Scholarly article on topic 'Hydrocarbons from near-surface sediments of the Barents Sea north of Svalbard – Indication of subsurface hydrocarbon generation?'

Hydrocarbons from near-surface sediments of the Barents Sea north of Svalbard – Indication of subsurface hydrocarbon generation? Academic research paper on "Earth and related environmental sciences"

CC BY
0
0
Share paper
Academic journal
Marine and Petroleum Geology
Keywords
{"Yermak Plateau" / "Hinlopen Margin" / "Southern Nansen Basin" / "Sorbed gases" / Methane / "Eocene shales" / "Arctic ocean" / "Surface geochemistry"}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — Martin Blumenberg, Rüdiger Lutz, Stefan Schlömer, Martin Krüger, Georg Scheeder, et al.

Abstract The Barents Sea is considered as an important target for oil and gas exploration, but the petroleum potential of its shelf and slope regions is unknown. Here we present results of a research cruise to the Northern Hinlopen Margin at the transition to the Southern Nansen Basin and the Eastern Yermak Plateau. Multichannel reflection seismic data acquisition, heat flow measurements, and geochemical analyses of near-surface sediments obtained by gravity coring were conducted to study the northern Barents Sea shelf and the early opening of the Nansen Basin and decipher their petroleum potential. Seismic data indicate high thicknesses of up to ∼2000 m of Cenozoic sediments. Heat flow density values in the study area range between 67 and 108 mW/m2. The sediment samples were analysed for bulk geochemistry and sorbed hydrocarbon gases and for two sites for extractable hydrocarbons. Data from extractable (n-alkanes > n-C25) and bulk (HI and OI from Rock Eval) organic matter demonstrate predominantly terrigenous organic material, most likely derived from ice-transported allochthonous sediments. None of the sediments revealed substantial amounts of methane in pore waters, arguing against active hydrocarbon seepage in the studied areas. However, thermogenic gases sorbed to the sediment matrix (clay minerals, organic matter and/or carbonates) were found in concentrations of up to 600 ppb (on sediment wet wt. basis). For the samples from the Northern Hinlopen Margin and particularly from the adjacent Nansen Basin, a paleo fluid flow of thermogenic gas is indicated and accompanied by higher n-alkanes with a modal, petroleum-like distribution. δ13C values of methane, ethane and propane and gas compositions point at a mainly marine source rock origin of all studied gases with early oil window maturities of the associated rocks (0.6–0.9%VR). From this data an admixture of Type III derived thermogenic gases is indicated for some of the Yermak Plateau sediments for which also the lowest abundances of sorbed gases (50–100 ppb) were observed. Gas geochemical characteristics in the samples with low gas abundances can partially be explained by an input of gases through ice-transport of allochthonous hydrocarbons, which were bound to mature organic matter. For a site on the Northern Hinlopen Margin NE of Svalbard, right at the southern termination of the Nansen Basin a different situation is indicated. In this area the highest concentrations of sorbed gases most likely derived from sediments with an early-oil window maturity and a marine kerogen Type II-typical isotopic distribution. At this location a pseudo well was constructed from 2D seismic data for reconstruction of thermal and maturity evolution. The simulation results indicate that an Early to Middle Eocene source rock would be in the early oil window since the Early Miocene. A possible source rock here and in the circum-Arctic region could have been formed by Azolla algae and other flourishing primary producers.

Academic research paper on topic "Hydrocarbons from near-surface sediments of the Barents Sea north of Svalbard – Indication of subsurface hydrocarbon generation?"

Contents lists available at ScienceDirect

Marine and Petroleum Geology

journal homepage: www.elsevier.com/locate/marpetgeo

Research paper

Hydrocarbons from near-surface sediments of the Barents Sea north of Svalbard - Indication of subsurface hydrocarbon generation?

CrossMark

Martin Blumenberg*, Rüdiger Lutz, Stefan Schlömer, Martin Krüger, Georg Scheeder, Kai Berglar, Ingo Heyde, Philipp Weniger

Federal Institute for Geosciences and Natural Resources (BGR), Stilleweg 2, 30655, Hannover, Germany

ARTICLE INFO

Article history: Received 23 March 2016 Received in revised form 24 May 2016 Accepted 26 May 2016 Available online 27 May 2016

Keywords: Yermak Plateau Hinlopen Margin Southern Nansen Basin Sorbed gases Methane Eocene shales Arctic ocean Surface geochemistry

ABSTRACT

The Barents Sea is considered as an important target for oil and gas exploration, but the petroleum potential of its shelf and slope regions is unknown. Here we present results of a research cruise to the Northern Hinlopen Margin at the transition to the Southern Nansen Basin and the Eastern Yermak Plateau. Multichannel reflection seismic data acquisition, heat flow measurements, and geochemical analyses of near-surface sediments obtained by gravity coring were conducted to study the northern Barents Sea shelf and the early opening of the Nansen Basin and decipher their petroleum potential. Seismic data indicate high thicknesses of up to ~2000 m of Cenozoic sediments. Heat flow density values in the study area range between 67 and 108 mW/m2. The sediment samples were analysed for bulk geochemistry and sorbed hydrocarbon gases and for two sites for extractable hydrocarbons. Data from extractable (n-alkanes > n-C25) and bulk (HI and OI from Rock Eval) organic matter demonstrate predominantly terrigenous organic material, most likely derived from ice-transported allochthonous sediments. None of the sediments revealed substantial amounts of methane in pore waters, arguing against active hydrocarbon seepage in the studied areas. However, thermogenic gases sorbed to the sediment matrix (clay minerals, organic matter and/or carbonates) were found in concentrations of up to 600 ppb (on sediment wet wt. basis). For the samples from the Northern Hinlopen Margin and particularly from the adjacent Nansen Basin, a paleo fluid flow of thermogenic gas is indicated and accompanied by higher n-alkanes with a modal, petroleum-like distribution. 513C values of methane, ethane and propane and gas compositions point at a mainly marine source rock origin of all studied gases with early oil window maturities of the associated rocks (0.6—0.9%VR). From this data an admixture of Type III derived ther-mogenic gases is indicated for some of the Yermak Plateau sediments for which also the lowest abundances of sorbed gases (50—100 ppb) were observed. Gas geochemical characteristics in the samples with low gas abundances can partially be explained by an input of gases through ice-transport of allochthonous hydrocarbons, which were bound to mature organic matter. For a site on the Northern Hinlopen Margin NE of Svalbard, right at the southern termination of the Nansen Basin a different situation is indicated. In this area the highest concentrations of sorbed gases most likely derived from sediments with an early-oil window maturity and a marine kerogen Type II-typical isotopic distribution. At this location a pseudo well was constructed from 2D seismic data for reconstruction of thermal and maturity evolution. The simulation results indicate that an Early to Middle Eocene source rock would be in the early oil window since the Early Miocene. A possible source rock here and in the circum-Arctic region could have been formed by Azolla algae and other flourishing primary producers.

© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license

(http://creativecommons.org/licenses/by/4.0/).

1. Introduction and geological setting

The Arctic region is one of the major frontier areas for

* Corresponding author. E-mail address: martin.blumenberg@bgr.de (M. Blumenberg).

hydrocarbon exploration. While in the European Arctic most of the exploration activity is focused on the Southern Barents Sea, little is known about the Norwegian shelf and slope north of the Svalbard archipelago, between the Yermak Plateau to the west and the Nansen Basin to the north and east (Fig. 1). The presence of sedimentary rocks older than Cenozoic in this study area is unclear.

http://dx.doi.org/10.1016/j.marpetgeo.2016.05.031

0264-8172/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

While some argue that respective rocks are missing north of Sval-bard (Minakov et al., 2012), others argue that locally Paleogene and Mesozoic sediments are present and are superimposed by Neogene deposits in deep basins (Grogan et al., 1999). Although speculative, this includes Late Jurassic and Paleogene (marine) source rocks on the Yermak Plateau (Grogan et al., 1999). Large Quaternary and Cenozoic sedimentary packages are present north of Svalbard, which mostly derive from erosional transport through (i) regional uplift in the area during the Early Cenozoic and subsequently (ii) glacial erosion and relocalisation of massive sediments from southerly located regions (Rasmussen and Fjeldskaar, 1996). It is questionable whether these sediments contain productive hydrocarbon source rocks. Based on their estimated thickness and their potentially underlying Mesozoic sedimentary sequences hydrocarbon generation from deeply buried TOC-rich rocks, however, could be possible (Rasmussen and Fjeldskaar, 1996). Furthermore, Paleozoic sediments might be present between Svalbard and Yer-mak Plateau, if the underlying basement consists of stretched continental crust (Engen et al., 2008; Geissler and Jokat, 2004). But, the exact thicknesses and composition of these sediments is poorly understood. To date no direct information from drill core samples exists on deep subsurface lithology, thickness, and composition of sediments of the Northern Barents region and Nansen Basin. Therefore indirect methods have to be used to explore for potential petroleum systems. These methods include geophysical surveys, as well as near-surface geochemical prospecting such as detection and analysis of migrated hydrocarbons bound to seabed sediments.

2. Sorbed (bound) gases in marine sediments

Since the pioneering work by Horwitz (e.g., Horvitz (1972) and references therein) near-surface hydrocarbon prospecting has become a frequently used method in hydrocarbon exploration (Abrams, 1996a, b; 2013; Bernard et al., 1976; Horvitz, 1972; Logan et al., 2010; Stahl et al., 1981), especially in frontier areas (Abrams, 2013; Cole et al., 2001; McConell et al., 2008; Polteau et al., 2014).

In offshore regions this involves detection and analysis of oil and gas migrating from the subsurface into the water column in form of macroseepage, as well as the geochemical analysis of hydrocarbons retained in near-surface seabed sediments, which can be indicative for recent (active or passive) or past (micro) seepage. Seabed sediment samples are commonly obtained by gravity coring. They can contain migrated hydrocarbons in form of 1) free gaseous hydrocarbons in the pore space, 2) hydrocarbons dissolved in pore water, 3) hydrocarbons "bound" to the mineral matrix or organic matter by sorption or enclosure e.g. in fluid inclusion in carbonates, and 4) higher molecular weight hydrocarbons which can be extracted from the sedimentary organic matter. Free and dissolved gases are commonly investigated by headspace analysis (Abrams and Dahdah, 2010; Bernard et al., 1978), whereas "sorbed" or "bound" (both terms are in the following used as synonyms) hydrocarbon gases can be released from the sediment by mechanical agitation or milling, as well as by acid treatment (Horvitz, 1972; Knies et al., 2004; Stahl et al., 1981; Whiticar, 2002). Free, dissolved and sorbed gases mostly comprise light hydrocarbons (C1—C7) but can also contain permanent gases such as CO2, N2 and traces of noble gases. However, the exact mechanisms by which

Fig. 1. Sampling sites on the Hinlopen Margin, southern Nansen Basin and at the Yermak Plateau. Red circles indicate locations of gravity cores. Green triangles show locations of heat flow measurements. Yellow circles are locations of ODP sites. PS2138-1 is a core site described in Knies and Stein (1998). White lines show the locations of seismic profiles BGR13-206 (Fig. 5) and BGR13-208 (Fig. 6). Bathymetry is from Jakobsson et al. (2012). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

methane and higher hydrocarbons are bound to sediments are still not entirely understood and a matter of debate. Molecules can be bound by sorption (adsorption or absorption) to the surface of pores in organic matter or in mineral matter particles. The sorption capacity positively correlates with the micropore volume and specific surface area, which is larger in organic matter than in inorganic matter such as clay minerals or carbonate.

However, the basic principles, whether these gases originate from hydrocarbons that migrated from deeper petroleum systems and are bound and trapped in near-surface sediments or whether they result from microbial and diagenetic processes within shallow sediments, are still a matter of controversial technical (Abrams and Dahdah, 2011) but also conceptual discussion. For instance Hinrichs et al. (2006) proposed that gas-sediment sorption is a reversible process and is thus interfered by gas produced in the sedimentary biosphere (Ertefai et al., 2010). Others, however, argue that gas dissolved in pore water and sorbed gases are separated from each other and that rising gas is transported through "handshake migration" on mineral surfaces and retained in layers of "structure" or "ordered" water (Whiticar, 2002) or through buoyancy-driven propagation of methane-filled fractures (Nunn and Meulbroek, 2002) and no exchange between sorbed gas and dissolved gas is possible. A direct proof of these concepts, however, is still lacking. Different extraction protocols and desorption techniques yield different abundances, quantitative compositions, and potentially isotopic distributions, further complicating the comparability of results from different near-surface gas surveys (Ertefai et al., 2010). To underline this problem, some consider clay minerals as key for gas adsorption processes and adapt their protocols in order to disintegrate clay minerals (e.g. by NaOH treatment; Hinrichs et al., 2006). The potential importance of clay minerals led Horvitz (1980) to suggest wet sieving and enrichment of the <63 mm fraction for the study of sorbed gases. Others found a positive correlation between carbonate (CaCO3) and abundances of sorbed gases (Abrams, 2005; Brekke, 1997). Such carbonate-associated gas would not be released if NaOH is used as sole extraction agent. Also not of prime interest was as yet the possibility of re-deposition of sediments and its sorbed gas although transport and reworking mechanisms are of particularly high importance in some marine settings like in the Arctic (e.g. Knies and Stein, 1998; Piggot and Abrams, 1996).

Despite the poor understanding of the principles of origin and trapping mechanism of near-surface gas, sorbed gases have been widely and often successfully applied and are established to provide information on paleoseepage and on the productive source rocks (e.g., Bjor0y and L0berg, 1993; Cramer and Franke, 2005; Faber et al., 1990, 1997; Faber and Stahl, 1983, 1984; Knies et al., 2004; Mani et al., 2011; Patil et al., 2013).

The major aim of our study is to better understand the petroleum geology of the working area in the Arctic. Therefore, we present data on extractable and sorbed hydrocarbons from a near-surface gas prospecting survey in two areas North of Svalbard. Furthermore, our data may serve as a reference for future studies on sorbed hydrocarbon gases in other Arctic areas where explicit data on the presence and thermal maturity of source rocks are lacking.

3. Materials and methods

3.1. Sampling

Sediment samples were collected using a gravity corer with a 3 m long open barrel and a 1000 kg lead weight. After sampling from the liners, sediment samples for bound gases and extractable hydrocarbons were immediately frozen and stored at -18 °C until further processing in the laboratory. Sediment samples for head-space gas analysis were sampled from gravity cores using a syringe

from which the top was removed and transferred into headspace flasks.

3.2. Bulk analyses (TOC, Ccarb, XRF)

Total sulphur (TS) and total organic carbon (TOC) were measured on a LECO CS-230 (Leco Instrumente, Germany). TOC was determined on decalcified (acidification with 10% hydrochloric acid; HCl) and dried (80 ° C for 18 h) samples. About 200 mg of each sample was burned in a high-frequency induction furnace in an oxygen atmosphere by use of the absorption signal at the IR detector. The instrument was calibrated using commercially available standards (from LECO). Reproducibility of TS- and C-measurements (TOC and carbonate carbon) was ±0.01%.

X-ray fluorescence (XRF) analyses were performed using a PANanalytical Axios WDRF System. 1 g of homogenized sample were heated up to 1030 ° C to determine the loss of ignition, than mixed with 5 g lithium borate to prepare a fusion at 1200 °C prior to the XRF analysis. Data were used to estimate clay mineral abundances in Table 1 from Al2O3 (calculated according to Brumsack (1989)).

3.3. Bulk geochemical analyses (RockEval)

Rock-Eval pyrolysis was performed with a Rock-Eval VI on 22 sediment samples according to the procedures described in Espitalie et al. (1977) and Lafargue et al. (1998). Hydrocarbons, released during the pyrolysis cycle below 300 ° C and between 300 and 650 ° C, are presented as S1 and S2 yields, respectively. Precision of the hydrocarbon determination was better than 6%. Tmax values represent maxima of the S2 peak and are a measure of the thermal maturity.

3.4. Headspace gas and sorbed (or bound) gases

Five ml sediment samples were placed in 50 ml glass serum vials pre filled with 10 ml saturated HgCl2 solution and immediately closed with gastight rubber stoppers and aluminium crimp caps. The vials were then vortexed until the sediment was completely suspended and stored at 4 °C until GC analysis. Methane concentrations in fixed sediment samples were analysed by measuring headspace samples at 60 °C using a GC-FID equipped with a 6' Hayesep D column (SRI 8610C, SRI Instruments, USA).

For the analyses of sorbed gases, free gas in pore spaces, proposed to be mainly composed of microbial methane and analysed as above, was removed by evacuating the sample according to Faber et al. (1997). Then, detained gas was released from about 150 g of the sediment samples by treatment with phosphoric acid and heating to the temperature of boiling water under vacuum (Faber and Stahl, 1983).

Gases were analysed with an RGA (Refinery Gas Analyser) and a pre-separation was achieved on a HP1 4 m column. Subsequently, C1 — C6 hydrocarbons were separated on a 50 m Al2O3 HP Plot column. The temperature program was: 50 °C (3 min) — 20°/min. to 180 °C (held for 5 min). Hydrocarbons were detected with an FID and the carrier gas was He.

The stable carbon isotopic compositions of C1 — C3 hydrocarbons were analysed with gas chromatography isotope ratio mass spec-trometry (GC-IRMS), using an Agilent GC 6890 coupled to a Thermo MAT253. Samples were injected via a split-splitless injector either directly on column or via a sample loop. C1 — C3 were separated on a Poraplot Q column (ID 0.32 mm, 25 m). The GC-program was: -20 °C (held for 2 min), 8 °C/min heating and 14 min at a final temperature 180 °C. C1 — C3 were oxidized on a CuO/Ni/Pt combustion furnace operated at 960 °C. The carrier gas was He. The

carbon isotope ratios are expressed in a per mil deviation from the Vienna Belemnite (VPDB) standard in the usual delta-notation: 513C [(Rsample/Rstandard)-1] x 1000, where R is the 13C/12C ratio and Rstandard = 0.0112372. GC-C-IRMS precision was checked daily using a laboratory standard with known isotopic composition. Standard deviations for replicate injections were less than ±0.5%o.

For the analyses of sulfate, 5 ml pore water were acidified with 50 ml HNO3 and measured using an ICP-MS instrument (Perki-nElmerSciexElan5000, USA; see for details Dekov et al. (2006)).

fatty acid fractions had been derivatized using trimethyl-silyl-trifluroacetamid (MSTFA) as reactant.

The distribution of compounds in the separate fractions was determined with an Agilent 7890A gas chromatograph (GC) equipped with a PTV inlet splitting and two Ultra 1 capillary columns (Agilent; 50 m x 0.25 mm inner diameter; 0.11 mm film thickness), one coupled to a flame ionisation detector and the other one to a mass spectrometer system (MS; Agilent 7000). Compounds were identified by comparison of mass spectra and retention times.

3.5. Extractable hydrocarbons

About 20 g of wet sample was weighted into a centrifuge tube with screw cap. After adding a 30 ml solution of 6% potassium hydroxide in methanol the sample was saponified using an ultrasonic bath for 2 h at 80 ° C. To achieve a neutral lipid extract 10 ml isohexane was added to the mixture after cooling and agitated for 15 min before the following centrifugation at 3000 rpm for 5 min. Subsequently the supernatant was removed and stored. The extraction procedure was repeated three times and the superna-tants were combined to the neutral lipids and evaporated to dry-ness by a nitrogen stream at 30 °C.

The residue in the centrifuge tube was acidified with 3N HCl to a pH of 1—2 and three times extracted with isohexane in the above described manner to receive a fatty acid extract.

The neutral lipid extract was further fractionated by flash chromatography on a silica gel column (activated at 240 °C for 12 h) into three different polarity fractions. An aliphatic fraction was eluted with isohexane followed by a ketone fraction including the aromatic hydrocarbons eluted with dichloromethane. Finally an alcohol fraction was received using a mixture of dichloromethane/ methanol 1:1 as eluting solvent.

Prior to the gas chromatographic analysis the alcohol and the

3.6. Heat flow measurements

Heat flow measurements were carried out with a special hard ground probe due to the rigid sediments (drop stones, relatively coarse, ice-rafted debris) at the seafloor (Delisle and Zeibig, 2007). The hard-ground heat flow probe features a 2.2 m long sensor rod made of steel with a diameter of 2 cm mounted along the long axis of a cage and was held in position by a special mechanism to prevent bending during penetration of hard ground sediments. It contains 7 thermistors with a spacing of 28 cm. The necessary force to press the sensor rod into the sediments is provided by a cylinder, which houses lead plates with a total weight of 600 kg and an electronic unit within a pressure vessel with a total weight of additional 144 kg. Taking into account the weight of the probe and the cable, the water depth until which the probe could be deployed safely was limited to about 1950 m. The accuracy of measurement is ~0.002 K. To further improve the accuracy of the measurements, an arithmetic mean of 20 consecutive measurements per sensor was formed and then accepted as one single measured value. For all measurements the same sensor rod was used. Based on the stabilised ground temperature values at the different positions of the rod the temperature gradient was calculated by linear regression. The in-situ thermal conductivity of the penetrated sediments was

Table 1

Bulk data of sediments from the Hinlopen Margin slope/southern Nansen Basin and the E Yermak Plateau. Cmbsf = cm below sea floor- *Calculated from Ccarb (%) assuming Ccarb to reflect CaCO3 (Carb). # calculated from XRF Al2O3 data (after Brumsack, 1989) as part of the sum of quartz, clay minerals, carbonates. HI = Hydrogen Index (S2*100/ TOC). OI = Oxygen Index (S3*100/TOC). Higher contributions of fresh terrigenous OM to 02SL (Yermak) than to 08SL (southern termination of the Nansen Basin) indicated (very high OI). This is corroborated by e.g. the hopane distributions, which has a more immature signature in 02SL (Fig. 4). Interestingly for all cores it was observed that the top sample has much higher OI values than the sample from the deeper parts of the core. That this is just a signature of a diagenetic change, is unlikely, although possible. More likely is that in the deeper part a different depositional setting is mirrored than in the top part. n.d.: not detected. Concentrations of pore water methane may eventually be slightly overestimated due to minor acidification from HgCl2 and release of carbonate-bound gases. HF = Positions for Heat Flow measurements (see Table 3 for data). Data in bold are discussed in detail in the text.

Study area (station and position) cmbsf TOC (%) Carb (%)* Clay (%)# Tmax HI OI Pore water CH4 (mmol/l) Pore water SO4~ (mg/l)

Yermak Plateau

01SL (82.2431CN; 17.4845°E) 30-40 0.82 11.8 53.2 436 74 495 2.1 2720

01SL 170-180 0.58 1.5 58.9 438 57 239 2.4 2250

02SL (81.9294 N; 18.2878 E) 0-10 0.55 4.33 55.7 437 56 470 n.d. 2720

02SL 165-175 0.69 2.75 54.9 434 49 275 3.6 2200

04SL (82.2546CN; 17.9461°E) 0-10 0.93 12.08 53.4 440 73 454 n.d. 2710

04SL 190-200 0.49 1.83 57.8 432 47 290 n.d. 2570

16SL (81.3037CN; 19.8625°E) 4-14 1.47 4.00 56.6 426 78 252 2.3 2820

16SL 218-228 0.92 3.50 50.7 437 84 100 3.2 2370

Hinlopen Margin/southern Nansen Basin

06SL (81.6326CN; 32.8253°E) 4-10 0.56 2.58 52.5 430 56 255 n.d. 2760

06SL 150-160 0.40 6.92 52.2 435 48 163 2.1 2520

07SL (81.6717CN; 32.5747°E) 0-10 1.26 3.33 56.3 432 80 234 n.d. 2770

07SL 165-172 1.06 3.67 55.7 427 78 143 2.3 2770

08SL (81.7938 N; 31.7833 E) 0-13 1.26 6.17 56.8 434 81 331 n.d. 2750

08SL 200-210 0.85 10.58 49.2 439 100 178 3.7 2638

11SL (81.1232CN; 34.1950°E) 1-10 1.0 2.42 56.2 424 86 226 2.9 2783

11SL 60-70 0.68 5.75 45.5 441 88 144 n.d. 2755

13SL (80.6493CN; 34.4156°E) 3-10 1.16 2.5 56.8 424 84 281 2.8 2793

13SL 127-133 0.71 12.0 51.2 437 73 197 4.7 2758

14SL (80.6234CN; 34.4268°E) 25-35 0.98 4.83 53.4 430 95 183 2.3 2821

14SL 60-70 1.02 2.83 55.2 427 89 204 2.3 2774

15SL (80.6231°N; 34.4250°E) 0-10 1.10 3.33 55.8 446 100 245 4.9 2786

15SL 70-80 0.43 24.0 31.5 439 111 193 2.3 2670

derived from their temperature increase during electrically heating-up of the sensor rod. Visual inspections after recovery showed that the probe penetrated at all stations the sediments till the maximum depth of 2.2 m.

3.7. ID-modelling

1D basin and petroleum system modelling (BPSM) was performed to assess thermal maturity of a possible Early to Middle Eocene source rock. The model was built at the location of sample site 08SL from a pseudo-well at the southern termination of the Nansen Basin against the continental margin. The pseudo-well was constructed using the 2D seismic profile BGR13-208. Stratigraphic information and interval velocities for depth conversion were taken from Engen et al. (2009). Generally, basin and petroleum system modelling reconstructs the evolution of a sedimentary basin and calculates the generation, migration and accumulation of petroleum. Principles of basin and petroleum system modelling and the numerical realization of geological processes are described in Hantschel and Kauerauf (2009).

4. Results and discussions

Gravity cores were taken from the seafloor of the Hinlopen Margin and the adjacent Southern Nansen Basin, NE of Svalbard and on the Eastern margin of the Yermak Plateau (Fig. 1). Samples from the cores were analysed for bulk, gas, and pore water geochemistry. Due to the relatively high amounts of sediment needed for the analyses of sorbed gases, bulk and pore water analysis were performed on samples adjacent to the depths from which samples for sorbed gases were taken (an overview is provided in Tables 1 and 2).

4.1. Recent seep activity in the studied areas?

Usually, concentrations of pore water methane are high in areas of active seepage or if samples originate from the methanogenic zone e.g. in the Baltic Sea Gotland Deep concentrations in pore waters increase to >2 mmol/l from the surface to the methanogenic zone (Piker et al., 1998), similarly to the Black Sea (Reeburgh et al., 1991). Concentrations of methane in pore waters in the working area were low in all gravity core samples ranging from 2 to 5 mmol/l (Table 1), arguing against the presence of active hydrocarbon macro-seepage at all coring sites. Concentrations of free gas were further low due to the shallow depth of the sediment samples. All studied samples still contain high levels of SO4- down to the bottom of the sediment cores (Table 1), which clearly shows that the methanogenic zone was not reached even with the deepest samples from the gravity cores. Similarly low pore water methane concentrations were also reported from areas W of Svalbard, where samples most likely also originated from the sulfate reduction zone (multi-coring and box-coring was applied; Knies et al., 2004).

4.2. Bulk geochemistry and sorbed gases

Selected samples from 12 cores were analysed for bulk geochemical parameters (Table 1) and the compositions, abundances and carbon isotopic signatures of sorbed hydrocarbon gases (Table 2). TOC concentrations did not exceed 1.3% and were normally <1%. Carbonate (CaCO3) concentrations were also low and mostly <10% (Table 1). Concentrations of sorbed CH4 maximized at ca. 600 ppb (per wet wt. sed.) and — except for one sample — exceeded a suggested threshold for background concentrations of sorbed methane of about 25—50 ppb defined in Faber et al. (1997). This background definition, however, has to be used with caution as

backgrounds can vary heavily between different areas. Therefore backgrounds should ideally defined for certain working areas (Abrams, 2013), which is, however, not available for our relatively remote and usually ice-covered study area. All concentrations were lower than concentrations previously measured W of Svalbard (Knies et al., 2004), although comparisons are complicate considering a potential bias through the different sediment fractions used for the studies (<63 mm in Knies et al., 2004; no size fractionation was used for our study). Highest concentrations in Knies et al. (2004) were found along tectonic lineaments and potentially mirror past seepage activity at re-activated fault systems. Due to the relatively low sampling resolution, such conclusions cannot be made from our data set, but seismic data did not indicate the presence of fault systems near the sampling sites.

Using a modified 'Bernard' diagram (Bernard et al., 1977) sorbed gases indicate a predominantly marine kerogen (Type II) origin (Fig. 2). However, some samples suggest slight variations, particularly samples from 02SL (E Yermak Plateau; low sorbed methane concentrations with ca. 100 ppb) and 08SL (Southern Nansen Basin close to Hinlopen Margin slope; high sorbed methane concentrations with ca. 600 ppb) were found to be different, with 02SL plotting slightly closer to the Type III (terrigenous kerogen) field. Differences between 02SL and 08SL among the data set are also visible in the 513CH4 versus 513C2H6 plot (Fig. 3). The data from 02SL are particularly off the trend line for Type II organic matter maturation, which can be best explained with an admixture from Type III organic matter and would thus be in line with the distributional and isotopic differences shown in the 'Bernard' diagram (Fig. 2). However, microbial oxidation of methane would also shift 513CH4 values towards less negative values (shifting data points upwards in Fig. 3A). In accordance with others we consider this influence on sorbed gases to be low (Faber et al., 1997) and more likely that sorbed gases from 02SL (E Yermak Plateau) originate from Type II kerogen with minor contributions from Type III kerogen. Despite the fact that 02SL and 08SL are different, samples cannot be generally taken as representatives for all samples from each of the studied areas (see for instance Fig. 3 demonstrating a scatter in the data sets). However, due to their characters as "end-members" in our data set we used 02SL and 08SL for further considerations in our study. Vitrinite equivalent maturities estimated from the relationship in Fig. 3 range from 0.6 to 0.8% vitrinite reflectance (VR) with slightly higher maturities for the samples from the Yermak Plateau (e.g. 02 SL). In the study area transport of organic rich terrigenous sediments from Svalbard by ice-rafting or sub-marine mass movement is possible, but we consider it unlikely that gases from all locations were predominantly released from allochthonous material. This is for instance indicated from the (i) much higher maturities of the organic matter in the prominent source rocks from Svalbard (e.g., Berner et al., 2012) compared to those for the sediments studied here (Tmax values in Table 1). Further, (ii) hydrogen and oxygen indices (HI and OI, respectively) clearly demonstrate that the source of the sedimentary OM is mostly terrigenous, which stands in contrast with the Type II specific hydrocarbon compositions and carbon stable isotope systematic of sorbed gases (Figs. 2 and 3). And finally, (iii) no relationship between TOC and abundances of sorbed gases were found (Fig. S2), which would be expected ifTOC is mainly controlled by allochthonous old sediments from e.g. Svalbard. On the other hand, Tmax values were too high to be entirely explained by recent marine sediments (Table 1), arguing for that at least a portion of OM in the sediments is mature and allochthonously transported into the sediments. Together, we suppose that maturities derived from sorbed gases are valid for Type II source rocks in the sedimentary succession.

Table 2

Composition and carbon stable isotope ratios of sorbed gases in sediments from the Hinlopen Margin/southern Nansen Basin and the E Yermak Plateau.

Station no.

Adsorbed CH4 (ppb)

(C2+C3)

S13CH4 S13C2H6 S13C3H8

Yermak Plateau (D

01SL 155-165 136 9.0 - - - c CO -C

01SL 180-190 75 8.6 -41.1 -33.7

02SL 115-125 108 10.0 ¡36.5 ¡ 32.2 ¡ 30.4 0)

02SL 160-170 54 11.1 ¡37.1 ¡ 31.7 ¡29.4 E

04SL 180-190 151 9.5 -42.7 -33.0 -30.3 O

04SL 200-210 205 9.7 -41.5 -32.5 -29.5 CO

16SL 180-190 310 8.5 -40.7 -33.2 -30.2 MO

16SL 200-210 169 7.9 -41.0 -33.4 -30.0

Hinlopen Margin/southern Nansen Basin

06SL 130—140 128 8.2 -42.1 -33.6 -29.8

06SL 140—150 218 9.1 -40.9 -32.6 -28.9

07SL 90—100 142 8.5 -41.8 -33.3 -29.9

07SL 130—140 381 7.2 -39.9 -33.2 -31.0

08SL 180—190 613 5.3 ¡42.7 ¡34.7 ¡ 31.5

08SL 200—210 602 5.3 ¡43.3 ¡34.4 ¡ 31.1

11SL 30—40 185 6.8 -42.0 -33.7 -30.8

11SL 45—60 241 6.6 -39.0 -32.9 -30.7

12SL depth unclear 17 7.5 -38.8 -32.5 -30.0

13SL 65—75 166 7.0 -39.9 -33.3 -30.4

13SL 95—105 127 6.7 -40.6 -33.6 -30.8

14SL 50—60 90 7.0 -40.1 -33.0 -29.8

15SL 60—70 309 5.4 -40.4 -34.3 -29.6

4.3. Extractable hydrocarbons

In addition to sorbed gases extractable higher hydrocarbons were also studied for the two gravity cores 02SL (Yermak Plateau) and 08SL (southern termination of the Nansen Basin close to the Hinlopen Margin) and chromatograms are shown in Fig. 4. Sediments directly taken from the Hinlopen Margin were not studied, because it is assumed that allochthonous material is dominating there. Both extractable portions consist of one or a mixture of the following sources: (i) allochthonous organic matter transported to the working area, (ii) seeping hydrocarbons from underlying mature source rocks and/or (iii) hydrocarbons from anthropogenic

Fig. 2. Composition versus S C of sorbed methane in a modified 'Bernard' diagram (after Bernard et al., 1977).

Fig. 3. Plot of carbon isotope ratios of sorbed (A) methane versus ethane and (B) ethane versus propane. Maturity trend lines are based on reservoir gases (and not sorbed gases) and inferred maturities of the source rocks and are taken from Faber et al. (2015) and references therein. Numbers represent vitrinite reflectance equivalents in VR% and are modified according to the kerogen precursors. For Type II kerogen -27.5%0 (Hayes, 1983) and for Type III -24%o (Smith et al., 1981) were used.

activity or contamination during sampling. The latter can most likely be excluded due to the pristine working areas and the sampling strategy and is further unlikely as hydrocarbon suites in both studied samples differ to some extent. Hydrocarbons in both samples are composed of low molecular n-alkanes with a modal distribution (maxima at n-C22) overlain by high molecular weight n-alkanes with an odd/even predominance (maxima at n-C27). The suite of long chained n-alkanes is more prominent in 02SL (E Yermak Plateau), while vice versa in 08SL (southern Nansen Basin) an unresolved complex mixture (UCM) is more prominent. The presence of long chain n-alkanes with an odd-over-even predominance is usually due to high input of terrigenous OM as these compounds are abundant in plant waxes (e.g., Eglinton et al., 1962). The high quantitative importance of a plant input to the Yermak Plateau (02SL) is also reflected in the occurrence of long-chained n-alcohols with even-over-odd predominance and plant derived sterols (C29-sterols high; Fig. S1) as well as terrigenous-specific high C/N ratios

Fig. 4. Total ion chromatograms (TIC) of 02SL (160—170 cmbsf.; Yermak Plateau) and 08SL (180—190 cmbsf.; southern Nansen Basin). Inserts are showing the m/z 191 ion traces, mainly highlighting pentacyclic triterpanes (hopanes). Numbers represent carbon chain lengths of corresponding n-alkanes. UCM = unresolved complex mixtures. The curves highlight the suite of mature n-alkanes. For better comparison the chromatogram from the Yermak Plateau sediment extract (SL02) was threefold (3x) increased.

in sediments in close vicinity (PS2138-1 in Knies (1999); see Fig. 1 for position). In contrast, higher contributions of mature hydrocarbons are indicated for 08SL. This conclusion bases on, (i) the relatively higher abundances of mature low carbon chain n-alkanes than in 02SL and the (ii) high UCM. The semi-quantitative absolute differences, about 3 times less in 02SL than in 08SL also mirror the different relative abundances of sorbed gases (see above). Vitrinite equivalent reflectances of the mature portion in the hydrocarbons were further inferred from the calculation of methyl phenanthrene indices (MPI-1) of 0.7 and 0.9 and 20S/(20S + 20R)-ratios for C29-steranes of ~0.31 and 0.35, which hint to early oil window maturities (Boreham et al., 1988; Radke and Welte, 1983; Seifert and Moldowan, 1978).

4.4. Sources of sorbed gases and extractable hydrocarbons — Paleoseepage, allochthonous input from glaciations and deglaciations or other origins?

In the studied area Cenozoic sediments are prominent, either overlying continental or oceanic crust (Grogan et al., 1999; Minakov et al., 2012). Whether below the Cenozoic sediments older sediments may also be present remains speculative (see e.g. profile A — A' in Grogan et al. (1999), but is not considered unlikely (Grogan et al., 1999)).

In our data set no correlations between abundances of sorbed methane and TOC-concentrations were found (Fig. S2). Also no positive correlation was found between clay or carbonate mineral concentrations and sorbed gases challenging the role of clay minerals or carbonates for the sorption or enclosure of gases (Abrams,

2005; Brekke, 1997; Ijiri et al., 2009).

Sediments in the study area have a strong input of allochtho-nous organic matter of mostly terrigenous origin. This is clearly evidenced by C/N ratios and Rock Eval data from the core PS2138-1 in the vicinity of the study area (Knies, 1999, Fig. 1) and own Rock Eval data (Table 1). Our data support a scenario where the majority of organic matter in the recent and sub-recent sediments on the Hinlopen Margin, Southern Nansen Basin as well as on the Yermak Plateau are controlled by ice-rafted material, which was transported from e.g. Svalbard to the studied areas (Knies and Stein, 1998). Such processes are also clearly visible from relatively high amounts of terrigenous n-alkanes (Fig. 4) and plant sterols (Fig. S1). Unless palaeoseepage is a plausible scenario explaining the observed sorbed gases as well as portions of the extractable mature hydrocarbons, transport processes have also the potential to explain parts of these signals as well.

The origin of the mature hydrocarbon signature, however, is most likely not ice-rafted organic-rich shales and coals from e.g. Svalbard (Fig. 3). These rocks have different Rock Eval signatures and biomarker distributions (e.g., Abdullah, 1999; Cmiel and Fabianska, 2004; Schou et al., 1984; and own unpublished data). We therefore consider it most likely and due to the reasons discussed above that mature hydrocarbons and consequently also the majority of sorbed gases record paleofluid flow from a thermogenic source.

4.5. Sediment thickness and thermal maturity from ID-modelling Seismic data show the presence of thick sediment packages

Fig. 5. Seismic profile BGR13-206 displaying the sediments and the underlying basement at sites 02SL and 03SL (top). Close-up of the seismic profile (middle) showing details of the sediment stratigraphy and the contact to basement at site 02 SL. Sediment echosounder data displaying the transparent seismic facies of the Hinlopen Slide (bottom).

below the two sites 02SL and 08SL on which we focus in this study. At site 02 SL, imaged in seismic line BGR13-206 (Fig. 5), the upper most part of the profile exhibits a discontinuous seismic facies. This facies is limited to the upper ~150 ms below seafloor and is imaged in echosounder data nearly devoid of reflections (Fig. 5, bottom). We interpret these sediments as part of the Hinlopen Slide (seafloor to red horizon) (Winkelmann et al., 2008). Below this unit high amplitude continuous and parallel reflectors are visible which we interpret as well stratified sediments of probably Plio-Plistocene age (red horizon to blue horizon). The upper two units comprise ~800 ms of sediments. The base of this well-stratified unit is formed by an unconformity which is onlapped by the parallel reflectors. The underlying unit is characterized by high amplitude discontinuous reflectors. Internally, parallel reflectors with a wavy to hum-mocky pattern are visible and comprises about 800 ms. We interpret this unit as a sedimentary package of unknown age. The sedimentary units together have an approximate thickness of

2000 m. The base of the discontinuous unit is formed by the acoustic basement with a speculative composition. It could either be formed by sediments and/or metasediments or magmatic continental crust. Regional uplift and erosion of Tertiary and pre-Tertiary potential source rocks may have also occurred on the Yermak Plateau (Rasmussen and Fjeldskaar, 1996). Under such a scenario it is feasible that sediment packages, which are at present not in a suitable depth for petroleum generation may have been situated deeper in the geological past. Respective organic rich sediments may include those deposited during the Paleozoic (e.g. Billefjjorden Gp.) and Mesozoic (e.g. Late Jurassic; Agardfjellet Fm. equivalent on Svalbard). However, the lack of age control and the speculative nature of the basement at this site does not allow the construction of a pseudo-well for thermal maturity calculations.

Site 08SL is located at the southern termination of the Nansen Basin against the continental margin of the Northern Barents Sea. There is no well in the Nansen Basin, therefore stratigraphy and

Fig. 6. Seismic profile BGR13-208 displaying the sediments and the underlying basement at sites 06SL, 07SL and 08SL. Close-up shows location of the pseudo-well and sediment stratigraphy (units NB-1B to NB-4B) adopted from Engen et al. (2009). The pseudo well is located close to border faults which separate this part from the continental crust of the Northern Barents Sea.

Table 3

Heat flow density stations with position, water depth, temperature gradient and heat flow density value. Station no. Depth (m) Temperature gradient [mK/m] Thermal conductivity [W/mK] Heat flow density [mW/m2]

Yermak Plateau

HF1 (82.2601CN; 17.5581°E) 1954 72.1 1.03 74.3

HF2 (82.2434°N; 17.8463°E) 1970 66.5 1.17 77.8

HF3 (82.2691°N; 18.1198°E) 1897 69.6 1.1 76.6

Hinlopen Margin/southern Nansen Basin

HF4 (81.6932CN; 32.4560°E) 1821 87.1 1.24 108

HF5 (81.8348°N; 33.8772°E) 1910 85.7 1.19 102

HF6 (81.7766°N; 33.9042°E) 1750 55.7 1.21 67.4

HF7 (81.6209°N; 30.5363°E) 1737 56.3 1.21 68.1

lithology of the sediments have to be derived from other sources. Based on sparse MCS data Engen et al. (2009) developed a seis-mostratigraphic concept for the Nansen Basin taking into account wells on the Yermak Plateau for the upper most sedimentary unit NB-4 of Quaternary-Pleistocene age and, below, onlapping contacts of prominent seismic reflectors on oceanic crust of known age. The age of the crust was determined by magnetic anomalies. Engen et al. (2009) further reviewed previously published stratigraphic models. One of their seismic lines intersects our line BGR13-208, thus enabling a direct transfer of the stratigraphic model. Fig. 6 shows our seismic profile and the interpreted seismic horizons according to the stratigraphic model of Engen et al. (2009). The whole sedimentary sequence at this location is characterized by subparallel reflections. The basement shows high amplitude reflections at its top and is onlapped by the subparallel reflections of unit NB-1B at the site of our pseudo-well. Internally strong reflectors are visible which were used for further subdivision of the sediments into the individual units. The upper Unit NB-4 (NB-4A and NB-4B) is frequently affected by erosion and slumping.

During the PANORAMA-1 cruise heat flow measurements were

conducted in areas where seismic lines gave us information on the sediment thickness. The measurements were restricted to areas with water depths between ~1700 and ~1950 m. Altogether at seven stations heat flow measurements were conducted (Table 3 and Fig. 1). The heat flow density values at the southeastern Yer-mak Plateau are very homogeneous with a mean value of 76 mW/ m2. The values at the Hinlopen Margin vary between 67 and 108 mW/m2, with the highest value measured about 15 km southeast of 08 SL. The mean heat flow density value at the four stations is 86 mW/m2. The larger scatter of the values can be attributed to the greater area covered on the Hinlopen Margin.

The 1D model consists of a 2000 m thick basement layer of mantle rocks followed by five sedimentary units with a thickness of almost 1900 m and covering the time period from 54 Ma to present. The dominant lithology is shale with increasing proportions of silt and sand towards the top. Heat flow enters at the base of the numerical model and is the most significant input parameter together with burial depth which governs the maturation of organic matter. We used a declining heat flow trend from 100 mW/m2 to 85 mW/ m2 at present for the simulation which corresponds to the

Time TMal

Fig. 7. 1D model of the sedimentary succession at the southern termination of the Nansen Basin (see location 08SL on Fig. 1). Left picture shows the calculated vitrinite reflectance evolution at this site, indicating the petroleum generation zones. Green is oil window, red is gas window, same color scale left and right. Right picture shows the calculated present day vitrinite reflectance (solid line) and the derived petroleum generation zones. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

measured mean heat flow values at the Hinlopen Margin. The water depth at the pseudo-well location is 3000 m and we assume a constant increase of water depth due to cooling of the basement rocks and therefore progressing subsidence.

The 1D model shows that petroleum generation in the lowermost sedimentary unit of early Eocene age (NB-1) in the Nansen Basin was possible since the early Miocene and these sediments are still in the early oil window (Fig. 7). Sensitivity analysis carried out with a higher heat flow of 100 mW/m2 at present, shifts the onset of petroleum generation to the late Oligocene at ~25 Ma but the lower most sedimentary unit is still in the oil window. Assuming a lower present day heat flow of 75 mW/m2, shifts the onset of petroleum generation to the Late Miocene at ~7 Ma and also leaves the oldest sediments in the early oil window. Assuming thinned continental basement below the sediments would slightly increase maturity of the potential source rock due to radiogenic heat production. Nevertheless, the source rock would still be in the early oil window. The results of our simulation are similar to the results of Moore and Pitman (2011). They modelled a pseudo well ~60 km to the NW and calculated that Eocene sediments are in the oil window. Early to Middle Eocene sediments drilled at the Lomonosov Ridge have good source rock properties (Stein, 2007) but they are immature at this location. Mann et al. (2009) carried out petroleum system modelling at the Lomonosov Ridge and in the Amundsen Basin and assumed an additional overburden of 1000—1200 m, which exists in the Amundsen Basin and heat flow values ~100 mW/m2, due to the proximity to the Gakkel Ridge and concluded that conditions for hydrocarbon expulsion were reached.

We calculated early oil window maturities for the lowermost shales directly overlying basement rocks in the Nansen Basin. This is in accordance with data from sorbed gases which indicate, although this has to be done with caution, a marine source rock with early oil maturities of ~0.6%VR. In concert with the regional geology and results from drillings at the Lomonosov Ridge, Lower to Middle Eocene eventually Azolla rich shales (Boucsein and Stein, 2009; Brinkhuis et al., 2006) would be a plausible candidate.

Our data of sorbed gases on the Hinlopen Margin, southern Nansen Basin and the E Yermak Plateau suggest oil window maturities for potential source rocks, with slightly higher maturities at the Yermak Plateau. These low maturities argue against abundant

gas generation in the studied areas, but the presence of petroleum systems with oil generation is at least feasible.

Data off the western coast of Svalbard suggest a less homogenous situation compared to our working area. There, isotope compositions of the sorbed gases in sediments indicate source rock maturities from 0.6 to 3% VR (Knies et al., 2004). However, their and our data are in accordance with regionally differing petroleum systems and variable source rocks in the entire region (e.g., M0rk and Bjor0y, 1984).

5. Conclusions

Sorbed gases and for selected samples extractable hydrocarbons of sediments from the E Yermak Plateau as well as from the Hin-lopen Margin slope and the Nansen Basin were studied for compositions and for gas isotopic signatures. Concentrations of sorbed gases were partially high with highest abundances in samples from the southern Nansen Basin (>600 ppb wet wt. sed.). Indications for active seepage were not found as pore water methane concentrations were very low. Interpretations of isotopic signatures in the sorbed gases (methane to propane) can be taken as indication for a low maturity of the potentially corresponding source rocks. However, differences were observed between the E Yermak Plateau, the Hinlopen Margin and particularly the southern Nansen Basin. For the latter two, maturities of about 0.6% VR are suggested, while those for the first were slightly higher (~0.8—0.9% VR). Higher C1/ C2+-ratios and the isotopic composition of methane may be taken as hint to a mixture of mostly Type II with minor contributions from Type III source rocks for the situation at the E Yermak Plateau, whereas for sorbed gases in sediments from the Hinlopen Margin and the Nansen Basin a marine Type II source rock is suggested. For two selected samples, the extractable organic matter was also studied. Samples from the Yermak Plateau and the southern Nansen Basin contained high molecular weight hydrocarbons as well as long chain n-alkanes (and n-alcohols and sterols), indicative for an immature terrigenous plant source. Results of sorbed gases and extractable organic matter analyses and 1D modelling can be explained by two plausible alternatives (i) the gas (and possibly higher hydrocarbons with a petroleum-like distribution) stems from migrated hydrocarbons from deeper source rocks and the

major portion of extractable organic matter is of ice-transported material from e.g. Svalbard or (ii) hydrocarbon gas and extract-able organic matter are both of allochthonous origin. We favor explanation one, because ID modelling has shown that petroleum generation is possible in the Nansen Basin and at the E Yermak Plateau similar or even higher sediment thicknesses suggest that petroleum generation is also possible there. Our observation that no correlation between abundances of TOC and sorbed gases exists and also Rock Eval data suggest that the vast majority of OM in the youngest sediments is less mature than the known source rocks and coals of Svalbard further supporting the hypothesis above. Consequently, paleofluid upflow from marine Type II source rocks is considered as most plausible scenario to explain the sorbed gases on the E Yermak Plateau and the Hinlopen Margin but particularly in the adjacent Southern Nansen Basin.

Acknowledgements

The crew and captain or RV OGS Explora are thanked for excellent cooperation during field work. We thank Jürgen Poggenburg, Dietmar Laszinski, Daniela Graskamp, and Monika Weiß (all BGR Hannover) for laboratory assistance. One anonymous reviewer and Michael A. Abrams are thanked for their helpful comments.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.marpetgeo.2016.05.031.

References

Abdullah, W.H., 1999. Organic facies variations in the Triassic shallow marine and deep marine shales of central Spitsbergen\Svalbard. Mar. Petrol. Geol. 16, 467—481.

Abrams, M.A., 1996a. Distribution of subsurface hydrocarbon seepage in near-surface marine sediments. In: Schumacher, D., Abrams, M.A. (Eds.), Hydrocarbon Migration and its Near-surface Expression - AAPG Memoir 66. AAPG, Tulsa, Oklahoma, USA, pp. 1—14. Abrams, M.A., 1996b. Interpretation of methane carbon isotopes extracted from surficial marine sediments for detection of subsurface hydrocarbons. In: Schumacher, D., Abrams, M.A. (Eds.), Hydrocarbon Migration and its Near-surface Expression - AAPG Memoir 66. AAPG, Tulsa, Oklahoma, USA, pp. 309—318.

Abrams, M.A., 2005. Significance of hydrocarbon seepage relative to petroleum

generation and entrapment. Mar. Petrol. Geol. 22, 457—477. Abrams, M.A., 2013. Best practices for the collection, analysis, and interpretation of seabed geochemical samples to evaluate subsurface hydrocarbon generation and entrapment. In: Offshore Technology Conference, Houston. Texas, USA, pp. 1—21.

Abrams, M.A., Dahdah, N.F., 2010. Surface sediment gases as indicators of subsurface hydrocarbons — examining the record in laboratory and field studies. Mar. Petrol. Geol. 27, 273—284. Abrams, M.A., Dahdah, N.F., 2011. Surface sediment hydrocarbons as indicators of subsurface hydrocarbons: field calibration of existing and new surface geochemistry methods in the Marco Polo area, Gulf of Mexico. AAPG Bull. 95, 1907—1935.

Bernard, B., Brooks, J.M., Sackett, W.M., 1977. A geochemical model for characterization of hydrocarbon gas sources in marine sediments. In: Offshore Technology Conference.

Bernard, B.B., Brooks, J.M., Sackett, W.M., 1976. Natural gas seepage in the Gulf of

Mexico. Earth Planet. Sci. Lett. 31, 48—54. Bernard, B.B., Brooks, J.M., Sackett, W.M., 1978. Light hydrocarbons in recent texas

continental shelf and slope sediments. J. Geophys. Res. 83, 4053—4061. Berner, U., Cramer, B., Piepjohn, K., Sobolev, P., 2012. Hydrocarbon potential of the Western Barents Sea — evaluation of Mesozoic to Cenozoic sedimentary rocks of Spitsbergen. Oil Gas Europ. Magaz. 1,10—15. Boreham, C.J., Crick, I.H., Powell, T.G., 1988. Alternative calibration of the methyl-phenanthrene Index against vitrinite reflectance: application to maturity measurements on oils and sediments. Org. Geochem 12, 289—294. Bjoroy, M., Loberg, R., 1993. Surface geochemical studies in the Norwegian Barents sea: comparison with drilling results. In: Spencer, A.M. (Ed.), Generation, Accumulation, and Production of Europes's Hydrocarbons III. Springer, Berlin, pp. 99—109.

Boucsein, B., Stein, R., 2009. Black shale formation in the late Paleocene/early

Eocene Arctic ocean and paleoenvironmental conditions: new results from a detailed organic petrological study. Mar. Petrol. Geol. 26, 416—426.

Brekke, T., Lonne, 0., Ohm, S.E., 1997. Light hydrocarbon gases in shallow sediments in the northern North Sea. Mar. Geol. 137, 81—108.

Brinkhuis, H., Schouten, S., Collinson, M.E., Sluijs, A., Damste, J.S.S., Dickens, G.R., Huber, M., Cronin, T.M., Onodera, J., Takahashi, K., Bujak, J.P., Stein, R., van der Burgh, J., Eldrett, J.S., Harding, I.C., Lotter, A.F., Sangiorgi, F., Cittert, H.v.K.-v, de Leeuw, J.W., Matthiessen, J., Backman, J., Moran, K., the Expedition S, 2006. Episodic fresh surface waters in the Eocene Arctic Ocean. Nature 441, 606—609.

Brumsack, H.-J., 1989. Geochemistry of recent TOC-rich sediments from the Gulf of California and the Black Sea. Geol. Rund 78, 851—882.

Cmiel, S.R., Fabianska, M.J., 2004. Geochemical and petrographic properties of some Spitsbergen coals and dispersed organic matter. Int. J. Coal Geol. 57, 77—97.

Cole, G.A., Yu, A., Peel, F., Requejo, R., DeVay, J., Brooks, J., Bernard, B.B., Zumberge, J.E., Brown, S., 2001. Constraining source and charge risk in deep-water areas. WorldOil.com online Mag. 222.

Cramer, B., Franke, D., 2005. Indications for an active petroleum system in the Laptev Sea, NE Siberia. J. Petrol. Geol. 28, 369—384.

Dekov, V.M., Kamenov, G.D., Savelli, C., Stummeyer, J., Marchig, V., 2006. Origin of basal dolomitic claystone in the Marsili Basin. Mar. Geol. 236,121—141.

Delisle, G., Zeibig, M., 2007. Marine heat flow measurements in hard ground offshore Sumatra. EOS 88, 38—39.

Eglinton, G., Hamilton, R.J., Raphael, R.A., Gonzales, A.G., 1962. Hydrocarbon constituents of the wax coatings of plant leaves: a taxonomic study. Nature 193, 739—742.

Engen, 0., Faleide, J.I., Dyreng, T.K., 2008. Opening of the Fram Strait gateway: a review of plate tectonic constraints. Tectonophysics 450, 51 —69.

Engen, 0., Gjengedal, J.A., Faleide, J.I., Kristoffersen, Y., Eldholm, O., 2009. Seismic stratigraphy and sediment thickness of the Nansen Basin. Arct. Ocean. Geophys. J. Int. 176, 805—82 .

Ertefai, T.F., Heuer, V.B., Prieto-Mollar, X., Vogt, C., Sylva, S.P., Seewald, J., Hinrichs, K.-U., 2010. The biogeochemistry of sorbed methane in marine sediments. Geochim. Cosmochim. Acta 74, 6033—6048.

Espitalié, J., Laporte, J.L., Madec, M., Marquis, F., Leplat, P., Paulet, J., Boutefeu, A., 1977. Méthode rapide de caracterisation des roches mètres, de leur potentiel pétrolier et de leur degré dévolution. Oil Gas Sci. Technol. Rev. IFP 32, 23—42.

Faber, E., Berner, U., Hollerbach, A., Gerling, P., 1997. Isotope geochemistry in surface exploration for hydrocarbons. Geol. Rund. D103,103—127.

Faber, E., Schmidt, M., Feyzullayev, A., 2015. Geochemical hydrocarbon exploration -Insights from stable isotope models. Oil Gas Sci. Technol. 2.

Faber, E., Stahl, W., 1983. Analytical procedure and results of an isotope geochemical survey in an area of the British North Sea. Geol. Soc. Lond. Spec. Publ. 12, 51 —63.

Faber, E., Stahl, W., 1984. Geochemical surface exploration for hydrocarbons in North Sea. AAPG Bull. 68, 363—386.

Faber, E., Stahl, W.J., Whiticar, M.J., Lietz, J., Brooks, J.M., 1990. Thermal hydrocarbons in Gulf coast sediments, Gulf coast oils and gases: their characteristics, origin, distribution, and exploration and production significance. In: 9th Annual Research Conference, Gulf Coast Oils and Gases: Their Characteristics, Origin, Distribution, and Exploration and Production Significance, pp. 297—308.

Geissler, W.H., Jokat, W., 2004. A geophysical study of the northern Svalbard continental margin. Geophys. J. Int. 158, 50—66.

Grogan, P., Ostvedt-Ghazi, A.M., Larssen, G.B., Fotland, B., Nyberg, K., Dahlgren, S., Eidvin, T., 1999. Structural elements and petroleum geology of the Norwegian sector of the northern Barents Sea. In: Petroleum Geology Conference Series, 5. Geological Society, London, pp. 247—259.

Hantschel, T., Kauerauf, A.I., 2009. Fundamentals of Basin and Petroleum Systems Modeling. Springer Science & Business Media.

Hayes, J., 1983. Practice and principles in biosynthetic processes. In: Meinschein, W.G. (Ed.), Organic Geochemistry of Contemporanous and Ancient Sediments, Great Lakes Section. SPEM, Bloomington, Indiana, pp. 5—31.

Hinrichs, K.-U., Hayes, J.M., Bach, W., Spivack, A.J., Hmelo, L.R., Holm, N.G., Johnson, C.G., Sylva, S.P., 2006. Biological formation of ethane and propane in the deep marine subsurface. Proc. Natl. Acad. Sci. U. S. A. 103, 14684—14689.

Horvitz, L., 1972. Vegetation and geochemical prospecting for petroleum. AAPG Bull. 56, 925—940.

Horvitz, L., 1980. Near-surface evidence of hydrocarbon movement from depth. In: Roberts III, W.H., Cordell, R.J. (Eds.), Problems of Petroleum Migration, AAPG Memoir, 10, pp. 241—269.

Ijiri, A., Tsunogai, U., Gamo, T., Nakagawa, F., Sakamoto, T., Saito, S., 2009. Enrichments of adsorbed methane in authigenic carbonate concretions of the Japan Trench. Geo-Mar. Lett. 29, 301 —308.

Jakobsson, M., Mayer, L., Coakley, B., Dowdeswell, J.A., Forbes, S., Fridman, B., Hodnesdal, H., Noormets, R., Pedersen, R., Rebesco, M., Schenke, H.W., Zarayskaya, Y., Accettella, D., Armstrong, A., Anderson, R.M., Bienhoff, P., Camerlenghi, A., Church, I., Edwards, M., Gardner, J.V., Hall, J.K., Hell, B., Hestvik, O., Kristoffersen, Y., Marcussen, C., Mohammad, R., Mosher, D., Nghiem, S.V., Pedrosa, M.T., Travaglini, P.G., Weatherall, P., 2012. The International Bathymetric Chart of the Arctic Ocean (IBCAO) Version 3.0. Geophys. Res. Lett. 39 http://dx.doi.org/10.1029/2012GL052219.

Knies, J., 1999. Late Quaternary paleoenvironment along the northern Barents and Kara seas continental margin. A multi parameter analysis. Ber. Polarforsch 304.

Knies, J., Damm, E., Gutt, J., Mann, U., Pinturier, L., 2004. Near-surface hydrocarbon anomalies in shelf sediments off Spitsbergen: Evidences for past seepages. Geochem. Geophys. Geosystems 5. Q06003.

Knies, J., Stein, R., 1998. New aspects of organic carbon deposition and its

paleoceanographic implications along the Northern Barents Sea Margin during the last 30,000 years. Paleoceanogr 13, 384—394.

Lafargue, E., Marquis, F., Pillot, D., 1998. Rock-Eval 6 applications in hydrocarbon exploration, production, and soil contamination studies. Rev. l'Institut français pétrole 53, 421—437.

Logan, G.A., Jones, A.T., Kennard, J.M., Ryan, G.J., Rollet, N., 2010. Australian offshore natural hydrocarbon seepage studies, a review and re-evaluation. Mar. Petrol. Geol. 27, 26—45.

Mani, D., Patil, D.J., Dayal, A.M., 2011. Stable carbon isotope geochemistry of adsorbed alkane gases in near-surface soils of the Saurashtra Basin, India. Chem. Geol. 280, 144—153.

Mann, U., Knies, J., Chand, S., Jokat, W., Stein, R., Zweigel, J., 2009. Evaluation and modelling of Tertiary source rocks in the central Arctic Ocean. Mar. Petrol. Geol. 26,1624—1639.

McConell, D., Gharib, J., Henderson, J., Digby, A., Orange, D.L., 2008. Seep-hunting in deepwater for frontier basin prospectivity assessment. WorldOil.com Online Mag. 229.

Minakov, A., Faleide, J.I., Glebovsky, V.Y., Mjelde, R., 2012. Structure and evolution of the northern Barents—Kara Sea continental margin from integrated analysis of potential fields, bathymetry and sparse seismic data. Geophys. J. Int. 188, 79—102.

Moore, T.E., Pitman, J.K., 2011. Geology and petroleum potential of the Eurasia Basin. In: Spencer, A.M., Embry, A.F., Gautier, D.L., Stoupakova, A.V., Sorensen, K. (Eds.), Arctic Petroleum Geology. Geol. Soc. London Mem., 35, pp. 731—750.

Mork, A., Bjoroy, M., 1984. Mesozoic source rocks on Svalbard. In: Spencer, A.M. (Ed.), Petroleum Geology of the Northern European Margin. Graham & Trotman Limited, Trondheim.

Nunn, J.A., Meulbroek, P., 2002. Kilometer-scale upward migration of hydrocarbons in geopressured sediments by Buoyancy-driven propagation of methane-filled fractures. AAPG Bull. 86, 907—918.

Patil, D.J., Mani, D., Madhavi, T., Sudarshan, V., Srikarni, C., Kalpana, M.S., Sreenivas, B., Dayal, A.M., 2013. Near surface hydrocarbon prospecting in Mesozoic Kutch sedimentary basin, Gujarat, Western India—a reconnaissance study using geochemical and isotopic approach. J. Petrol. Sci. Eng. 108, 393—403.

Piggot, N., Abrams, M.A., 1996. Near-surface coring in the Beaufort and Chukchi Seas, Northern Alaska. AAPG Mem. 66, 385—399.

Piker, L., Schmaljohann, R., Imhoff, J.F., 1998. Dissimilatory sulfate reduction and methane production in Gotland Deep sediments (Baltic Sea) during a transition period from oxic to anoxic bottom water (1993—1996). Aquat. Microb. Ecol. 14, 183—193.

Polteau, S., Planke, S., Mazzini, A., Myklebust, R., 2014. Seafloor sampling of frontier basins for detection of active hydrocarbon systems and stratigraphic tie. In: EAGE Workshop Conference Proceedings.

Radke, M., Welte, D.H., 1983. The Methylphenanthrene Index (MPI): a maturity parameter based on aromatic hydrocarbons. Adv. Org. Geochem. 1981, 504—512.

Rasmussen, E., Fjeldskaar, W., 1996. Quantification of the Pliocene-Pleistocene erosion of the Barents Sea from present-day bathymetry. Glob. Planet. Change 12,119—133.

Reeburgh, W.S., Ward, B.B., Whalen, S.C., Sandbeck, K.A., Kilpatrick, K.A., Kerkhof, L.J., 1991. Black Sea methane geochemistry. Deep-Sea Res. 38, S1189—S1210.

Schou, L., Mork, A., Bjoroy, M., 1984. Correlation of source rocks and migrated hydrocarbons by GC-MS in the middle Triassic of Svalbard. Org. Geochem 6, 513—520.

Seifert, W.K., Moldowan, J., 1978. Applications of steranes, terpanes and mono-aromatics to the maturation, migration and source of crude oils. Geochim. Cosmochim. Acta 42, 77—95.

Smith, J.W., Gould, K.W., Rigby, D., 1981. The stable isotope geochemistry of Australian coals. Org. Geochem 3, 111—131.

Stahl, W., Faber, E., Kirksey, D.L., 1981. Near-surface evidence of migration of natural gas from deep reservoirs and source rocks. AAPG Bull. 65,1543—1550.

Stein, R., 2007. Upper Cretaceous/lower Tertiary black shales near the North Pole: organic-carbon origin and source-rock potential. Mar. Petrol. Geol. 24, 67—73.

Whiticar, M.J., 2002. Characterization and application of sorbed gas by micro-desorption CF-IRMS. In: AAPG Hedberg Conference.

Winkelmann, D., Schaofer, C., Stein, R., Mackensen, A., 2008. Terrigenous events and climate history of the Sophia Basin, Arctic Ocean. Geochem. Geophys. Geosyst 9. Q07023.