In this issue:

The paper by Shahriar et al. (pp 363‐382) presents 2D seismic, well log and other data from the Sangu gasfield, offshore Bangladesh, in order to reinterpret the field’s lithostratigraphy and structure and to evaluate possible remaining gas reserves. The field was discovered in 1996 and had reported Gas Initially In‐Place (GIIP) of about 1612 BCF. Production ceased in 2013 due to low reservoir pressures, by which time the field had produced about 487 BCF of gas. Of the remaining reserves, an estimated 358 BCF are now thought to be recoverable using advanced production techniques. The structure at Sangu is an asymmetric NW‐SE trending anticline at the western margin of the Chittagong Tripura Fold Belt which bounds the Bengal Basin to the east. Gas is generated by organic‐rich shale source rocks in the Miocene Bhuban Formation. The reservoir units consist of stacked, upper Miocene ‐ Pliocene deltaic and deeper‐water sandstones which at Sangu occur within a km‐thick Neogene sandstone‐shale succession. The reservoir units were investigated using wireline logs from two wells and seven 2D seismic profiles (two strike lines and five dip lines) across the Sangu structure with a total line length of about 300 km. Seven gas‐bearing reservoir units were identified, and their petrophysical characteristics were interpreted from well log and limited core sample data. The average total porosity of the reservoir units was >13% (maximum porosity ~25%); the permeability was in general less than 10 mD, and the gas saturation ranged from 24% to 80%. From the seismic data, prominent reflecting horizons correlating with the reservoir units were mapped in terms of depth allowing the extent and geometry of the Sangu anticline to be defined. Amplitude extraction maps for the tops of the reservoir units appear to show the presence of a series of palaeo channels which are interpreted to be clay‐filled, and were useful in locating potential gas‐filled reservoir sands. These palaeo‐channels will probably result in compartmentalisation of the reservoirs whose drainage areas therefore need carefully to be assessed, for example using 3D seismic, before there is any further development of the field.

Milkov (pp. 383‐400) investigated the forecasting ability of individual participants in a survey which asked them to predict the success or failure of eleven exploration wells worldwide, six of which were subsequently drilled. Participants were asked to make seven predictions regarding the outcomes of the wells, and the predictions were then compared with the actual exploration results. Predictions of the wells’ probability of success (PoS: i.e. the probability of finding movable petroleum) were very variable; point forecasts (answers to binary or multiple choice questions) were found on average to be only slightly better than random chance. The anonymous survey participants also provided personal information such as their academic achievements, employment status and experience, and familiarity or otherwise with the eleven wells. On the basis of this information, the survey results appeared to suggest that years of experience had little influence on forecasting ability, and nor did the level of a participant’s highest university degree. Moreover and perhaps counter‐intuitively, the duration of time spent on the survey did not affect a participant’s forecasting quality. In addition, those survey participants who described themselves as having some background knowledge of the wells on average made worse predictions than less well‐informed participants who relied only on limited prospect information. However, the aggregated opinion of all survey participants regarding forecast geological PoS for the six drilled wells was better than that of individual participants, consistent with the concept of the “wisdom of the crowd”. The total number of participants in the survey was 104 and not all participants answered all questions. These survey results are to some extent therefore preliminary and await validation and confirmation by other forecasting studies based on larger datasets.

Middle‐Upper Jurassic organic‐rich marls and shales are believed to have source rock potential in basins throughout much of Ethiopia and adjacent parts of Somalia and beyond. In the Blue Nile Basin in eastern‐central Ethiopia, the Early Callovian ‐ Late Tithonian Antalo Limestone Formation is thought to have particular potential as a source rock and is the subject of the paper by Mohammedyasin et al. on pp 401‐418. The paper reports the results of organic geochemical, biomarker and petrographic analyses of samples from two outcrop locations in the centre of the basin. Samples analysed included organic‐rich black shales from the Antalo Limestone together with coals and coaly shales from a directly‐underlying Middle Jurassic unit. The coaly shales contained terrigenous organic matter dominated by vitrinite macerals, and this landplant‐derived material is interpreted to have some potential to generate natural gas. By contrast, organic matter in the Antalo Limestone black shale samples was found to be dominated marine algal‐ derived material (Type II kerogen) of fairly uniform composition and with high HI values averaging 575 mg HC/ g TOC. The black shale samples were interpreted to have been deposited in generally anoxic‐suboxic conditions as indicated by ratios of pristane and phytane over the adjacent n‐alkanes, sterane patterns and other biomarker parameters. These samples were assessed as having very good to excellent source rock potential. Although the Antalo Limestone samples from the study area were found to be thermally immature with %R_r of 0.21 to 0.47%, maturities may be higher in other parts of the Blue Nile Basin. Thus a seepage oil in the NE of the basin, indicating the occurrence there of a working petroleum system, has been related in previous studies to a marine source rock which is inferred to be the Antalo Limestone.

Galushkin et al. present the second and concluding part of a numerical modelling study of the conjugate margins of SW Australia and East Antarctica. The eastern passive margin of Antarctica formed during Middle Jurassic to Late Cretaceous rifting, with separation of the Australian and Antarctic continents interpreted to have taken place in the Turonian. The first part of the study (July 2020 JPG, 43, 323‐340) reported on the subsidence and thermal histories of the Bremer sub‐basin, offshore SW Australia. Part 2 on pp 419‐436 of this issue investigates the Mawson Sea Basin on the stretched continental margin of eastern Antarctica. Both investigations used the GALO numerical modelling programme which simulates heat transfer in the crust and underlying lithosphere and asthenosphere down to depths of up to 100 km and models tectonic subsidence over time. Input data for the present paper came from a roughly north‐south seismic profile in the western Mawson Sea acquired in 2014, together with geophysical assessments of the deep crustal and lithospheric structure including the depth of the Moho. The lithostratigraphy of the Mawson Sea Basin was interpreted from a seismic stratigraphic model and was assumed in general to be analogous to that in the better studied Bremer sub‐basin, where six seismic stratigraphic units have been identified ranging from Upper Jurassic to Upper Cretaceous. However no deep formation temperature or vitrinite reflectance data for the Mawson Sea Basin are available so the thermal modelling results are to some extent unconstrained. The modelled subsidence history suggests that a phase of lithospheric extension occurred before the onset of synrift deposition at —160 Ma (Late Jurassic). Maturation modelling at pseudo‐well locations along the profile suggested that potential Lower Jurassic source rocks may be sufficiently mature (R_o —1.30%) to generate light oil and gas where deeply buried; heavier oil may also have been generated but was subsequently thermally degraded. Source rock maturities are inferred to be lower outside of these locations.

The paper by Cheng et al. (pp. 437‐452) reports on the biomarker characteristics of light oils and condensates recovered from wells in the western margin of the Xihu Sag, a structural depression in the east of the East China Sea Shelf Basin, offshore SE China. The hydrocarbons are interpreted to be derived from coals and coaly source rocks in the Eocene Pinghu Formation, and to have been generated at early to peak oil window maturities. Twelve drill‐stem test samples were recovered from Paleogene reservoir rocks at seven wells and were analysed by gas chromatography ‐ mass spectrometry (GC‐MS). The analytical results showed that the liquid hydrocarbons contained relatively high concentrations of diterpanes, including bi‐, tri‐ and tetra‐cyclic components with varying structures. The presence of n‐alkanes over a range between C₉ and C₃₄ together with other data show that the analysed samples have not been biodegraded. The relative abundance of the diterpanes is therefore inferred to be a primary compositional feature of the light oils analysed and is not due to selective biodegradation of petroleum in the reservoir. The presence of the diterpanes and other observations including the high pristane: phytane ratio (5.7‐10.7), and the dominance of C₂₉ regular steranes over C₂₇ and C₂₈ homologues together suggest a terrigenous depositional setting for the source rock and that organic matter was dominated by higher landplant material. The diterpanes are interpreted to be derived from precursors which are known to be common in conifer resins, and gymnosperm (mainly conifer) pollen has been recorded in Pinghu Formation source rocks at a nearby well. The petroleum analysed therefore appears to have been generated by a compositionally unusual coaly source rock which was deposited in relatively oxic fluvial‐deltaic conditions. Precursor organic matter is interpreted to have contained significant gymnosperm (mainly conifer) material although there was also an unquantified contribution from angiosperms (flowering plants).