment process in 2017, new resource estimates were published at least once for several of the deposits. Commonly, the ore tonnages had increased and the lithium grades had slightly decreased in the newer resource estimates, resulting in an increased lithium metal content in the deposit. This suggests that the grade-tonnage model to some extent underestimates the lithium contents of LCT pegmatite lithium deposits. Consequently, the amounts of lithium in the undis-covered resources estimated based on the grade-

tonnage model are somewhat smaller than the true lithium contents in the undiscovered deposits.

Finally, care must always be taken when applying the results of assessments such as these. Although the assessment method predicts the existence of a number of undiscovered deposits, it gives no guarantee that these deposits will ever be discov-ered. Many of the undiscovered deposits estimated to exist by this work can be under hundreds of metres of barren rock, whereas others may crop out at the surface. Some of the buried deposits are likely to be beyond the reach of present day explo-ration technology, or their discovery may require exploration expenditures so large they are unlikely to be discovered in the near future. As the grade-tonnage models used in three-part assessments typically contain uneconomic occurrences in addi-tion to operating mines, the resulting estimated undiscovered resources are also partly located in uneconomic occurrences. Although technological advances act over time to lower mining costs and allow formerly uneconomic occurrences to become operating mines, some of the undiscovered deposits estimated here might never be mined for one or more reasons, including relatively low tonnages or grades, deep burial, or occurrence in or near envi-ronmentally sensitive areas or in areas designated for other land uses than mining.

6 SUMMARY

Experts in lithium pegmatite deposits, metallogeny and geostatistics at the Geological Survey of Finland produced this assessment of undiscovered LCT peg-matite lithium resources within the uppermost one kilometre of the Finnish bedrock. The resources

were assessed using the three-part quantitative method developed at the U.S. Geological Survey. This report provides numerical estimates of the expected endowment of lithium in undiscovered, potentially exploitable LCT pegmatite lithium deposits.

2626

The main results include the following:1. The three-part quantitative assessment method

is suitable for assessing LCT pegmatite lithium deposits.

2. The number of well-known LCT pegmatite lith-ium deposits in Finland is six, and all of these are located in the Kaustinen area. The identi-fied resources in these deposits are 45,529 t of lithium.

3. Compared with global LCT pegmatite lithium deposits, the Finnish deposits have an approxi-mately one order of magnitude smaller median tonnage, but similar median lithium grade.

4. A global grade-tonnage model was constructed using information on 29 LCT pegmatite lith-ium deposits from seven countries. Most of the deposits in the grade-tonnage model are prob-ably not totally delineated.

5. Nineteen permissive tracts were delineated for LCT pegmatite lithium deposits in Finland. The tracts cover 22,404 km2, which is approximately 7% of the total land area of Finland.

6. The expected number of undiscovered LCT pegmatite lithium deposits in Finland is 6.8 deposits.

7. The undiscovered LCT pegmatite lithium depos-its are estimated to contain, at 50% probability, at least 510,000 t of lithium.

8. The assessment results indicate that at least 90% of the remaining lithium endowment within the uppermost one kilometre of the Finnish bedrock is in poorly explored or entirely unknown deposits.

9. Over 90% of the estimated undiscovered lith-ium resources are located within either the Kaustinen permissive tract or the surrounding larger Järvi-Pohjanmaa tract.

10. The identified lithium resources in Finland constitute 0.1% of the known world resources, 1.5% of the European and 2.6% of the EU lithium resources. There are no global assessments of undiscovered lithium resources to compare with the results from Finland.

ACKNOWLEDGMENTS

We thank many of our colleagues at GTK for provid-ing data and insights into details of the geology of lithium deposits in Finland. Jussi Pokki is thanked for preparing the map figures for all the permissive

tract reports in Appendix 3, Alan Butcher and an anonymous reviewer for reviewing the manuscript and Roy Siddall for correcting the English language.

REFERENCES

Ahtola, T., Kuusela, J., Käpyaho, A. & Kontoniemi, O. 2015. Overview of lithium pegmatite exploration in the Kaustinen area in 2003–2012. Geological Survey of Finland, Report of Investigation 220. 28 p.

Alviola, R. 2003. Pegmatiittien malmipotentiaalista Suomessa. Geological Survey of Finland, archive re-port M10/-03/1/85. 6 p. (in Finnish)

Alviola, R. 2012. Distribution of rare element pegmatites in Finland. In: Kukkonen et al. (eds) Lithosphere 2012. Seventh Symposium on the Structure, Composition and Evolution of the Lithosphere in Finland, Espoo, November 6−8, 2012, programme and extended ab-stracts. University of Helsinki, Institute of Seismolo-gy, report S-56. Helsinki: Institute of Seismology, 1−4.

Alviola, R., Mänttäri, I., Mäkitie, H. & Vaasjoki, M. 2001. Svecofennian rare-element granitic pegmatites of the Ostrobothnia region, western Finland; their metamor-phic environm

bulletin 406 research report · rasilainen kalevi, eilu pasi, ahtola timo, halkoaho tapio,...

Documents