Modelling of hydrogen gas generation from overmature organic matter in the Cooper Basin, Australia

Introduction

Hydrogen gas (H₂) is the most energy-rich gas weight-for-weight and is considered crucial for a sustainable clean energy future (International Energy Agency 2022). The production of H₂ from geological sources (geologic H₂) is increasingly recognised as a low-cost complement to manufactured H₂ (Lapi et al. 2022). Although there are numerous sources of geologic H₂ within sedimentary, metamorphic and igneous rocks (Boreham et al. 2021), the most studied and effectively modelled are those reactions associated with fluid–rock interactions e.g. ferrous to ferric iron oxidation accompanied by water reduction and the radiolysis of water to produce H₂. Calculation of H₂ production rates, coupled with a geological understanding, enables estimation of geologic H₂ resource potential. The majority of sources for geologic H₂ are abiotic and volumetric estimates of H₂ generation are mostly qualitative. Biogenic (microbial and thermogenic) sources of H₂ also exist. In this paper, the chemical kinetics of H₂ generation from coals and carbonaceous shales containing land-plant-derived organic matter has been applied within a petroleum system modelling framework to the Cooper Basin to quantify the generation of H₂ in a free state (free H₂).

Cooper Basin petroleum generation

The Cooper Basin is a Pennsylvanian–Middle Triassic intracratonic basin containing predominantly terrigenous sediments, which covers an area of 127 000 km² with sediment thicknesses typically being up to 2500 m, although the base of the section within the deepest trough (Nappamerri) reaches over 4400 m. The overlying Jurassic–Cretaceous Eromanga Basin extends beyond the preserved depositional edge of the Cooper Basin and is over 2500 m thick. The Palaeocene to Quaternary Lake Eyre Basin overlies the Eromanga Basin (Fig. 1a, b; Hall et al. 2019 and references therein) but it is thin (up to 300 m) across the Cooper Basin region.

Fig. 1. 

(a) Areal extent of the Cooper Basin showing the main depocentres, troughs and ridges and the distribution of oil and gas occurrences (note: ‘PSM well’ identifies the location of the well used in the petroleum system model of Hall et al. (2019)) and (b) cross-section showing the distribution of the sedimentary formations of the Cooper and Eromanga basins across the Nappamerri, Patchawarra and Tenappera troughs. Modified after Hall et al. (2019).

Hydrocarbon generating source rocks are dominated by land-plant-derived organic matter with the Toolachee and Patchawarra formations being considered prospective. Both Permian formations collectively host the main commercial reservoirs for gas in the Cooper Basin (Fig. 1b; Hall et al. 2019 and references therein). Petroleum system modelling indicated that the main phase of hydrocarbon generation began in the latest Permian in the Nappamerri Trough depocentre in South Australia while the mid-Cretaceous corresponded to a time when hydrocarbons were generated in the adjacent troughs, including the Patchawarra and Windorah troughs. Source rock maturity varies between depocentres, with large areas of the Nappamerri, Patchawarra and Windorah troughs being gas mature to overmature at present day (Hall et al. 2016).

Modelling free H2 generation

Modelling of gas generation employed the Zetaware software suite (Kinex version 4.94, Genesis Version 5.68b and Trinity T3 version 6.61). Input parameters from Hall et al. (2019) were used, including basin history (burial, thermal and pressure), source characterisation (net thickness, richness and quality) and the primary methane, late methane and free H₂ generation kinetics of Mahlstedt et al. (2022). The maturation history model is shown for Burley 2 (Fig. 2a) because it is a deep well located in the centre of the Nappamerri Trough (Fig. 1a) where the Patchawarra Formation source rocks are overmature (measured vitrinite reflectance (Ro) >2%) with predominantly dry gas generation. The modelled Ro is displayed in Fig. 2a, which also highlights the variation in the vitrinite reflectance kinetic models available in the literature and employed within the Zetaware software suite; for modelling vitrinite reflectance in Fig. 2a, b, this paper uses the ARCO vitrinite reflectance kinetic model (see Falvey and Middleton 1981). Additionally, well histories of Hall et al. (2019) were slightly modified for those wells where measured Ro was >2% to best honour the fit between the modelled and measured Ro profiles over this high-maturity region.

Fig. 2. 

For Burley 2 well (a) modelled and measured vitrinite reflectance (Vitrinite Ro) versus depth, (b) transformation ratio for primary methane, late methane and free H₂, and modelled vitrinite reflectance versus modelled temperature and (c) modelled temperature history versus age.

Utilising primary methane, late methane and free H₂ generation kinetics (Mahlstedt et al. 2022), the cumulative transformation history (i.e. transformation ratio (TR) from 0 to 1) of a source rock positioned in the middle of the Patchawarra Formation in Burley 2 is modelled and displayed in Fig. 2b. Methane generation is exhausted at the hydrocarbon deadline where TR >0.9, which occurs at the time of maximum paleotemperature, in this case ~250°C and a modelled Ro of ~3.5%, in the Late Cretaceous just prior to the Winton Formation uplift and erosion (Fig. 2b, c). Importantly, at this same time, the transformation of the initial free H₂ potential is only ~50% completed (TR ~0.5). Thus, where the Patchawarra Formation source rock experienced higher maturity and temperature, methane-free H₂ would be generated.

Employing a 2D surface of a mid-Patchawarra Formation source rock (coal and carbonaceous shale), the cumulative gas generation (m³ gas/m² rock) for methane (primary + late) and free H₂ is mapped in Fig. 3a, b, respectively. Methane generation is prevalent across much of the Cooper Basin, while free H₂ generation is predominantly restricted to a ‘sweet spot’ in the central Nappamerri Trough. Interestingly, both methane and free H₂ have similar maximum ‘m³ gas/m² rock’ generation rates. With consideration of rock thickness and a conservative H₂ expulsion efficiency of 90%, Mahlstedt et al. (2022) estimated 86 Tcf free H₂ gas is still retained, much more must have expelled from the overmature Patchawarra Formation source rocks.

Fig. 3. 

Areal distribution of modelled cumulative gas generation volumes m³/km² from a mid-Patchawarra Formation source rock for (a) methane (primary + late) and (b) free H₂ (the enclosed black line marks the 50 km boundary from the edge of the H₂ ‘sweet spot’. Note: initial primary methane potential = 40 mg/g TOC (total organic content), initial late methane potential = 50 mg/g TOC, initial free H₂ potential = 20 mg/g TOC (Mahlstedt et al. 2022)) and (c) well location with reported maximum H₂ mol% where H₂ >0.1 mol% (Supplementary Table S1 lists data from WCRs and five well gases in Boreham et al. (2021). Note: well names for maximum H₂ ≥0.8 mol%; Burley 3 with a maximum H₂ = 0.82 mol% is for pre-well-fire gas whereas post-well-fire gas has a maximum H₂ = 40 mol%). Note: Apply structural elements in Fig. 1 to Fig. 3.

The H2 system

The concentration of H₂ in Australian natural gas is typically low with ~80% of the dataset having H₂ <0.1 mol%. Nevertheless, natural gases with H₂ concentration >1 mol% are found in all Australian states and the Northern Territory (Boreham et al. 2021). In the Cooper Basin, the areal extent of those gases reported in Boreham et al. (2021) shows that the five gases with H₂ >0.1 mol% (Table S1) fall either on or within a 50 km boundary extended from the outline of the H₂ ‘sweet spot’ in the Nappamerri Trough (Fig. 3b). While the migration and fate of H₂ gas is beyond the scope of this paper, if the source of H₂ is from the thermal decomposition of organic matter then this may define the extent of the H₂ system, consistent with the critical elements of a petroleum system defined by Magoon and Dow (1994).

In order to further constrain and have confidence in the areal extent of an inferred H₂ system, well completion reports (WCRs) were consulted for petroleum wells drilled in the Cooper and Eromanga basins. WCRs from 3346 wells were examined with a low proportion (557 wells) containing gas data and a limited number (35 wells and 108 gas analyses) containing H₂ mol% values. Of these, 26 wells have gases with H₂ >0.1 mol% and 9 of these wells having gases with H₂ >1.0 mol% (Fig. 3c) (Table S1). The south-west margin of the Nappamerri Trough hosts H₂-rich (>10 mol% H₂) natural gases in South Australia where migration of H₂ may be up to ~40 km from the edge of the thermogenic H₂ generation ‘sweet spot’ (kitchen) (Fig. 3b, c). Natural gases with much lower H₂ contents typify the northern extent over the Patchawarra Trough in South Australia where source rock maturities are lower. The lack of wells both fewer in number and without H₂ mol% data lends to an uncertain and unconstrained H₂ system in Queensland (Fig. 3c).

Summary and conclusions

The need for a reliable and volumetrically significant supply of H₂ at the right price is a challenge for the future low-carbon economy. Experience gained over more than a century by petroleum and mineral explorationists guarantee the right tools are now available for the exploration for geologic hydrogen. Geologic hydrogen gas has the advantage over petroleum since it has around 10 times more sub-surface sources and hence greater opportunities, compared to sources of petroleum. Thermogenic hydrogen is seen as a viable candidate in the hydrogen ‘mix’ because an organic-rich, highly overmature rock can be modelled to determine where geologic hydrogen is generated, in this case the central Nappamerri Trough in the Cooper Basin, and hence influence future exploration opportunities and activities.

Supplementary material

Supplementary material is available online.

Data availability

Burial history and maturation model for Burley 2 in Zetaware format is available on request to the corresponding author.

Conflicts of interest

The authors declare that no conflicts of interest exist.

Declaration of funding

This study was funded by Geoscience Australia’s Exploring for the Future program. No funding from external organisations was received for this research.