promine project – wp1promine.gtk.fi/documents_news/promine_final_conference/10_10_be… ·...

ProMine Project – WP1

PART 2: Mineral Potential and Predictive Mapping

ProMine final conference

Levi, April 23rd 2013

G. Bertrand, D. Cassard, M. Billa, B. Tourlière, and the WP1 Team

ProMine WP1 Introduction

WP1 has produced a huge homogeneous dataset of primary and secondary mineral resources that covers the whole EU. The objective of WP1 was to go farther, and to valorize this dataset by producing added value layers. Two types of added value layers : - Mineral potential maps, to

assess geographic distribution of known ores

- Mineral predictive maps, to

identify areas that could favorably contain targeted commodities

ProMine final meeting – Levi – 23rd April 2013

ProMine WP1

Summary • Introduction

• Main commodity associations/deposit types

• Mineral potential mapping

- Methodology - Results and examples

• Mineral predictive mapping with the Weight of Evidence method


• Mineral predictive mapping with the Database Querying method


• Conclusion

Introduction



ProMine WP1 Main commodity associations

In order to analyse homogeneous mineralization, a list of 16 most characteristic commodity associations – or deposit types - has been established by WP1 partners. The ProMine MD database was then queried to extract all deposits of these 16 major types. As a result, 16 homogeneous deposit populations were obtained, that were processed for potential and predictive mapping.


Number Association Name Commodity Association 'Type’ codes queried

in MD database (including ‘sons’)

1 Alkaline & Peralkaline intrusions

Nb, REE, P, (Ta, Zr, Sc, F, U, Fe)

C10, C20

2 Epithermal Au, Ag, Sb, Hg, Te, Cu, In D

3 Igneous Felsic Sn, W, Ta, Nb, (Mo, Li, Be, B, In, F)

C40

4 Igneous Intermediate Cu, Mo, Au, (Re) C50 5 Igneous Replacement Fe, W, Pb, Zn, Cu, Au C70 6 IOCG Fe, Cu, Au, (P, REE, U, Co) K 7 Mafic intrusion Fe, Ti, V B30

8 Mafic or UltraMafic Ni, Cr, Cu, PGE, (Co, Bi, U, Ag)

B, except B30

9 Orogenic Gold Au, (Ag, As, W, Cu, Sb, Bi) A, plus 'commodity Gold'

10 Pegmatites Nb, Ta, Sn, Li, Be, (U, REE) C60 11 Carbonate-hosted deposits Zn, Pb, Ag, Ba F40

12 Sandstone- and shale-hosted deposits

Cu, U, Pb, (Ni, Co, Zn, V, PGE, Re)

F20, F30, F60

13 Sedimentary deposits Fe, Mn, Ba,K,Na,Sr F50

14 VMS Cu, Zn, Pb, (Ag, Au, Te, Sn, In)

E

15 Residual deposits Fe, Al, Ni, Cu, (Mn, Au, P, REE)

H20, H30

16 Base metals veins Pb, Zn, Cu, U, (Ba, F) A, without 'commodity Gold'

ProMine WP1 Mineral potential mapping - methodology

Principles: The goal of potential mapping is to identify areas of high mineral resources potential, based on the distribution and size of known deposits of a given type. The principle of the method was to combine: - a statistical study of the spatial distribution of deposits (kernel density in a first stage);

- the introduction, as a weight, of the size (class of the main commodity) of the deposits.

- geological constrains, by selecting lithology polygons containing deposits of the selected type.

The interest of such a methodical approach is that it is reproducible, whatever the type and the density of deposits.


ProMine WP1

Potential map calculation: For each population (i.e. commodity association), the potential map is the sum of: 1) A density grid: kernel density analysis using the ninth decile of proximity statistics as search radius and weighting deposits according to their class (from 1 for class E or 'no class', to 25 for class A, based on the highest class of the deposit) 2) A density coverage grid: contouring of the weighted density grid (interval equal to one decile of the density statistics) and extraction of the surface inside a buffer equal to the mean value of the density statistics 3) A geology coverage grid: extraction of the surface covered by geology polygons containing at least one deposit of the population.

Map of “geology- controlled deposit density”

Geology coverage grid Density coverage grid Density grid

Mineral potential mapping - methodology


ProMine WP1

One map for each commodity association (or deposit type)

1 - Alkaline-peralkaline intrusions 2 - Epithermal and volcanic systems

3 - Igneous felsic 4 - Igneous intermediate

5 - Igneous replacement or skarn 6 - IOCG (Iron Oxide Copper Gold)

7 - Mafic intrusions 8 - Mafic-ultramafic

9 - Orogenic gold 10 - Pegmatites

11 - Carbonate hosted 12 - Sandstone and shale hosted

13 - Sedimentary deposits 14 - VMS (Volcanogenic Massive Sulfides)

15 - Residual deposits 16 - Base metals veins

Mineral potential mapping - results


ProMine WP1

Orogenic gold: Distribution of potential is guided by a single main commodity (Au) and a well constrained type of mineralization. Major districts belong to two groups: - Paleoproterozoic orogenic deposits

related to greenstones in the Fennoscandian shield;

- Hercynian gold-bearing districts related

to late Hercynian (~300 Ma) deformation belts (N. Iberian peninsula, French Massif Central, Bohemian Massif).

Additional more scattered deposits can be found in other Hercynian (Salsigne) or Caledonian (Great Britain and Norway) domains and the Balkan-Carpathian region.



ProMine WP1

Igneous intermediate: Potential of igneous intermediate mineralization clearly highlights the western Tethyan suture, with large porphyry-type districts: - Kremnica in Slovakia, - Telkibanya in Hungary, - The Apuseni mounts in Romania, - Bor in Serbia, - Assarel in Bulgaria, - Bucim in Macedonia.

Other domains have porphyry-type mineralization, such as the Fennoscandian shield (Aitik, Tallberg, Kopsa), the Caledonian belt (Coed y Brenin in Wales, Black Stockarton Moor in Scotland), and the upper Paleozoic Hercynian magmatism (Beauvin, Sibert, Auxelles-le-Haut in France)



ProMine WP1 Mineral predictive mapping

Overview: The goal of predictive mapping is to identify areas where unknown mineral raw materials could favourably be discovered. Methodologies used for the calculation of mineral predictive maps vary depending on the fact that the studied commodity is a main element in the deposit or a by-product. Up to now, a very large majority of predictive studies dealt with a relatively low number of elements which are the main commodities in their deposits and/or which belong to the main paragenesis (e.g. Cu or Au in porphyries). However, ProMine project bears a great attention to strategic and critical commodities, and especially to the 14 critical raw materials identified by the European Commission. In fact, these critical commodities may either be - the main commodities with a proper mineral expression in mined ores (e.g. W, Sb, Fl or Sn), - or commodities which are by-products from mining (e.g. Ge, Ga, In or Ta). As a consequence, predictive methods needed to be adapted to consider both possibilities. In the present work, we used, for the first case, a geographic prediction method (Weight of Evidence, or WofE) and, for the second case, a database querying method.


ProMine WP1 Mineral predictive mapping, Weight of Evidence – methodology

principles: The WofE method is a probability-based approach that uses Bayes’ rule to combine evidence with an assumption of conditional independence. Where sufficient data are available, it can be applied to estimate the relative importance of evidence by statistical means. For details of the method, see Bonham-Carter (1994) and Kemp et al. (2001). The spatial analysis was performed on the “deposits” (training points) and “geology” (evidential theme) layers, using the Arc-SDM (Spatial Data Modeller) extension, developed for ArcView 3.x / Spatial Analyst (ESRI®). The same process was used for all targeted commodities. Wh was calculated separately for deposits (i.e. class A, B, C, D, E) and showings (i.e. class N/A) in the database. Results were added as follow: (Wh deposits) + 0.5 (Wh showings)



ProMine WP1

One predictive map was calculated for each of the 6 studied commodities. Deposits containing the targeted commodity are also placed on the maps to easily identify “new” favourable areas.

Copper Fluorite

Antimony Tin

Tungsten Zinc in carbonate hosted deposits

Mineral predictive mapping, Weight of Evidence – results

ProMine WP1

Antimony (Sb), table of results: 342 records from the ProMine MD database contain Sb: - 109 deposits, - 233 showings,

Dominant deposit types are: - Base metals veins (55,05%) - Epithermals (17,8%) - Orogenic gold (8,78%) - Carbonate hosted (<5%) - VMS (<5%)

Dominant associated commodities are Au and As with, in smaller proportion (Pb, Zn, Cu, Ag) and Hg in epithermal parageneses.

Geological Code Comment

WofE (deposits)

WofE (showings)

WofE (combined) code

mp_vi Mioc 1,80 2,14 2,87 25

m_va Mioc 1,33 0,57 1,61 24

Oph-jc Mesoz. 1,88 1,13 2,45 23

t_vi Mesoz. 4,87 0,00 4,87 22

t2-3 Mesoz. 0,75 1,38 1,44 21

Plt-h2 Psup_Mag 1,86 1,79 2,76 20

Plt-d3 Psup_Mag 2,53 3,16 4,11 18

PltGn-ko2 Pinf_Mag 2,67 2,83 4,09 17

PltGn-b2k Pinf_Mag 3,79 1,24 4,41 16

hr Psup 1,34 1,27 1,97 15

h3 Psup 2,56 2,50 3,81 14

d3_vb Psup 4,03 3,27 5,66 13

h2_va Psup 3,25 3,18 4,84 12

Gla-ko Pinf 2,63 3,25 4,25 11

s1 Pinf 1,74 0,00 1,74 10

s Psup 1,04 1,38 1,73 9

os Pinf 0,69 1,72 1,55 8

o1-2 Pinf 1,18 1,52 1,94 7

ko Pinf 3,97 3,71 5,82 6

b2k Pinf 2,31 2,24 3,43 5

bk Pinf 2,53 2,36 3,71 4

bo Pinf 3,11 2,70 4,47 3

kr P 1,68 0,22 1,79 2




ProMine WP1

Antimony (Sb), predictive map: Most favourable area is the French Hercynian domain (Massif Central and Brittany) and correspond to late orogenic veins (~300 Ma) on Paleozoic basement. Another favourable area is the Balkan-Carpathian region, in relation to Mezosoic and Cenozoic series. Predictive results, however, allow to extend Sb-favourability zones to other areas of the Paleozoic domain (e.g. Wales, the Alps, eastern Bohemian massif,…) and the Balkan-Carpathian domain.


ProMine WP1

Principles: The objective of this approach is to assess favourability for rare metals commodities (i.e. by-products of other commodities) in deposits of the MD ProMine database. To achieve this objective, deposits of the database are ranked according to their similarity to deposits containing the targeted commodity. • Step 1: frequency of the targeted commodity is evaluated per deposit type and then compared to

the whole database, in order to calculate an “enrichment ratio” (ER) per deposit type;

• Step 2: In all deposits that contain the targeted commodity, for each favourable deposit type, a list of associated commodities and their frequency is calculated. The result is a table of characteristic polymetallic association, or “signature” favouring the presence of the targeted commodity. This characteristic polymetallic signature will be search in all deposits of favourable types.

• Step 3: deposits are ranked relatively to their degree of similarity with the polymetallic signatures. The ranking is the sum, for all commodities in the polymetallic association of the product of a Boolean value (i.e. “commodity is present” = 1 and “commodity is not present” = 0) and the frequency of the commodity in the signature.

• Step 4: to give a more global and synthetic view of results, ranks of deposits are weighted with the ER of the type they belong to.

Mineral predictive mapping, Database Querying – methodology



ProMine WP1

One predictive map was calculated for each of the 5 studied commodities. Results are mapped as density. Deposits containing the targeted commodity are also placed on the maps to easily identify “new” favourable areas.

Cobalt Gallium

Germanium

Indium

Tantalum

Mineral predictive mapping, Database Querying – results

ProMine WP1

Germanium (Ge), enriched deposit types: 111 deposits in the ProMine MD database contain germanium. Enrichment ratios show that germanium is preferably present in: - Carbonate hosted deposits

(ER=3,35)

- VMS (ER=1,78)

- Epithermals (ER=1,50)

- Base metals veins (ER=1,14)

These 4 types contains 79% of Ge-bearing deposits.

Metallogenic Type N

Enrichment ratio

Base metals veins 31 1,14 Carbonate-Hosted 28 3,35 Epithermal 10 1,50 VMS 19 1,78 Igneous Felsic 8 0,69 Igneous Intermediate 1 0,59 Igneous Replacement 5 0,98 Orogenic Gold 1 0,20 Residual deposits 1 0,13 SandStone- and Shale-Hosted 5 1,09 Sedimentary deposits 2 0,24



ProMine WP1

Germanium (Ge), polymetallic signatures: The frequency of commodities associated with germanium is calculated for each enriched type. The resulting table shows, per deposit type, the commodities that are preferentially associated with germanium, or “characteristic polymetallic associations”. For instance, zinc that is the most frequent element is in 87% of base metals veins deposits that contain germanium, while Pb and Ag are in 74% and 68% (respectively) of these deposits.

Commodity Global %

Base metals veins

Carbonate- Hosted

Epithermal VMS

Zn 74 87 79 40 89 Pb 68 74 71 30 68 Ag 58 68 43 70 37 Cd 40 48 32 20 32 Cu 33 19 25 80 26 Ga 25 32 18 10 37 In 17 16 7 30 16 Au 15 3 4 70 11 Ba 12 13 14 20 11 As 11 13 11 30 0 Sb 10 6 7 30 5 Bi 9 6 4 10 5 Fe 8 0 7 10 16 Sn 7 10 0 0 5 Tl 7 3 4 10 0 Ni 6 3 7 10 5 Te 6 0 4 20 0 Co 5 3 4 0 11 U 5 6 0 0 0 F 4 10 4 0 0

Hg 4 0 7 20 0 Se 4 0 0 10 5



ProMine WP1

Ranking calculation for germanium: Rank 1 (base metals veins): ((0.87 * [Comm.Zn]) + (0.74 * [Comm.Pb]) + (0.68 * [Comm.Ag]) + (0.48 * [Comm.Cd]) + (0.19 * [Comm.Cu]) + (0.32 * [Comm.Ga]) + (0.16 * [Comm.In_]) + (0.03 * [Comm.Au]) + (0.13 * [Comm.Ba]) + (0.13 * [Comm.As_]) + (0.06 * [Comm.Sb]) + (0.06 * [Comm.Bi])) Rank 2 (carbonate-hosted): ((0.79 * [Comm.Zn]) + (0.71 * [Comm.Pb]) + (0.43 * [Comm.Ag]) + (0.32 * [Comm.Cd]) + (0.25 * [Comm.Cu]) + (0.18 * [Comm.Ga]) + (0.07 * [Comm.In_]) + (0.04 * [Comm.Au]) + (0.14 * [Comm.Ba]) + (0.11 * [Comm.As_]) + (0.07 * [Comm.Sb]) + (0.04 * [Comm.Bi])) Rank 3 (epithermal): ((0.40 * [Comm.Zn]) + (0.30 * [Comm.Pb]) + (0.70 * [Comm.Ag]) + (0.20 * [Comm.Cd]) + (0.80 * [Comm.Cu]) + (0.10 * [Comm.Ga]) + (0.30 * [Comm.In_]) + (0.7 * [Comm.Au]) + (0.2 * [Comm.Ba]) + (0.3 * [Comm.As_]) + (0.3 * [Comm.Sb]) + (0.1 * [Comm.Bi])) Rank 4 (VMS): ((0.89 * [Comm.Zn]) + (0.68 * [Comm.Pb]) + (0.37 * [Comm.Ag]) + (0.32 * [Comm.Cd]) + (0.26 * [Comm.Cu]) + (0.37 * [Comm.Ga]) + (0.16 * [Comm.In_]) + (0.11 * [Comm.Au]) + (0.11 * [Comm.Ba]) + (0.05 * [Comm.Sb]) + (0.05 * [Comm.Bi]))

Germanium (Ge), ranks calculation: For each deposit in each favorable type, a rank is calculated, in order to measure its degree of similarity with deposits of the same type that contain the targeted commodity. Calculation is based on the polymetallic signature of the type. Ranks are then weighted with the ER of the type the deposits belong to.

Global ranking for germanium = (1.14 [RANK1]) + (3.35 [RANK2]) + (1.58 [RANK3]) + (1.78 [RANK4])




ProMine WP1

Germanium (Ge), predictive map: The resulting map highlights a germanium province in southern Europe with various types of mineralization emplaced in lower Paleozoic, Mesozoic (in relation to carbonates of the Tethyan margin), and upper Cretaceous-Cenozoic (porphyries and epithermal deposits). Also, some areas, such as massive sulfides domains, show relatively high favourability, and should be investigated in more details by future studies.


ProMine WP1 Conclusion

• 16 maps of mineral potential have been produced, for the 16 main commodity associations (or deposit types)

• 6 predictive maps have been produced for 6 main elements (W, Sn, Sb, Fl, carbonate hosted Zn and Cu), using the Weight of Evidence method

• 5 predictive maps have been produced for 5 by-products elements (Ge, Ga, In, Ta and Co), using a newly developped database querying method

• These 27 maps and their comments are available on the ProMine web portal.

• The ProMine databases will provide data for numerous future researches on mineral resources in Europe!


ProMine WP1

Thank you for your attention!


promine project – wp1promine.gtk.fi/documents_news/promine_final_conference/10_10_be… ·...

Documents