First results regarding the influence of mineralogy on the mechanical properties of seafloor massive sulfide samples

Engineering Geology 214 (2016) 127–135

Contents lists available at ScienceDirect

Engineering Geology journal homepage: www.elsevier.com/locate/enggeo

First results regarding the influence of mineralogy on the mechanical properties of seafloor massive sulfide samples Giovanni Spagnoli a,⁎, Andreas Jahn b, Peter Halbach b a b

Department of Maritime Technologies, BAUER Maschinen GmbH, BAUER-Str. 1, 86529 Schrobenhausen, Germany Institute of Geological Sciences, Freie Universität Berlin, Malteserstrasse 74-100, 12249 Berlin, Germany

a r t i c l e

i n f o

Article history: Received 10 August 2016 Received in revised form 14 October 2016 Accepted 17 October 2016 Available online 18 October 2016 Keywords: SMS deposits Unconfined compressive strength Brazilian test Porosity Vickers tests Mineralogical analysis

a b s t r a c t Seafloor Massive Sulfides (SMS) are increasingly accepted as important marine raw material resources for the future, in particular because of their polymetallic character. Many industrial nations are researching not only on the scientific importance of these deposits but also on their economic value. Regarding the current state of the international exploration, mainly two-dimensional surface-close observations are accessible, whereas only some core data and measurements are publicly available to date. In fact, there are few successful drilling campaigns containing information about the size and shape, which also give information on the structure and content of modern massive sulfide ore bodies in their third dimension. Regarding the mechanical properties of SMS samples, only few data are available. Geotechnical data of these deposits are important in order to develop an efficient mining technology for exploitation. In this research, 12 SMS samples from two different locations were investigated. Based on the mineralogical characterization of the studied samples the geotechnical properties were preliminary correlated with the mineralogical results. The comparative study indicates how far the geotechnical data are controlled by mineral type and composition, including the porosity. A regression relation between compressive strength and the porosity based on the mineralogy shows a distinct relation between these parameters. Therefore, the geomechanical and mineralogical features have a strong importance for deep-sea mining applications and this should be kept in mind considering the hyperbaric effects on the rock cutting. © 2016 Elsevier B.V. All rights reserved.

1. Introduction In the deep-sea environment, hydrothermal fields are preferentially related to extensional tectonics. These fields are parts of a global energy dissipation system in which mantle heat is transported to the seafloor by a combination of conductive and convective flux; the latter mainly represents the hydrothermal circulation. The preferred locations for the convective energy transport and the related mineral formation processes are the globe-encircling mid-ocean ridges as well as the spreading zones of back-arc basins behind volcanic island arcs. Typical water depths of the modern hydrothermal fields are 1500–4000 m. Seafloor Massive Sulfide (SMS) deposits are considered to be a very important future source for metals such as Cu, Zn, Pb, Au, Ag, Se, Cd and other high-tech elements. However, the necessary techniques and systematic details for mining and processing these minerals have not been tested yet and are still in the development stage (Nautilus Minerals, 2016). For the process of marine excavation the geotechnical properties of the respective sulfidic ores are extremely important, in particular ⁎ Corresponding author. E-mail addresses: [email protected], [email protected] (G. Spagnoli).

http://dx.doi.org/10.1016/j.enggeo.2016.10.007 0013-7952/© 2016 Elsevier B.V. All rights reserved.

since these deposits are located in water depths with remarkable hydrostatic pressures (Spagnoli et al., 2015). More than 100 sites exhibiting more or less recent activity, emanating mineral-bearing hot fluids at temperatures of up to 410 °C, and containing metal-rich solutions, are now known from a variety of tectonic settings, particularly in the Pacific Ocean and its marginal back-arc seas, but also in the Atlantic and Indian Oceans as well as in the Mediterranean Sea. Larger accumulations of polymetallic sulfides are known from about 300 sites (Hannington et al., 2011). They contain groups of chimneys, fragmented chimneys, and hydrothermal mounds that are up to several hundred meters in diameter and tens of meters high within vent fields that can be traced for kilometers. Commonly, the larger occurrences consist of several generations of chimneys, with the older portions generally fragmented, forming mound deposits that are subsequently topped by the youngest (active) chimney generation. The evolved hydrothermal fluids reach the seafloor where, due to rapid cooling by mixing with cold seawater, most of their mineral content precipitates in the form of chimney structures and stockwork mineralizations in the permeable subsurface. Hot fluid-mineral suspensions emanate as black and white smoke and enter into the near-bottom water layer, since not all the mineral content precipitates immediately on and below the seafloor. The chimney structures are subject to

128

G. Spagnoli et al. / Engineering Geology 214 (2016) 127–135

cooling, alteration and fragmentation and become eventually incorporated into mound deposits. These mound deposits are primarily the sites for seafloor excavation. This research provides new information on the influence of the mineralogy on the geotechnical characteristics of the SMS samples. This is an important part of the project development for the mining systems (e.g. Spagnoli et al., 2015; Spagnoli et al., 2016). Many of the results available in the literature regarding deep-sea mining applications are based on tests performed on non-SMS materials (e.g. Waquet et al., 2011; Alvarez Grima et al., 2015). Therefore, the data introduced in this research give a first indication about the geotechnical properties with regard to the mineralogical characteristics of SMS samples. However, it has to be pointed out that further detailed tests are necessary for designing specific mining systems for the marine deep-sea environment.

through by reactivated hot solutions and becomes cemented by epigenetic sulfide precipitates which also contribute to the seafloor massive sulfide deposit. Depending on the respective maximum fluid temperature in this process, the massive ore body can be dominated by either Fe- and Cu- sulfides or Zn- sulfides. At the surface contact between the mound and cool surrounding seawater, predominantly amorphous silica (jasper) is precipitated (Halbach et al., 2003), but also Fe-free sphalerite, barite and anhydrite in larger cracks and vugs. The caprock often forms a covering layer above the subjacent mound sulfide material. From the theoretical point of view, the black smokers are interesting because of the Cu, Au and Ag, whereas the white smokers are for the Zn. However, it can also be that black smokers have larger amounts of pyrite, which is economically irrelevant and also very hard and abrasive. Therefore, for each exploratory drilling the mineralogy plays an important role in order to assess the ore content.

2. Materials and methods

2.2. Geotechnical tests

2.1. Geological background

In the present study, geotechnical properties were measured for 22 cylindrical samples cut from 12 sulfide samples of known origin. Unconfined compressive strength (UCS), Brazilian test strength (BTS), mineral density, porosity, bulk density and Vickers tests were performed. With these geotechnical data, it was possible to provide preliminary information regarding future deep-sea mining cutting projects. UCS tests were performed according to the recommendations of Mutschler (2004) on 10 test cylinders (Table 1). The cylindrical wet core samples (height 100 mm and diameter 50 mm) were drilled and cut with a diamond saw to cylinders with a length-to-diameter ratio of about 2:1. The load was strain controlled and applied with a constant deformation rate of 0.05%/min (i.e. 0.1 mm/min) until no residual strength was reached. BTS tests, an indirect tensile strength test, were performed according to the Brazilian tests on 12 cylinder (Table 1) samples according to the recommendations of Lepique (2008). The disc shape specimens of the rock were loaded by two opposing normal strip loads at the disc periphery. The specimen diameter was 50 mm with a thickness/diameter ratio ranging between 0.5 and 0.6. The load was continuously increased at a constant rate until failure of the sample occurred after a few minutes. The loading rate depending on the material, varied between 10 and 50 kN/min. At the failure, the tensile strength of the rock is calculated as follows:

The samples tested for these geotechnical experiments originate mainly from the North Fiji Basin (Halbach et al., 1996, 1998) and the MESO-site in the Central Indian Ocean (Halbach et al., 2002). The samples were collected with a grab sampler. The sample set comprises black smoker, white smoker and caprock mineralizations. The black smoker chimneys mostly consist of copper minerals (chalcopyrite and subordinately covellite) as well as recrystallized iron sulfides (pyrite and marcasite) and relics of anhydrite (see sample in Fig. 1A). The corresponding chalcopyrite-concentrations vary between 5 and 18%. The Fe-rich “kies-type” sulfides typically form the outer part of the chimney structures; these sulfides are dominated by pyrite and marcasite with one sulfide often replacing the other. All these features indicate that the black smoker chimneys represent high-temperature mineralizations and were formed from emanating vent fluids with temperatures of up to 410 °C. “White smoker” chimneys are characterized by particularly high sphalerite concentrations (between 19 and 52%) and a great variability with respect to interstitial amorphous silica content. Fepoor sphalerite is the main sulfide phase in the white smoker mineral assemblage. The chimneys begin to grow at temperatures of around 300 °C or less, and get their colour from greyish white particles (amorphous silica, some barite and anhydrite) suspended in the meso-thermal fluids. White smoker chimneys form mainly at a later stage in high-temperature systems (see sample in Fig. 1B). Often, white and black smoker chimneys occur in the same large vent complex indicating that they are supplied from the same high-temperature plumbing system at depth. The succession of chimney formation and their subsequent erosion causes the build-up of a debris mound at the hydrothermal vent site. This porous debris material is also flown

BTS ¼

2∙F π∙d∙l

ð1Þ

where BTS is the tensile strength in MPa, F is the applied load, d is the specimen diameter and l is the specimen length. Compression-induced extensional fracturing generated in this test is also more representative of the in situ loading and failure of rocks (Aydin and Basu, 2006). The

Fig. 1. A: Original sample for the specimen defined as 07 (see Table 3); B: Original sample for the specimen defined as 10 (see Table 3).

G. Spagnoli et al. / Engineering Geology 214 (2016) 127–135

129

Table 1 Results of the UCS tests performed according to Mutschler (2004). Specimen Diameter [mm]

Height [mm]

Mass [g]

Density [g/cm3]

UCS [MPa]

Elastic modulus [MPa]

Poisson's ratio

Mineral density [g/cm3]

Porosity [%]

SMS type

01 03_A

49.2 49.2

102.0 100.8

481 485

2.5 2.5

30.4 36.6

6650 10,554

0.04 0.06

3.20 3.14

22.6 19.4

04_A

49.1

101.5

495

2.6

10.1

4238

0.03

3.25

20.8

05

49.2

101.2

369

1.9

5.3

2654

0.06

3.14

38.8

06_B

49.1

100.7

482

2.5

17.8

5943

0.06

3.14

19.6

07_A 08_B 10_B

49.2 49.1 49.2

100.4 101.7 99.7

578 453 401

3.0 2.4 2.1

41.4 8.5 13.3

11,100 1118 4942

0.17 0.51 0.10

3.68 3.34 2.96

17.5 29.6 28.5

13_A 14_A

49.1 49.1

103.2 103.6

573 415

2.9 2.1

17.0 12.2

4815 4324

0.04 0.07

4.04 2.95

27.5 22.3

Black smoker White smoker White smoker White smoker White smoker Black smoker Black smoker White smoker Black smoker Caprock

mineral densities were obtained according to the DIN 66137-2. This method is based on the displacement of gas (helium) from the sample. Only the solid volume VF is displaced, whereas open pore volume VP and interstitial volume VS can be freely filled by the gas. The porosity and bulk density values were obtained according to the DIN EN 1936. For mineralogical analyses, 13 standard polished sections were prepared from the 10 UCS test cylinders after the measurement. Vickers tests were performed on the 13 polished sections according to DIN EN 6507-1 by using a Reichert-Jung MicroDuromat 4000. The Vickers hardness test method consists of indenting the test material with a diamond indenter, in the form of a right pyramid with a square base and an angle of 136° between opposite faces subjected to a load of 1 to 100 kgf. The full load is normally applied for 10 to 15 s. The two diagonals of the indentation left in the surface of the material after removal of the load are measured using a microscope and their average was calculated. The area of the sloping surface of the indentation is determined. The Vickers hardness is the quotient obtained by dividing the kgf load by the square mm area of indentation: HV ¼ 1:854∙

F

ð2Þ

2

d

where F is the applied load in kgf and d is the arithmetic mean of the two diagonals. 2.3. Mineralogical tests In order to characterize the relation between the geotechnical properties and the mineralogy of the ore samples, a petrographical survey of 13 polished sections was conducted. The round polished sections have standard dimensions (30 mm diameter, 5 mm thickness) and were prepared from the residues of the UCS test cylinders after testing (see

above). For the analyses a petrologic microscope with polarizer (Carl Zeiss Axioskop 40) with a digital camera (Carl Zeiss AxioCam MRc5) was used. Within the scope of the petrographic investigation, the quantitative mineral composition was determined by using the freeware image interpretation software ImageJ. While studying the polished sections with the petrologic microscope, the investigation was focused on the detailed description of the dominant ore mineral phases, their intergrowth pattern, the growth textures of individual minerals or intergrown minerals as well as the portion and size distribution of the pore spaces. All these data were collected for the three different ore types of black smokers, white smokers and caprock. Subsequently, the results of this survey were correlated with the geotechnical data and interpreted. 3. Results and discussion 3.1. Geotechnical interpretation Not many results about the geotechnical properties of massive sulfides are in the public domain (Spagnoli et al., 2014). Some geotechnical data are available for massive sulfide deposits on land, which however represent more mature sulfide ores with higher resistance (e.g. Andrieux and Simser, 2001). Yamazaki et al. (1990) and subsequently Yamazaki and Park (2003) performed several geotechnical tests on eight SMS samples coming from the Okinawa Trough. The authors state that porosity and Cu + Pb + Zn + Fe content influence the mechanical properties of those samples. The results of the mechanical tests from this study are shown in Table 1 and Table 2. Unfortunately, for the specimens 09, 11 and 12 it was not possible to determine the porosity and the mineral density, as during the preparation the samples collapsed. From the results shown in Table 1 and Table 2 the m value, defined by Hoek and Brown

Table 2 Results of the BTS tests performed according to Lepique (2008). Specimen

Diameter [mm]

Height [mm]

Mass [g]

Density [g/cm3]

BTS [MPa]

Mineral density [g/cm3]

Porosity [%]

SMS type

01 04 05 06 07 08 09 10 11 12 13 14

99.9 99.0 99.2 98.8 99.5 99.1 98.6 99.2 99.2 99.8 99.2 99.1

50.7 50.5 50.7 49.8 50.2 50.1 50.0 50.1 50.6 50.2 50.0 50.3

976 895 908 822 1034 954 1018 967 1161 798 558 439

2.5 2.3 2.3 2.2 2.6 2.5 2.7 2.5 3.0 2.0 2.5 2.3

3.6 1.4 1.7 1.3 3.6 1.2 1.3 6.0 3.3 1.1 1.0 0.9

3.20 3.27 3.14 3.21 3.75 3.38 n.a. 3.14 n.a. n.a. 4.19 3.10

22.6 22.6 38.8 21.3 16.6 26.6 n.a. 27.7 n.a. n.a. 34.1 23.8

Black smoker White smoker White smoker White smoker Black smoker Black smoker Black smoker White smoker Black smoker Black smoker Black smoker Caprock

130

G. Spagnoli et al. / Engineering Geology 214 (2016) 127–135

Fig. 2. Comparison of the bulk density values of our study, Yamazaki et al. (1990) and Yamazaki and Park (2003).

Fig. 3. Elastic-modulus vs UCS values for our study, Tomishima et al. (1989), Yamazaki et al. (1990) and Yamazaki and Park (2003).

(1980) as the ratio of UCS to BTS, ranges from 3.14 (for specimen 05) to 16.5 (for specimen 13). These values agree with the data presented by Spagnoli et al. (2015). The correlation of the UCS and BTS vs. the bulk density is shown in Fig. 2 compared with the values of Yamazaki et al. (1990) and Yamazaki and Park (2003). It is possible to observe that

the UCS values increase with higher values of bulks density, whereas this trend is not marked with BTS tests. The values shown by Yamazaki et al. (1990) and Yamazaki and Park (2003) have a similar path. Fig. 3 shows the relation between UCS and elastic modulus. A similar comparison has been also done for the data furnished by Tomishima et al. (1989) (for the Kuroko ore deposit), Yamazaki et al. (1990) and subsequently Yamazaki and Park (2003). The path is very similar and shows a good relation between UCS and elastic modulus, however, because of the small number of specimens tested for this research, the correlation E-modulus = f(UCS) has not been explicitly modelled. The data from Tomishima et al. (1989) are by far, the most different. Very high values of UCS and elastic-modulus are reported. Unfortunately, no other information about the maturity, metal content, type of minerals is provided. However, because of the lack of public data on the geotechnical properties of SMS, it is worth showing all available information. Fig. 4 shows the influence of the porosity on the UCS and BTS values. For this diagram a distinction between black smoker, white smoker and caprock was done. Regarding the black smoker results, porosity decreases with increasing UCS values. This is due to the maturation stage of the SMS deposits. According to Waquet and Fouquet (2011) porosity decreases along with maturation, the closer the sulfides samples are to the surface, the less mature and the more porous they are. Normally, this behavior is commonly observed for sedimentary rocks as UCS

Fig. 4. Influence of porosity on the UCS (left) and BTS (right). The UCS of all massive sulfide samples shows a strong decrease with increasing porosity. No correlation is observed between the tensile strength and the porosity, and the BTS values are distinctly lower than the UCS values.

G. Spagnoli et al. / Engineering Geology 214 (2016) 127–135

131

therefore it is not reasonable to propose a detailed regression analysis. The influence of the mineralogy seems to play an important role on the physical properties; therefore the results are also evaluated by the mineralogical point of view. 3.2. Mineralogical interpretations

Fig. 5. Relation UCS vs BTS for the SMS samples compared with Yamazaki et al. (1990) and Yamazaki and Park (2003). A proper regression function cannot be derived.

increases with age due to increased lithification and reduced porosity (e.g. Chang et al., 2006; Waltham, 2009). Regarding the BTS, the correlation is not as clear as for the UCS, although a trend exists. This agrees very well with the findings of Palchik and Hatzor (2004), who investigated the relation BTS with porosity in chalks. Unfortunately, not many data on UCS and BTS are available for SMS samples, therefore an indirect comparison is only possible with other rock types. It has to be pointed out that the porosity values for the same sample number shown in Table 1 and Table 2 are not identical. In fact, we measured the porosity for each UCS sample, whereas the porosity associated with the BTS data was not measured, as it originates from another part of the sulfide block. UCS and BTS values are very important geotechnical parameters because they are linked together by the m value: igneous rocks have generally high m values, whereas sedimentary clayey rocks have low m values (Miedema, 2014). Regarding the UCS and BTS relation (Fig. 5), we also observed a weak trend. With respect to our data, the only point aside from the correlation is the sample 10, which has high porosity (28.5%) and an average grain size (108 μm). The data compared with the results of Yamazaki et al. (1990) and Yamazaki and Park (2003) create a preliminary database. Several studies have been conducted to assess a correlation of UCS to BTS on rock samples (e.g. Farah, 2011; Kahraman et al., 2012); however there are not many data available on SMS samples up to date, and

The results of the geotechnical properties of the SMS samples will need some more explanation in order to understand their behavior. Therefore, an analysis from the mineralogical point of view is performed to define the respective mineral phases and their intergrowth in relation to the porosity. The basic mineral phases in the studied SMS samples are pyrite, marcasite, melnikovite, chalcopyrite and sphalerite. The respective chemical formula of the composition is shown in Table 4. The gangue minerals are amorphous silica and subordinately anhydrite and barite. All of these minerals except the latter two are quantitatively documented in the table of the mineral composition (Table 3). These minerals are also identified by the ore-microscopic investigations of polished sections and are presented in the respective photomicrographs (Fig. 6). Regarding the relative hardness of the sulfide minerals and amorphous silica significant differences can be observed, which is also documented by the Mohs scale values (Table 4). The hardest mineral phase is pyrite, which in our set of measurements has an average Vickers Hardness Number (HV units) of about 1750 which is somewhat higher than the respective value shown in the literature (1546 HV units; Anthony et al., 2003). The second hardest mineral is marcasite, which is also a Fe sulfide compound with a measured HV unit of 1569; also this value is higher than the respective literature number. The melnikovite phase is a collomorph mixture of extremely fine-grained pyrite and marcasite and shows correspondingly a HV value between pyrite and marcasite. The reason for this intimate intergrowth is the high rate of the fast precipitation (Halbach et al., 2003). Chalcopyrite and sphalerite represent very important minerals with regard to the economic value of the ore (Hannington et al., 2011). However, with regard to the relative hardness they are significantly softer and have similar hardness values: chalcopyrite 232 HV and sphalerite 239 HV. The respective hardness numbers in the literature (Anthony et al., 2003) are somewhat smaller. With regard to the mineral composition, it has to be noted that in all samples, mixtures of different minerals with varying hardnesses are on hand. In the investigated sample set a distinction was made between black smoker, white smoker and caprock composition. Based on the polished sections, the mineral composition of the different ore types using the software package ImageJ for quantitative image analysis was determined. These results show remarkable differences with regard to the

Table 3 Quantitative mineral composition divided by black smoker, white smoker and caprock ore types. Sample

Quantitative mineral composition [vol.%] Chalcopyrite

Pyrite

Marcasite

Melnikovite

Sphalerite

SiO2

Porosity [%]

Average grain size [μm]

Black smoker 01 07_A 07_B 08_B 13_A 13_B

14.1 10.0 4.9 16.2 0.0 18.3

5.8 5.8 34.6 7.2 3.1 2.9

3.0 1.1 6.1 3.3 29.3 0.0

7.5 17.2 18.8 7.9 1.2 4.6

0.0 2.2 0.0 2.2 1.3 3.2

47.0 46.3 18.1 33.7 37.6 43.5

22.6 17.5 17.5 29.6 27.5 27.5

98 80 63 95 85 113

White smoker 03_A 04_A 05 06_B 10_B

1.1 0.9 0.6 2.5 1.1

0.6 2.1 0.4 2.8 0.9

0.0 1.5 0.0 0.4 0.8

0.0 2.0 0.0 0.1 0.3

45.1 39.3 19.3 52.1 34.1

33.8 33.4 40.9 22.6 34.3

19.4 20.8 38.8 19.6 28.5

95 100 93 193 108

Caprock 14_A 14_B

0.7 0.0

2.0 1.2

0.5 0.3

0.2 0.1

21.3 0.5

52.9 75.6

22.3 22.3

127 n.d.

132

G. Spagnoli et al. / Engineering Geology 214 (2016) 127–135

Fig. 6. Microphotographs of polished sections of the studied massive sulfide samples showing different mineral assemblages. All pictures shown in Fig. 6 are obtained with parallel polarizers. Fig. 6A and 6B show a typical black smoker mineral association which is dominated by melnikovite and pyrite; 6C is a chalcopyrite-rich black smoker sample, 6D shows large xenomorphic chalcopyrite grains; 6E, 6F and 6G are typical white smoker mineral associations dominated by sphalerite with 6F showing the fine-grained massive sphalerite dominance; 6H shows the caprock mineral assemblage, where sphalerite is the dominant sulfide mineral.

relative composition of the above mentioned ore types (Table 3). In the black smoker substance the mineral sphalerite plays only a subordinate role (Fig. 6A – D). The leading constituents are pyrite, marcasite, chalcopyrite and melnikovite. In samples with high Cu content (sample 13_B, Table 3) the chalcopyrite can be highly enriched within distinct zones (Fig. 6C); mostly the chalcopyrite grains are hypidiomorphic to idiomorphic. The main gangue phase is amorphous silica, which in all black smoker samples is at least to more than 18 vol.%. As a minor constituent, anhydrite was observed. The highest relative portion of pyrite amounts to about 35 vol.% in sample 07_B (Fig. 6A, Table 3). The maximum value of marcasite is about 29 vol.%, whereas melnikovite, which typically represents a late-stage precipitate at lower temperatures, is

only about 19 vol.%. Chalcopyrite is typically a high-temperature precipitate located in the central parts of the chimney structures (Halbach et al., 2003) and shows relative portions in the studied samples of up to 18 vol.% (Table 3). The average grain sizes of the ore minerals based on the microscopic investigations on the polished sections vary in the black smoker samples between 63 and 113 μm while individual grains of chalcopyrite can reach up to 350 μm; often the coarser chalcopyrite grains are grown as idiomorphic aggregates. The two Fe sulfide minerals pyrite and marcasite are often intimately intergrown (Fig. 6A and B). Besides the supporting mineral framework, the porosity and their degree of interconnection in the samples has a decisive influence on the strength and stability of the mineral substance. As will be shown later,

G. Spagnoli et al. / Engineering Geology 214 (2016) 127–135 Table 4 Measured Vickers Hardness Numbers of the main constituents of the SMS samples compared with values from Anthony et al. (2003) and the respective hardness values of the Mohs scale. Vickers Hardness Number (HV)

Chalcopyrite Pyrite Marcasite Melnikovite Sphalerite SiO2, amorph. a

Average Min

Max

Anthony et al. (2003)

Mohs scale

Chemical formula

232 1748 1442 1569 239 621

262 1974 1662 1890 289 717

206 1546 1006 n.d. 216 550

4.0 6.0–6.5 6.0–6.5 n.d. 3.5–4.0 5.5–6.0

CuFeS2 FeS2 FeS2 a FeS2 Zn(Fe)S SiO2

210 1471 1200 1362 210 506

Collomorph mixture of marcasite and pyrite.

the porosity is more or less inversely related to the UCS of the individual samples. The measured porosity is shown in Table 3 and varies in the group of the black smoker samples between 17.5% and 29.6%. The white smoker samples represent a sulfide material which has been precipitated at lower temperatures, in general at temperatures underneath 300 °C (Halbach et al., 2003). Therefore, the dominating mineral phase of this kind of sulfide assemblage is sphalerite (Fig. 6E-G); within the sphalerite lattice the metal Zn can be partially replaced by Fe. Also the minerals pyrite, marcasite and chalcopyrite can be observed in the mineral composition but represent only minor fractions. The sphalerite content varies between 19 vol.% and 52 vol.%. The porosity has also been determined and varies between 19% and 39%, i.e. the sample with the highest porosity value in our set (38.8%) is a white smoker ore type. The average grain sizes of the minerals in the white smoker assemblage varies between 93 μm and 193 μm while individual grains of well-crystallized sphalerite can reach maximum diameters of 500 μm. The white smoker sample 05 has the lowest compressive strength value (Table 1) combined with a very high porosity (38.8 vol.%; Fig. 7). The microscopic study of this sample also shows that the pore spaces are characterized by a high degree of interconnection. The caprock material (sample 14_A, Fig. 6H) represents a late stage mineralization of the hydrothermal ore precipitation and forms at temperatures below 200 °C (Halbach et al., 2003). The main constituent in the considered sample is amorphous silica with contents between 52.9 vol.% and 75.6 vol.%; a large portion of this collomorph silica is impregnated by Fe-oxyhydroxide and can be described as jasper. In parts

Fig. 7. Plot of the UCS values (MPa) vs. measured porosity of the investigated SMS samples. The 6 samples along the dashed line represent the linear relationship (R2 = 0.99) and consist of 4 black smoker and 2 white smoker samples. The solid line represents the exponential regression of the interrelationship of all 10 studied samples (R2 = 0.61). The mineral composition of the respective samples is shown in Table 3.

133

the caprock can contain high local concentrations of ZnS indicated by a 21.3 vol.% content of sphalerite (Fig. 6H). The grain sizes vary between 100 μm and 200 μm. Fig. 6A shows the sample 07_B; the typical black smoker mineral association is dominated by melnikovite (mel) and pyrite (pyr). The finegrained Fe-sulfide ore material is densely intergrown and combined with a low irregular porosity. The melnikovite shows typical dendritic growth textures; in the central part the sulfide aggregates are compact and massive. Some larger crystals of chalcopyrite (cpy) and pyrite are distributed in the interstices. This sample has the highest compressive strength value of all samples (41.4 MPa). Fig. 6B shows the sample 07_A: a typical black smoker mineral association dominated by melnikovite and pyrite. In this case the melnikovite shows a microporosity, in the marginal zones of the Fe-sulfides, rims of pyrite crystals can be observed. One pyrite crystal is covered by chalcopyrite. Fig. 6C shows the sample 13_B; a chalcopyrite-rich black smoker sample. The hypidiomorphic chalcopyrite grains have diameters of 100–500 μm, the individual grains are loosely arranged. In the interstices between these grains, amorphous silica exists. In the upper part, micro-zones of hypidiomorphic pyrite are intercalated. Fig. 6D shows the sample 13_B; large xenomorphic chalcopyrite grains (up to 200 μm) intimately intergrown with collomorph silica; the silica microsurfaces are covered by thin layers of marcasite (mrc). Fig. 6E shows the sample 03_A, a typical white smoker mineral association dominated by sphalerite (sph). The ore texture shows a week zonation. The grain sizes of the sphalerite vary between 50 μm and 150 μm. The larger interstices are filled with amorphous silica. The open pore spaces in the sphalerite zones are small, irregular and are not interconnected. The section contains finegrained aggregates of pyrite and chalcopyrite. In the central part, some remarkable marcasite crystals exist. Fig. 6F shows the sample 04_A, which is a typical white smoker mineral assemblage dominated by fine-grained massive sphalerite. Some xenomorphic pyrite grains are intergrown with sphalerite. The observed chalcopyrite lamellae are oriented intergrown with sphalerite and represent products of an exsolution process described as “chalcopyrite disease” (Halbach et al., 1993). Fig. 6G shows the sample 04_A, which is the typical white smoker mineral assemblage dominated by sphalerite. The sphalerite is fine- to mediumgrained and mostly hypidiomorphic to idiomorphic. The porosity is well interconnected and controlled by the surface constitution of the sphalerite grains. This specimen has, for example, a fairly low compressive strength value (10.1 MPa). Xenomorphic chalcopyrite is subordinately present; the pyrite grains are often hypidiomorphic to idiomorphic. Marcasite can also be observed and has a mottled to irregular distribution. Fig. 6H shows the sample 14_A; in this caprock mineral assemblage, sphalerite is also the dominant sulfide mineral. In general, the amorphous silica is the most abundant phase in the caprock. The sphalerite grains are coarse-grained up to 500 μm in diameter. In some microzones of smaller particle sizes, pyrite and chalcopyrite can also be observed. In between the sphalerite grains, large crystals of barite (bar) are distributed which often have a thin cover of amorphous silica. The porosity was also considered in this investigation by the microscopic point of view, since the intergrowth patterns of the porosity with the supporting mineral framework will have an influence on the geotechnical strength behavior. However, there is one major problem related to this step of study, since SMS samples, in general, have a very heterogeneous and strongly zoned setup and structure. The two mineralogical properties, porosity and mineral framework, are mainly caused by the complex history of formation (Halbach et al., 2003) based on multi-phase hydrothermal activity with varying temperatures of precipitation. The zonation patterns are scaled in the cm range. For example, a sulfidic chimney structure may have vertical zonation because the central part is marked by high-temperature precipitates (dominated by chalcopyrite) whereas the outer zones show sphalerite, pyrite and finally melnikovite-marcasite mineralizations, the latter are often mixed with anhydrite. This mineral sequence is caused by the decreasing temperature of mineral formation.

134

G. Spagnoli et al. / Engineering Geology 214 (2016) 127–135

In some cases, already over the length of the test cylinder, a change of ore facies was observed, thus, it is difficult to estimate which mineral facies eventually controlled the values of the UCS tests. One step to avoid the influences of changing mineral composition is to also carry out UCS tests with smaller sulfide cylinders. The number of samples and also the spectrum of the type of mineralizations were limited. Nevertheless, it can be stated that the UCS behavior of the individual measurements can be better described and understood when it is related to the mineralogical observations including the porosity texture (Fig. 7). For example, the sample with the highest UCS value (41.4 MPa, sample 07_A and B; Table 3) has not only the lowest porosity but is also characterized by a distinct mineralogy and texture. The dominating minerals are pyrite and melnikovite. Both minerals are very finegrained and densely intergrown, the observed porosity is very irregular and only partly interconnected. It can be stated, that these mineralogical observations (Fig. 6) will cause the high material strength. On the other hand, the sample with the lowest material strength (sample 05, Fig. 7) shows mineralogical features characterized by a very high porosity combined with a high degree of pore interconnection and very high amorphous silica content. The samples 04_A and 06_B, which are located in the diagram (Fig. 7) underneath the line of linear regression, are marked by a coarse-grained texture and a moderate degree of interconnected pore distribution. An additional parameter which influences the porosity as well as the material UCS is the maturation of the SMS. This maturation is age-controlled and related to processes caused by an epigenetic overprint of the mound deposits by later reactivation of the hydrothermal activity. The typical result of this secondary alteration is that the porosity is additionally filled with sulfide minerals, which increases the degree of cementation. The results of Waquet and Fouquet (2011) show that this cementation combined with renewed mineral precipitation increases the UCS values up to 100 MPa. Such high values do not exist in our dataset. The only example which already shows a low degree of maturation is the sample 07_A, which has the highest UCS value (see above) in this study. The strongest influence on the UCS of the SMS sample is due to the porosity: in general, the porosity exhibits an inverse relationship to the stability of the mineral substance. In Fig. 7 the individual measured values are plotted versus the porosity. As this diagram shows, we can distinguish between two kinds of interrelationship: 1) 6 samples (4 black smoker and 2 white smoker samples) can be linearly correlated with a very high coefficient of determination (R2 = 0.99; dashed line in Fig. 7); 2) the correlation of all 10 samples do not represent a linear relationship but show that an exponential relationship can be used, however, with a lower coefficient of determination (R2 = 0.61). Since we have only 10 samples to study the reasons for this dual behavior, it is difficult to derive clear mathematically based interdependencies, at present. Further influences controlling the UCS could be the average grain size, the average amorphous silica content (Table 3) and the degree of connected porosity. Also mineral processes of maturation (Waquet and Fouquet, 2011) which, in general, are caused by epigenetic hydrothermal overprint of the primary chimney fragments, may result in a decrease in porosity and thus in an increase in the overall strength. This kind of maturation is, in general, related to the formation of sulfide mounds. The microscopy of the sample 07_A/B shows that in the pore spaces of the sulfide material some well-crystallized pyrite grains can be observed which probably indicate a first evidence of maturation. However, most of the other samples do not show any remarkable influence of maturation. Considering the 6 samples of the linear correlation, we can state that the average grain size varies between 62 μm (sample with the highest UCS value) and 108 μm (sample 10_B). On the other hand, the samples 04_A, 06_B and 14_A show a lower porosity and larger grain sizes (100– 193 μm). The two extreme samples with regard to strength and porosity

(07_A/B and 05; Table 3) have some remarkable properties: sample 07_B is very fine-grained (63 μm), has a low porosity with a lower degree of interconnection and is dominated by a densely interlocked and finely intergrown texture of pyrite, marcasite and melnikovite (Fig. 6A and B); the sample 05 has the highest porosity and the SiO2 content is distinctly above the average. Considering the texture of the pore spaces, a high degree of interconnection can be observed. Summarizing, it can be stated that a trend between UCS and porosity is observed; however, it is necessary to study and investigate more samples to obtain more significant results, which can also be statistically interrelated. 4. Conclusions The research presents a detailed analysis on the physical and mineralogical properties of SMS samples coming from the North Fiji Basin and the MESO-site in the Central Indian Ocean. From the geotechnical point of view it was observed that the porosity influences the UCS values and to a lesser degree the BTS values. According to Waquet and Fouquet (2011) porosity decreases along with maturation; the closer the sulfides samples are to the surface, the less mature and the more porous they are. Normally, this behavior is common for sedimentary rocks as UCS increases with age due to increased lithification, compaction and reduced porosity. Regarding the BTS, the correlation is not as clear as for the UCS, although a trend exists. The basic mineral phases in the studied massive sulfide samples are pyrite, marcasite, melnikovite, chalcopyrite and sphalerite. Although the number of investigated samples is fairly low, some interdependencies between the physical properties and the mineralogy including the porosity were observed. The samples were subdivided in black smoker, white smoker and caprock mineralization distinguished by different mineral compositions. In general, massive sulfide samples have a very heterogeneous setup and structure. Based on the complex history of formation caused by a multi-phase hydrothermal activity with varying temperatures of precipitation, massive sulfides are marked by zonation patterns scaled in the cm-range. In some cases already over the length of the test cylinder a change of the ore phases and their textures was observed. One step to avoid the influences of small-scaled changing mineral composition is to also carry out UCS tests with smaller sulfide cylinders. For example, the sample with the highest UCS value (41.4 MPa, sample 07_A and B; Table 3) has not only the lowest porosity but is also characterized by a distinct mineralogy and texture. Pyrite and melnikovite are very fine-grained and densely intergrown, the observed porosity is very irregular and only partly interconnected, causing a high material strength. Sample 05, with the lowest material strength shows mineralogical features characterized by a very high porosity combined with a high degree of pore interconnection and very high amorphous silica content. In Fig. 7 the attempt was made to correlate the measured UCS values with the porosity. The respective diagram reveals two different kinds of interrelation: 6 samples could be described by a linear regression, but the total data set is controlled by an exponential regression of lower probability. It has to be assumed that the geotechnical properties strongly influence the kind of technology and the energy consumption to loosen and to excavate the respective minerals on the seafloor. The silica content, with the quartz can have a strong negative impact (high abrasivity) on the mining tool. Beyond it, the hydrostatic pressure under deep-sea conditions will influence the geotechnical properties (Miedema, 2014). Thus, one logical step of future research must be to study the physical properties under hyperbaric conditions, considering also the abrasivity under very high pressure conditions. Acknowledgement The authors wish to thank BAUER Maschinen GmbH for the financial support for this project and for the permission granted to publish these

G. Spagnoli et al. / Engineering Geology 214 (2016) 127–135

results and to the Bundesanstalt für Materialforschung und– prüfung (BAM) for the support during the geotechnical tests. References Alvarez Grima, M., Miedema, S.A., van de Ketterij, R.G., Yenigül, N.B., van Rhee, C., 2015. Effect of high hyperbaric pressure on rock cutting process. Eng. Geol. 196, 24–36. http://dx.doi.org/10.1016/j.enggeo.2015.06.016. Andrieux, P., Simser, B., 2001. Ground stability-based mine design guidelines at the Brunswick mine. In: Hustrulid, W.A., Bullock, R.L. (Eds.), Underground Mining Methods: Engineering Fundamentals and International Case Studies, Society for Mining, Metallurgy, and Exploration, Englewood, p. 207. Anthony, J., Bideaux, R., Bladh, K., Nichols, M., 2003. Handbook of Mineralogy. Mineralogical Society of America, Chantilly, VA 20151-1110, USA. Aydin, A., Basu, A., 2006. The use of Brazilian test as a quantitative measure of rock weathering. Rock Mech. Rock. Eng. 39, 77–85. Chang, C., Zoback, M.D., Khaksar, A., 2006. Empirical relations between rock strength and physical properties in sedimentary rocks. J. Pet. Sci. Eng. 51, 223–237. http://dx.doi. org/10.1016/j.petrol.2006.01.003. Farah, R., 2011. Correlations Between Index Properties and Unconfined Compressive Strength of Weathered Ocala Limestone. UNF Theses and Dissertations, p. 142 Paper. Halbach, P., Fouquet, Y., Herzig, P., 2003. Mineralization and compositional patterns on deep-sea hydrothermal systems. In: Halbach, P., Tunnicliffe, V., Hein, J.R. (Eds.), Energy and Mass Transfer in Marine Hydrothermal Systems. Dahlem University Press, Berlin, pp. 85–122. Halbach, P., Herzig, P., Giere, O., et al., 1996. Hydrothermalism in the North Fiji Basin: evolution of fluids, mass fluxes and special biological activity. Technical Cruise Report SO 99, Berlin, p. 106. Halbach, P., Koschinsky, A., Rahders, E., et al., 1998. Hydrothermal fluid development, material bilancing and special biological activity in the North Fiji Basin. Technical Cruise Report SO 134, Berlin, p. 148. Halbach, M., Halbach, P., Lüders, V., 2002. Sulfide-impregnated and pure silica precipitates of hydrothermal origin from the Central Indian Ocean. Chem. Geol. 182, 357–375. http://dx.doi.org/10.1016/S0009-2541(01)00323-0. Halbach, P., Pracejus, B., Märten, A., 1993. Geology and mineralogy of massive sulfide ores from the Central Okinawa trough. Japan. Econ. Geol. 88, 2210–2225. Hannington, M., Jamieson, J., Monecke, T., Petersen, S., Beaulieu, S., 2011. The abundance of seafloor massive sulfide deposits. Geology 39, 1155–1158. Hoek, E., Brown, E.T., 1980. Underground Excavations in Rock. Institution of Mining and Metallurgy, London. Kahraman, S., Fener, M., Kozman, E., 2012. Predicting the compressive and tensile strength of rocks from Indentation Hardness Index. J. South. Afr. Inst. Min. Metall. 112, 331–339.

135

Lepique, M., 2008. Empfehlung Nr. 10 des Arbeitskreises 3.3 Versuchstechnik Fels der Deutschen Gesellschaft für Geotechnik e. V.: Indirekter Zugversuch an Gesteinsproben – Spaltzugversuch. Bautechnik 85, 623–627. http://dx.doi.org/10. 1002/bate.200810048. Miedema, S.A., 2014. The Delft Sand, Clay and Rock Cutting Model. IOS Press, Delft. Mutschler, T., 2004. Neufassung der Empfehlung Nr. 1 des Arbeitskreises Versuchstechnik Fels der Deutschen Gesellschaft für Geotechnik e. V.: Einaxiale Druckversuche an zylindrischen Gesteinsprüfkörpern. Bautechnik 81, 825–834. http://dx.doi.org/10. 1002/bate.200490194. Nautilus Minerals, 2016. Technology Overview. http://www.nautilusminerals.com/irm/ content/technology-overview.aspx?RID=329 (accessed 13.10.16). Palchik, V., Hatzor, Y.H., 2004. The influence of porosity on tensile and compressive strength of porous chalks. Rock Mech. Rock. Eng. 37, 331–341. http://dx.doi.org/10. 1007/s00603-003-0020-1. Spagnoli, G., Freudenthal, T., Strasser, M., Weixler, L., 2014. Development and Possible Applications of Mebo200 for Geotechnical Investigations for the Underwater Mining. Proceedings of the Offshore Technology Conference http://dx.doi.org/10.4043/ 25081-MS. Spagnoli, G., Miedema, S.A., Hermann, C., Rongau, J., Weixler, L., Denegre, J., 2015. Preliminary design of a trench cutter system for deep-sea mining applications under hyperbaric conditions. IEEE J. Ocean. Eng. http://dx.doi.org/10.1007/s001090000086. Spagnoli, G., Hannington, M., Bairlein, K., Hördt, A., Jegen, M., Petersen, S., Laurila, T., 2016. Electrical properties of seafloor massive sulphides. Geo-Mar. Lett. 36, 235–245. http://dx.doi.org/10.1007/s00367-016-0439-5. Tomishima, Y., Yamazaki, T., Handa, K., Tsurusaki, K., 1989. Physical and Engineering Properties and Basic Consideration of Mining Machinery of Massive Submarine Polymetallic Sulfide Deposits. Proceedings of the 5th Symposium Deep-sea Research Using the Submersible “SHINKAI 2000” System, pp. 283–291 (in Japanese with English abstract). Waquet, B., Fouquet, Y., 2011. Evolution of Geotechnical Properties in Hydrothermal Sulphide Mounds: A Maturation Threshold. Proceedings of the SME Annual Meeting. Denver, CO. Waquet, B., Faulds, D., Benbia, A., 2011. Understanding the Effects of Deep-Sea Conditions on Seafloor Massive Sulfide Deposits Crushing Process Proceedings of the Offshore Technology Conference. Waltham, T., 2009. Foundations of Engineering Geology. CRC Press, Boca Raton. Yamazaki, T., Tomishima, Y., Handa, K., Tsurusaki, K., 1990. Engineering Properties of Deep-sea Mineral Resources. Proceedings of the 4th Pacific Congress on Marine Science and Technology, pp. 385–392. Yamazaki, T., Park, S.H., 2003. Relationship Between Geotechnical Engineering Properties and Assay of Seafloor Massive Sulphides. Proceedings of the 13th International Offshore and Polar Engineering Conference, pp. 310–316 Honolulu.