Petroleum systems in the Amadeus Basin: Were they all oil prone? Petroleum systems in the Amadeus Basin, central Australia: Were they all oil prone? TR Marshall1, IA Dyson2 and Keyu Liu3

The Amadeus Basin in central Australia is a broad portion of the Amadeus Basin succession. This important intracratonic feature, covering some 170 000 km2 (65 500 result builds on source rock studies performed over the square miles). It has been the subject of petroleum last 40 years, and demonstrates that not only was the exploration for almost 50 years, and is productive Neoproterozoic capable of producing oil, in fact, it has. from Ordovician reservoirs at both the Mereenie (oil/ This opens up a rich new vein of exploration gas) and Palm Valley (gas) fields. As part of an NTGS possibilities in the Amadeus Basin, and future exploration project, a number of wells were sampled for analysis models should be refined to focus on the Neoproterozoic by QGF, QGF-E and GOI methods, patented CSIRO for oil, as well as gas. Petroleum fluid history analysis techniques that detect oil exclusively. The Neoproterozoic part of the Amadeus Keywords: , Amadeus Basin, Basin succession was targeted by this study to ascertain Neoproterozoic, Cambrian, Ordovician, fluid history if oil had been present in any of the wells sampled. analysis, QGF analysis, QGF-E analysis, GOI analysis, Failure analysis indicates most dry wells were drilled off FOI analysis, stratigraphy, Finke-1, petroleum exploration, structure; therefore, at those locations, analytical results petroleum geology, hydrocarbons, petroleum accumulation, should reflect a migration pathway. migration, petroleum potential QGF and QGF-E data confirmed that oil had been present in the samples, and subsequent GOI analysis Introduction confirmed that it was not captured in accumulations, therefore the analysed zones are interpreted to be The Amadeus Basin is an asymmetric, east–west migration pathways. This is consistent with the trending, intracratonic depression covering approximately interpretation that, due to lack of information (seismic) 170 000 km2 (65 500 square miles) of central Australia available pre-drill, most wells were drilled off structure, (Figure 1). It contains a Neoproterozoic to Late Devonian which in turn suggests that there may still be untested sedimentary section that reaches a maximum thickness of potential in some of the ‘drilled’ prospects in the Amadeus approximately 17 km (Figure 2). The basin is bounded to Basin. the north by the Precambrian Arunta Complex, and to the The results of analyses from Finke-1 are the most south by the Musgrave–Mann Complex and the Olia Gneiss. interesting as they strongly suggest that a 50+ m palaeo- To the west and east, the present basin margins are obscured oil column is present at the well location. The source of by a cover of Permian and younger sediments. some of the oil is likely to be from the Neoproterozoic The distribution of Late Proterozoic sediments was controlled by an east–west trending hinge-line (Angas 1 Vibrante Solutions, 74 Alice Street, Sefton Park, SA 5083. Lineament, Marshall and Dyson 2007), across the north 2 Salt Tectonics Australia Research, PO Box 221, Aldgate SA 5154. central part of the basin, from which the thickness of section 3 CSIRO Petroleum, PO Box 1130, Bentley WA 6102. increases both northward and southward. The Palaeozoic

0°

131° 133°

134° 132° 13

Alice Springs Palm Valley AY HW Mereenie G -24° HI

-25° LASSETER HIGHWAY

Yulara N

T STUAR

0 50 100 km

129° A06-137.ai Figure 1. Location map showing the Amadeus Basin in central Australia. Black lines represent axial traces of anticlines defined by outcrop, seismic and airborne magnetic data.

 Marshall et al

section is thickest along the present northern margin of the AGE SW STRATIGRAPHY NE basin, and thins depositionally to the south. PERMIAN BUCK FM OROGENY The Amadeus Basin is a proven petroleum province containing initial recoverable, proved, probable and possible LATE BREWER CGL reserves estimated at 37 million barrels of oil and 427 billion DEVONIAN HERMANNSBURG SST cubic feet of gas (Jackson et al 1984). Most hydrocarbon PARKE SLTST occurrences, including the oil and gas reserves at Mereenie

PERTNJARA GP PERTNJARA EARLY MOVEMENT and Palm Valley, have been found in Ordovician sandstone. DEVONIAN Cambrian reservoirs, particularly in the Arumbera Sandstone MEREENIE SST (see Figure 2), have been the main exploration objective in the eastern Amadeus Basin. The small Neoproterozoic P SILURIAN Dingo gasfield (currently uneconomic) also occurs there. CARMICHAEL SS TSST The Amadeus Basin is a salt-cored fold and thrust

RODINGAN belt, with multiple episodes of deformation (both tectonic LATE MOVEMENMOVEMENTT STOKESSTOKES SLTST SLTST and halokinetic), spanning some 800 million years. These ORDOVICIAN continuous events have strongly impacted on the distribution

STAIRWAYSTAIRWAY SST SST and type of source, seals, and reservoirs at both regional and EARLYEARLY LARAPINTA GROUP GROU 5 local scales. ORDOVICIAN HORNHORN VALLEY VALLEY SLTS T SLTST The aim of this paper is to briefly review the lithostratigraphic petroleum systems present in the P 1 PACOOTA SSTSST P 2 Amadeus Basin and offer insights, using new data, into their ? ? ? P 3 hydrocarbon phases. It presents additional new evidence ? ? ? P 4 that gas-prone Neoproterozoic sedimentary rocks have also produced oil, but it is beyond the scope of the paper to discuss ? upper GOYDER FM source rock characteristics in detail. Previous authors (eg LATE Kurylowicz 1976, McKirdy 1977, Oterdoom 1982, Gorter lower P lower GOYDER FM GOYDER FM 1982, 1983, Jackson et al 1984, Summons and Powell 1991, PETER- JAY CREEK MANN LST Marshall 2004) have presented geochemical studies on the FM SHANNON capacity of various parts of the Amadeus Basin succession to DECEPTION FM CLELAND SST produce hydrocarbons (gas and liquids); this paper presents SST HUGH RIVER SHALE MIDDLE ILLARA evidence that the Neoproterozoic did generate liquids. The SST GILES CREEK CAMBRIAN DOL CSIRO techniques discussed in this paper do not measure/ TEMPLE FM detect the presence of gas, and therefore no comments are ? CHANDLER FM FM TODDTODD made on gas generation from the Neoproterozoic of the RIVERRIVER IV DODOLL basin in this paper. PERTAOORRTA GROU MOUNMOUNTT ENINTAENINTA EARLY CURRICURRIE E ARUMBERARUMBERA A CGL FMFM IIII SSSSTT 4 Petroleum Systems of the Amadeus Basin II NAMATJIRA FM PETERMANN RANGES OROGENY I RANGES This paper follows the petroleum systems framework of JULIEJULIE FM FM MAURICE FM FM Marshall (2003) for the Amadeus Basin (see Figure 2). Given the paucity of data in the basin (35 exploration wells WINNALL BEDS BEDS 2 PERTATATAKA FM FM over 170 000 km or 65 500 square miles), it is currently SIRSIR ELLISELLIS 3 FREDERICKFREDERICK SSSSTT CGL very difficult to develop a sequence stratigraphy-style system framework, although it is currently being attempted SOUTHS RANGE RANGE MOVEMENTS MOVEMENTS

C PIONEER SS SST T ? (Dyson in prep). ? ? ? OLYMPIC FM FM Marshall (2003) defined 5 petroleum systems in the LATE CARNEGIE FM FM ININDIA BEDS BEDS Amadeus Basin, numbered 1 (oldest: Neoproterozoic) to 5 ARALKA FM FM (youngest: Ordovician, see Figure 2). Currently, the only BOORD 2 PROTEROZOI PROTEROZOIC AREYONGA FM FM FM commercially productive petroleum system is number 5

AREYONGANAREYONGAN MOVEMENT (Marshall 2003). This Ordovician system encompasses

JOHNNY’SJOHNNYS CK CREEKBEDS BEDS the Mereenie, Palm Valley and West Walker hydrocarbon MEMBER LOVESLOVES CREEK CREEK MEMBER occurrences and is responsible for the accumulations of PINYINNA BED BEDSS oil and gas at both the Mereenie and Palm Valley fields. GILLEN MEMBER

BITTER SPRINGS FM BITTER SPRINGS FM Sub-economic Neoproterozoic discoveries of hydrocarbons

HEAVITREEHEAVITREE QUARTZITE QTE (gas) have been made within system 4 (Orange-1 and -2, DEAN QTE QUARTZITE 1 Dingo-1 to -5), system 2 (Ooraminna-1) and system 1 ARUNTA OROGEN OROGENY Y (Magee-1). Despite numerous oil shows, the general view of ? MIDDLMIDDLEE ARUNTA COMPLE COMPLEX X A06-139 .ai the exploration community has been that anything older than Figure 2. Generalised stratigraphic column for the Amadeus Basin Ordovician would be a gas-prone target only. (modified after Weste 1990), showing petroleum systems defined Previous workers (eg Kurylowicz 1976, McKirdy 1977, by Marshall (2004). Oterdoom 1982, Gorter 1982, 1983, Marty et al 1982,

 Petroleum systems in the Amadeus Basin: Were they all oil prone?

Jackson et al 1984, Summons and Powell 1991), generally Finke-1 was drilled to a total depth of 509.3 m during considered the source-rock potential of the Neoproterozoic April and May 1983. It was drilled primarily to assess the succession in the Amadeus Basin to be negligible to nil and hydrocarbon potential of the Arumbera Sandstone and test predominantly gas prone, largely based on limited, and in the pre-Arumbera section (Gorter 1984), and was sited on the some cases, poor-quality data. However, Marshall (2003, maximum gas seepage anomaly detected by a Vaporsearch 2004) demonstrated that there was source potential within this gas sniffer survey of the James Ranges, conducted in part of the succession, by including more recent data gathered July 1981. The well tested the validity of the Vaporsearch by the Northern Territory Geological Survey (NTGS), and technique, as well as an independent closure west of the using the same style of analysis (geochemical plots) as used dry James Range-1A well to the east (Gorter 1984). No gas by previous authors. This conclusion was based on newer seepage anomaly was associated with James Range-1A. geochemical results (effectively indicating some source units Finke-1 penetrated 214 m of Cambrian shale and carbonate were ‘type 2’ (oil and gas), not just ‘type 3’ (gas) source with minor sandstone, before encountering a very much rocks), laced with analogies to the productive west Siberian reduced clastic succession, tentatively assigned to the Arumbera basins. Essentially, the source-rock characteristics that are Sandstone (Gorter 1984). No hydrocarbons were noted as measured today are only remnants of the original potential occurring in this section in the well completion report. of the intervals; they have been severely degraded due to the Below the 24 m of ?Arumbera Sandstone, the well production and expulsion of hydrocarbons, particularly their penetrated 271.3 m of Late Neoproterozoic dolostone and total organic carbon values (TOC). In the nearby Officer Basin, shale, and reached its total depth in Bitter Springs Formation immature sediments of the same age (?Neoproterozoic), in dolostone (Loves Creek Member). Minor shows of live oil the NJD-1 drillhole, yielded TOC values in excess of 6% were found in non-effective vuggy porosity in the upper (Hocking 2002). This represents a legitimate Australian dolomitic/shale section, but where effective porosity was analogy for the potential source characteristics of sediments encountered, it proved to contain slightly brackish water, in the Neoproterozoic Amadeus Basin. possibly indicative of flushing (Gorter 1984,Table 1). The well penetrated two units that contained live oil Finke-1: a Neoproterozoic oil discovery bleeds within Neoproterozoic sedimentary rocks (Gorter in the Amadeus BASIN? 1984). Geochemical aspects of the recovered oils were detailed in Gorter (1984); in brief terms, the analytical work The Finke Prospect is located 100 km southwest of Alice confirms the former presence of two different oil families Springs (Figure 3). The nearest wells to the prospect are and/or charges. A full description and discussion of the James Range-1A, 7.5 km to the east, Palm Valley-1, 25 km geochemistry of these oils is in Gorter (1984) and Marshall to the northwest and Highway-1, 55 km to the southeast. (in prep). Prior to this study, data (from visual inspection