Thermal profiles inferred from fluid inclusion and illite geothermometry from sandstones of the Athabasca basin: Implications for fluid flow and unconformity-related uranium mineralization

Highlights

• Higher-than-normal paleo-temperatures recorded in the Athabasca Basin

• Lack of obvious thermal gradients suggesting fluid convection

• Convection important for ground preparation for U mineralization

Abstract

The Proterozoic Athabasca basin and underlying basement host numerous unconformity-related uranium deposits that were formed from extensive fluid circulation near the basement-cover interface. Although it is generally agreed that the mineralizing fluids were basinal brines, it is still unclear what driving forces were responsible for the circulation of the basinal fluids. Because different fluid flow driving forces are associated with different thermal profiles, knowing the basin-scale distribution of paleo-fluid temperatures can help constrain the fluid flow mechanism. This study uses fluid inclusions entrapped in quartz overgrowths and authigenic illite in sandstones from three drill cores (WC-79-1, BL-08-01, and DV10-001) in the central part of the Athabasca basin as thermal indicators of paleo-fluids in the basin. A total of 342 fluid inclusions in quartz overgrowths were studied for microthermometry. The homogenization temperatures (T_h) range from 50° to 235 °C, recording the minimum temperatures in various diagenetic stages. Temperatures estimated from illite geothermometry (121 points) range from 212° to 298 °C, which are systematically higher than (partly overlapping) the T_h values, suggesting that illite was precipitated in hotter fluids following the formation of quartz overgrowths. Neither the fluid inclusion T_h values nor the illite temperatures show systematic increase with depth in individual drill cores. This, together with the high illite temperatures that cannot be explained by burial at a normal geothermal gradient (35 °C/km), is interpreted to indicate that basin-scale fluid convection took place during the diagenetic history of the basin. Prolonged fluid convection is inferred to be responsible for delivering uranium (extracted from the basin or the upper part of the basement) to the unconformity, where uranium mineralization took place due to redox reactions associated with fluid-rock interaction or structurally controlled fluid mixing.

Introduction

Basin-scale fluid flow in sedimentary basins played a significant role in the formation of many mineral and petroleum deposits (Bethke and Marshak, 1990, Garven and Raffensperger, 1997, Cathles and Adams, 2005). The mechanism of such fluid flow, for example that responsible for the formation of Mississippi Valley-type (MVT) Zn–Pb deposits, has been a subject of scientific debates for over three decades, one of the focuses being whether or not a given fluid flow model can explain the heat anomaly observed in the deposits (Cathles and Smith, 1983, Anderson and Macqueen, 1988, Bethke and Marshak, 1990, Garven et al., 1993, Garven and Raffensperger, 1997, Cathles and Adams, 2005). This is understandable because fluid flow is always associated with heat transport, and different fluid flow mechanisms may result in different thermal profiles (Duddy et al., 1994, Deming, 1994, Jessop and Majorowicz, 1994, Phillips, 2009, Ingebritsen and Appold, 2012, Chi, 2015). Thus, the study of thermal profiles in sedimentary basins is important for constraining fluid flow models. Fluid flow mechanisms related to the formation of sedimentary basin-hosted (especially unconformity-related) uranium deposits have also been extensively studied (Sanford, 1992, Raffensperger and Garven, 1995a, Raffensperger and Garven, 1995b, Chi et al., 2011, Chi et al., 2013, Chi et al., 2014, Cui et al., 2010, Cui et al., 2012a, Cui et al., 2012b, Chi and Xue, 2014), but so far little attention has been paid to the thermal effects of the fluid flow as was done for the MVT deposits.

The unconformity-related uranium (URU) deposits, which are best developed in Proterozoic basins in northern Canada and northern Australia, especially the Athabasca basin in northern Saskatchewan (Canada; Fig. 1a), represent the richest uranium deposits in the world (Jefferson et al., 2007, Fayek, 2013). The formation of these deposits has been related to circulation of large amounts of basinal fluids, facilitated by high permeabilities due to dominance of sandstones in the basins (Hoeve and Sibbald, 1978, Hoeve and Quirt, 1984, Wilson and Kyser, 1987, Kotzer and Kyser, 1990, Hiatt and Kyser, 2000, Cuney et al., 2003, Kyser et al., 2000, Richard et al., 2011, Richard et al., 2014, Mercadier et al., 2012). However, the driving forces controlling fluid flow are still controversial. Large-scale convection related to a normal geothermal gradient was proposed by Hoeve and Sibbald (1978) and Boiron et al. (2010) as the main mechanism responsible for fluid flow related to URU mineralization in the Athabasca basin, and was shown to be plausible using numerical modeling (Raffensperger and Garven, 1995a, Raffensperger and Garven, 1995b, Cui et al., 2010, Cui et al., 2012a, Cui et al., 2012b). Topography-driven fluid flow was implied in some schematic models (Derome et al., 2005, Hiatt and Kyser, 2007, Boiron et al., 2010) and proposed to be responsible for URU mineralization in the Athabasca basin (Alexandre and Kyser, 2012). Both thermal convection and topography-driven flow are consistent with the near-hydrostatic fluid pressure regime in the Athabasca basin, as demonstrated by numerical modeling (Chi et al., 2013, Chi et al., 2014). Compaction-driven fluid flow was implied in hydrostratigraphic studies of sandstones of the Athabasca basin (Hiatt and Kyser, 2007), but it has been shown that such fluid flow was too slow to result in any significant thermal disturbance in the basin (Chi et al., 2013, Chi et al., 2014). Furthermore, based on the observation that most unconformity-related uranium deposits are spatially associated with faults crosscutting the unconformity, it was demonstrated that fluid flow related to uranium mineralization may be related to deformation along fault zones (Cui et al., 2012a), and mixed or alternative convection and deformation-driven fluid flow models have been advocated (Hoeve and Quirt, 1984, Hoeve and Quirt, 1987, Raffensperger and Garven, 1995b, Cui et al., 2012a, Li et al., 2015). The uncertainties on fluid flow models related to URU mineralization are in part related to the poor understanding of thermal profiles related to fluid flow either at the basin scale or the deposit scale.

Various temperatures (ranging from 60° to 250 °C) have been estimated from fluid inclusions in quartz overgrowths, quartz veins and euhedral quartz for the basinal fluids and the mineralizing fluids in the Athabasca basin in previous studies (Pagel, 1975, Pagel et al., 1980, Kotzer and Kyser, 1995, Derome et al., 2005). However, most of these studies were focused on areas of mineralization, and the background temperatures of the basin, the ambient temperatures in the host rocks near the sites of mineralization and the temperatures of the actual mineralization fluids were not well distinguished. Ore-forming fluids were inferred to be hotter than the background diagenetic fluids in some studies (e.g., Kotzer and Kyser, 1995), and cooler in other studies (e.g., Derome et al., 2005). Few studies were carried out about the background thermal regime of the Athabasca basin. Pagel (1975) studied fluid inclusions entrapped in quartz overgrowths in sandstones from two long drill cores penetrating a major part of the stratigraphy of the Athabasca basin, and estimated that the burial temperature reached 180°C at the base of the Rumpel Lake drill core near the central part of the basin (Fig. 1b), and 220 °C at the base of and the MP-73-183 core near the Carswell structure in the central-west part of the basin. He further deduced a thermal gradient of 35 °C/km based on the assumption of a lithostatic pressure regime, a maximum burial depth of 4.8 km at the site of the Caswell structure, and a thickness of 3.2 km of eroded strata above the top of the Rumpel Lake drill core. Scott and Chi (2014) studied fluid inclusions in quartz overgrowths and illite in the sandstones from the Rumpel Lake drill core, and obtained higher maximum temperatures than those reported by Pagel (1975), mostly from 200° to 250 °C. Furthermore, they did not find the systematic increase of temperature with depth as inferred by Pagel (1975). Maximum burial temperatures of ~ 160° to 200 °C were obtained through a vitrinite reflectance study of the Douglas Formation near the top of the current Athabasca basin by Stasiuk et al. (2001), which are also hotter than those predicted by Pagel (1975)'s study (< 130 °C).

In view of the controversies on basinal fluid flow models and discrepancies on burial temperatures and geothermal gradients in the Athabasca basin, a study of basin-wide distribution of paleo-temperatures both in horizontal and vertical directions is required. Such a study is best carried out in barren areas away from mineralization, so that the results represent the background thermal regime of the basin and can be used as a reference for comparison with the temperatures of the hydrothermal fluids in the mineralization areas. In this study, we collected 164 samples from three drill holes (WC-79-1, BL-08-01, and DV10-001) in the central part of the basin (Fig. 1b), which are widely spaced, far away from known mineralization and penetrate into the basement, and used fluid inclusion (in quartz overgrowths) microthermometry and illite (in interstitial space) geothermometry to characterize the paleo-temperatures during the diagenetic evolution of the sandstones. The results were used to construct vertical thermal profiles at different localities of the basin, to compare the thermal profiles from different localities, and to infer the fluid flow mechanisms that may explain the observed thermal profiles. Furthermore, the significance of the results for URU mineralization, in terms of fluid flow and metal extraction from source rocks, was explored.