Des galènes ont été prélevées (i) au sein d'un échantillon minéralisé dont les oxydes d'U ont été datés à 1281±6 Ma [7], et (ii) dans des fractures non minéralisées à 50–80 m au-dessus du corps minéralisé. Les galènes (i) sont extrêmement uranogènes (Tableau 1), avec un âge maximal de dépôt de la source uranifère de Pb (983±46 Ma) nettement plus jeune que l'âge des oxydes d'uranium du même échantillon. Ces galènes montrent que le plomb uranogène peut être piégé dans la zone minéralisée, mais qu'il peut migrer sur une plus longue distance que le voisinage immédiat de l'oxyde d'uranium père. Les galènes (ii) sont moins radiogéniques (Tableau 1), avec une contribution thorigène non négligeable. Ces galènes témoignent d'une redistribution de plomb, par suite du lessivage de minéraux riches en Th de la base du bassin ou des granitoı̈des du socle [4,5] et, dans une moindre mesure, d'oxydes d'uranium.

Les teneurs en U et Pb ont été analysées sur 2943 échantillons prélevés le long des forages d'exploration (Fig. 1, Tableau 1). Cette base de données a permis de construire des modèles 3D de distribution des teneurs en U et Pb (Fig. 2) et des rapports Pb/U (Figs. 3 et 4) à l'aide du logiciel GOCAD [17], dans et autour de la zone minéralisée, et de calculer le rapport Pb/U moyen du corps minéralisé (équations (1) et (2)) : 0,071±0,015. Ce rapport correspond à celui qu'aurait eu un gisement formé en un épisode il y a 485±97 Ma, en supposant, de plus, que l'essentiel du plomb est d'origine uranogène et qu'il n'y a eu de perte, ni en plomb, ni en uranium (équation (4)).

Les plus fortes teneurs en U et Pb sont localisées le long de la discordance (au maximum 50 m au-dessus) et localement dans le socle (Fig. 2b). De plus faibles anomalies Pb dans les formations supérieures des grès, ne correspondant pas à des anomalies en U, semblent indiquer des chemins de migration de Pb, probablement le long de structures intersectant la stratigraphie.

Excepté pour les faibles teneurs en U et Pb (<10 ppm, groupe 1, Fig. 3a), Pb est globalement corrélé avec U, en accord avec une origine uranogène du plomb. Cette corrélation n'est cependant pas linéaire. Les échantillons faiblement minéralisés (U<500 ppm, groupe 2, Fig. 3a), localisés principalement à l'aplomb du corps minéralisé (Fig. 4), ont des rapports Pb/U supérieurs au rapport théorique de 0,211 (équation (4), en considérant un âge de 1315 Ma). En notant que les âges les plus anciens enregistrés à Shea Creek sont d'environ 1315 Ma [7], des rapports supérieurs à 0,211 indiquent un gain en plomb uranogène, et/ou une perte en U, et/ou la présence de plomb commun ou thorigène.

À l'inverse, les échantillons fortement minéralisés ont des rapports inférieurs à 0,211 (Fig. 3a). Les plus faibles rapports Pb/U sont notamment distribués dans la minéralisation massive le long de la discordance (Fig. 4), suggérant que, dans cette zone, U a été déposé plus récemment et/ou qu'une grande quantité de plomb radiogénique y a été lessivée.

Dans le bassin Athabasca et le gisement de Shea Creek, en particulier, les oxydes d'uranium ont perdu une quantité importante de plomb radiogénique, comme le montrent les âges isotopiques UPb, souvent discordants [1,2,6,7,9,10,15,18,23].

Ce plomb radiogénique a été au moins en partie redéposé sous forme de galène dans la zone minéralisée, puisque certaines galènes (type (i)) ont des compositions isotopiques extrêmement uranogènes.

Pour quantifier la proportion de plomb radiogénique éventuellement perdue en dehors du gisement, nous émettrons tout d'abord les hypothèses (1) que l'essentiel du plomb est uranogène, (2) qu'il n'y a pas eu de perte en plomb à l'échelle du corps minéralisé depuis le ou les dépôts uranifères.

L'hypothèse (1) est vérifiée par l'existence d'une corrélation entre les teneurs en U et Pb (Fig. 3), et par les compositions isotopiques UPb des oxydes d'uranium [6,7,9,10,18] et des galènes de type (i).

Avec l'hypothèse (2), le gisement de Shea Creek peut être vu comme un système fermé, dans lequel l'uranium aurait été apporté en un ou plusieurs épisodes et tout le plomb radiogénique aurait été conservé. Dans ce cadre, le rapport Pb/U de 0,0711 indique qu'au moins la moitié du gisement a dû s'être formée au cours d'événements plus récents que 485 Ma (équation (5)). Ce résultat est même sous-estimé, puisque du plomb commun et du plomb thorigène sont aussi présents, comme le montre la composition isotopique des galènes de type (ii).

Ce résultat est improbable, puisque d'après les études géochronologiques antérieures dans l'Athabasca [1,2,6,9,10,15,18,23], les gisements d'uranium se sont essentiellement formés au Méso-Protérozoı̈que. Les âges plus jeunes représentent, soit des dépôts mineurs de pechblende [6,18], soit des âges de remobilisation d'anciennes uraninites [1,6,14,15].

Par conséquent, l'hypothèse d'une rétention totale de plomb radiogénique dans le corps minéralisé apparaı̂t trop restrictive. Seule une partie du plomb y a été piégée, mais l'essentiel (environ 60 %) a migré à plus de 700 m du corps minéralisé.

Le plomb a préférentiellement migré le long de la discordance où sont observées les minéralisations les plus riches avec les plus faibles rapports Pb/U (Fig. 4). La discordance constitue en effet une zone de forte perméabilité, le long de laquelle les fluides tardifs ont pu facilement circuler [12,15].

PbPb isotopic studies have been performed on six galena crystals at IGEM (Moscow) on a multicollector mass-spectrometer Sector 54. Three samples have been selected immediately above the Anne uranium orebody in the MFc formation, in non-mineralised (U<5 ppm) vertical fractures filled with pyrite and galena. Three other galena crystals have been selected, within the uranium orebody, from a vein crosscutting some U oxides that have been dated by [7] (concordant UPb upper intercept at 1281±6 Ma and lower intercept at 53±18 Ma with a MSWD of 19 on a Concordia diagram).

105 vertical drill holes and 54 directional drill holes diverging from some of the vertical drill holes have been completed by COGEMA along the SLC in the Anne mineralised area. The vertical boreholes have been drilled down to more than 800 m through the Athabasca sandstone to the basement, whereas directional drill holes were restricted to the ore zone close to the unconformity. The drilling pattern had a 20–25 m spacing along a 300 m strike length and 50 to 70 m perpendicularly (Fig. 1). One barren drill hole, Erica 1, has been used as a reference for background U and Pb contents, because it is located about 5 km away from the SLC and it does not intersect any graphite-rich structure. In each drill hole, composite samples have been taken every 10 m in the sandstone and in the basement, except in the ore zones where the sampling interval was in the range 10–50 cm. Two thousand nine hundred and forty-three samples were analysed at the Saskatchewan Research Council by ICP–OES for Pb following a tri-acid total digestion (HFHNO₃HClO₄). U was analysed by fluorimetry.

The 3D distribution of U, Pb and Pb/U was modelled with the GOCAD software [17]. The values were interpolated in a 3D grid by the Discrete Smooth Interpolation method [17] within a 750×250×750 m volume with grid cell dimensions of 20×30×10 m in the Anne mineralised area, decreased to 20×30×1 m in the orebody. In this paper are shown vertical sections across these 3D models: (i) along Erica 1 drill hole (Fig. 2a), (ii) in the Anne mineralised area through the sandstone cover, the ore zone and the basement (Fig. 2b), and (iii) zoomed on the orebody (Fig. 2c).

The average Pb/U ratio at the scale of the orebody (Figs. 3 and 4) was calculated from the following formula:

The density of a 10⁶ gram cell was calculated taking into account the U and Pb contents:

The estimated error was calculated as follow:

The amount of lead produced from the radioactive decay of U, $m_{{\mathrm{Pb}}^{*}}$, for a U deposition at time t, can be calculated from the formula of [3], expressed here in mass ratio:

Hence, for a single episode of U deposition at 1315 Ma, $m_{{\mathrm{Pb}}^{*}}=0.211{\mathrm{m}}_{\mathrm{U}}$. Consequently, a rock that has remained a closed system with respect to U and Pb and that contains negligible amount of common and thorogenic Pb should have a present-day m_Pb/m_U ratio of 0.211. Higher ratio would indicate uranogenic Pb gain and/or U loss and/or the presence of non-uranogenic Pb. Inversely, lower ratio would indicate uranogenic Pb loss and/or U gain.

If several events of U deposition occurred, then

The $m_{{\mathrm{Pb}}^{*}}/{\mathrm{m}}_{\mathrm{U}}$ ratios will be considered as ‘theoretical ratios’ that will be used as references to estimate the Pb and/or U gain and/or loss.

Considering the ²⁰⁶Pb/²⁰⁴Pb, ²⁰⁷Pb/²⁰⁴Pb, and²⁰⁸Pb/²⁰⁴Pb ratios for common Pb at 1350 Ma from the model of [22] (16.43, 15.43 and 36.09, respectively), the three galenas from the non-mineralised fractures above the Anne orebody (Table 1) are partly uranogenic, but common and thorogenic lead are also present. The least square regression of the ²⁰⁷Pb/²⁰⁴Pb vs ²⁰⁶Pb/²⁰⁴Pb ratios yields an age of 1391±430 Ma (MSWD=0.32). This age would correspond to the oldest deposition age of the uranium source, if galena crystals were formed at present time, which is however not proved. The deposition age of the source would be younger for an older age of galena crystallisation.

Conversely, the three galenas collected from the vein crosscutting some U oxides (Table 1) are extremely uranogenic. For these galenas, the deposition age of the uranium source cannot be older than 983±46 Ma (MSWD=0.19).

U and Pb contents in barren areas are below 2 ppm and 5 ppm in the sandstones, respectively, and below 6 ppm and 44 ppm in the basement, respectively (Erica 1, Table 2, Fig. 2a). These contents will be used as thresholds values to define the U- and Pb-enriched areas in the distribution models (Fig. 2).

The U-richest areas are located along the unconformity in the Anne area, in the 30 lower metres of the MFc formation and down to 70 m into the basement (U contents systematically above 10 ppm and locally up to a few tens of wt%, Fig. 2b and c, Table 2). The Pb-richest contents globally follow a similar distribution (Fig. 2b and c), suggesting that Pb is of uranogenic origin and/or that Pb was also added during the mineralising events.

In the other sandstone formations (MFd, LzL, WPa, WPb, LL), U contents only very locally exceed 2 ppm (up to 4.4 ppm, Fig. 2b, Table 2), but these local anomalies are not supported by Pb enrichment. The Pb anomalous zones appear more widely developed higher in the basin, especially in the WP formations (up to 17.7 ppm). The relative distribution of these U and Pb anomalies may suggest that some uranogenic Pb might have migrated from the U-rich zone upward, probably along some fractures crosscutting the lithostratigraphy.

Except for samples with low U and Pb contents (<10 ppm, group 1, Fig. 3a), Pb is globally positively correlated with U. This correlation is however not linear, as the Pb/U ratios tend to decrease with increasing U-grade.

Most samples with less than 500 ppm U (group 2, Fig. 3a) display Pb/U ratios higher than 0.211, which is the theoretical ratio for a U deposition at 1315 Ma. Those samples are essentially located on the margins of the orebody, especially toward the sandstone, as shown on the Pb/U spatial distribution model (Fig. 4). The ∼1315 Ma age being approximately the oldest event recorded at Shea Creek [7], the high Pb/U ratios observed in this zone might reflect radiogenic Pb gain and/or U loss and/or the presence of non-uranogenic Pb minerals.

Samples richer in U (U>500 ppm) have Pb/U ratios lower than 0.211 (Fig. 3a), suggesting either increasing radiogenic Pb loss and/or increasing proportion of more recently deposited U oxides, with increasing ore grade. The high-grade samples plot themselves in two groups: group 3 with Pb/U ratios between 0.089 and 0.138, and group 4 with Pb/U ratios between 0.021 and 0.028 (Fig. 3b). Group-3 samples are located in the basement, while group-2 samples are located in the massive ore zone along the unconformity, where the Pb/U ratios are the lowest (Fig. 4). The particularly low Pb/U ratios distributed in this zone indicate that the U ore was deposited more recently or that radiogenic lead has been more massively leached along the unconformity.

The average Pb/U ratio calculated at the orebody scale (Fig. 4) is 0.071±0.015. It corresponds to the theoretical ratio expected for a U orebody formed 485±97 Ma ago, containing only uranogenic lead and that has remained closed relative to U and Pb.

Uranium oxides in the Athabasca Basin, and in the Shea Creek deposit in particular, experienced significant episodic Pb loss [1,6,7,9,10,15,18,23], as evidenced from the numerous discordant data in Concordia diagrams, and the frequent non-zero lower intercepts ranging from 700 to 200 Ma.

Radiogenic Pb has however been trapped within the orebody, at least partly, as shown by the extremely uranogenic Pb composition of the three galenas collected in the mineralised sample. Their compositions compare well with that of the galenas from the mineralised zone of the Midwest deposit [5]. However, the oldest possible deposition age of the uranium source (983±46 Ma) does not correspond to the age of the U oxides present in the same thin section (1281±6 Ma, [7]). This indicates that the source was external to the sample, which highlights the possibility of ‘long-distance’ Pb migration within the orebody, i.e., Pb is not systematically redeposited in the immediate vicinity of the U oxides. Such Pb migration has already been observed at the scale of a hand sample in the Key Lake deposit [23], by comparing the UPb isotopic composition of uraniferous phases with that of their corresponding matrix after mechanical separation of the two components.

These data however do not indicate whether all radiogenic lead was retained within the ore zone or not. The anomalous Pb contents observed upper in the sandstone (Fig. 2b) and the increase of the Pb/U ratios toward the margins of the orebody (Fig. 4), where some galenas have also been observed, may suggest a Pb migration upward. However the non-negligible amount of common and thorogenic Pb in those galenas indicates that the source cannot be the U oxides exclusively because the amounts of Th, ²⁰⁸Pb and ²⁰⁴Pb are very low in those minerals (²⁰⁸Pb/²⁰⁴Pb ratios from 38.3 to 43 [7]). Hence, common and thorogenic lead from these galenas may derive either from Th-rich sandstone layers (heavy-mineral-rich layers of the basal MF formations) or from Th-rich high K-calcalkaline granitoids in the basement [4]. Such Pb isotopic compositions correspond to those of the ‘basement-type’ galena around the Midwest orebody [5], which have been interpreted as resulting from the alteration of the regolith below the unconformity.

To quantitatively constrain the proportion of lead that may have been lost outside the ore zone, the average Pb/U ratio of 0.071±0.015 will be now discussed with respect to the two following hypothesis: (1) Pb was essentially of uranogenic origin, (2) the orebody has remained closed relative to Pb.

5.2.1 Is Pb essentially of uranogenic origin?

5.2.2 Has the orebody remained closed?

Possible Pb migration pathways may be deduced from the relative 3D distribution of U and Pb. As shown in Fig. 2b, Pb anomalies that are not supported by U anomalies seem to follow some structures extending from the basement up to the WP formations. Such anomalies may result from the migration of uranogenic lead along fractures crosscutting the lithostratigraphy.

At the scale of the orebody (Fig. 4), the increase in the Pb/U ratios from the high-grade ore zone at the unconformity toward the sandstone cover is consistent with a Pb migration upward. However, the galenas selected in the Pb-enriched zone above the orebody are only slightly uranogenic, indicating that the main part of uranogenic lead probably did not migrate upward.

Conversely, the location of the lowest Pb/U ratios along the unconformity, indicates that uranogenic lead has been more massively leached from this zone (Fig. 4). It suggests that Pb preferentially migrated along the unconformity, which is consistent with the presence of highly permeable lithologies at the base of the Athabasca Basin (regolith and microconglomerate). This is also supported by the very radiogenic isotope ratios measured in rocks close to the unconformity, up to several kilometres away from some unconformity-related deposits in the Thelon and Athabasca Basins [12].