|Résumé||(disponible en anglais seulement)|
The Mackenzie Valley of Canada's Northwest Territories is a lightly explored, northern extension of the heavily developed, petroleum-rich Western Canada
Sedimentary Basin. Exploration costs are high in this remote area because of the lack of infrastructure, short drilling season, and land accessibility issues. Petroleum has been encountered in Paleozoic and Mesozoic reservoirs but, in general,
results have been disappointing, in part due to the complicated and poorly understood thermal history of the area. This region has experienced multiple phases of burial, exhumation, and deformation (further complicated by salt mobilization) during
the Phanerozoic that resulted in the development of multiple and widespread unconformities and substantial loss of the stratigraphic record. Recent exploration successes (2005 Summit Creek B-44 Devonian gas/condensate and 2006 Stewart D-57 Cretaceous
gas discoveries) and subsequent failures have underscored the need to constrain aspects critical to a functioning petroleum system such as the timing of petroleum generation relative to trap formation. As part of this process, Issler et al. (2005)
undertook an integrated thermal history study of the East MacKay I-77 well located approximately 80 km southeast of the Norman Wells oil field (one of Canada's largest and longest producing fields and the only producing pool in the area).
MacKay I-77 is located in the Keele Tectonic Zone (KTZ), a north-trending zone that experienced four major phases of subsidence and erosion (> 1 km) throughout the Phanerozoic that were amplified with respect to sections preserved in the bounding
Mackenzie Mountains to the west, and the Franklin Mountains to the east (MacLean and Cook, 1999). The well penetrated 2400 m of Tertiary through Upper Cambrian section and crossed two major unconformities (Cambrian-Lower Devonian; Upper
Devonian-Upper Cretaceous), with the uppermost unconformity marked by a significant increase in vitrinite reflectance with depth (0.58 to 0.76%Ro). East MacKay is drilled on a complicated structure involving pre-Cretaceous thrusting of the Cambrian
through Devonian section. Seismic data indicate that the localization of these structures may be related to Early Cambrian normal faults disrupting Neoproterozoic rocks that underlie the KTZ.
The high cost of field work, limited number of well
penetrations, and sample size requirements for apatite fission track (AFT) analysis have limited the pace at which regional thermal history studies can proceed in this area. Where possible, samples are being obtained directly from companies currently
engaged in field and drilling operations. Drill cores are rarely available, requiring reliance on well cuttings samples, supplemented by outcrop samples where possible. The use of drill cuttings leaves open the possibility of sample contamination.
The Issler et al. (2005) study used state-of-the-art knowledge of the AFT system to define two kinetic populations of apatite within a Devonian sandstone sample (Imperial Formation) from the East MacKay I-77 well on the basis of Cl content (see
Figure 1a) and achieved a thermal history that is consistent with available geological constraints. Other kinetic parameters such as rmro and Dpar appeared to yield poor results, with substantial overlap of younger and older single grain
ages with respect to each parameter. Given that Carlson et al. (1999) had serious doubts concerning the general application of their empirically calibrated rmro equation, and that variable factors can influence AFT etching characteristics,
it appeared that the widely used Cl parameter was successful at defining the two kinetic populations. However, subsequent analysis of a large suite of probe data for AFT core and outcrop samples from western Newfoundland, the Beaufort-Mackenzie Basin
and the Western Canada Sedimentary Basin (unpublished data) suggests that the rmro parameter (as defined by Carlson et al., 1999) is the best and most consistent method for defining AFT kinetic populations compared with using only Cl
content or Dpar. Such kinetic populations commonly show a good correspondence to age populations defined using the binomial-peak fitting program, Binomfit (Brandon, 2002).
Given that rmro works so well for apatite with
a broad range of compositions from different basins (including compositions similar to those of the East MacKay sample), the alternative interpretation is that the East MacKay sample set is severely contaminated (70% AFT age grains from Cretaceous
sediments). In hindsight, some level of contamination is consistent with the sample processing history and was expected: initial processing yielded a coarser apatite fraction with dominantly Late Cretaceous AFT ages (now interpreted to be from
Cretaceous cuttings) whereas subsequent sample re-crushing produced abundant silt-sized grains with Jurassic to late Paleozoic AFT ages (now interpreted to be in situ from siltstones and fine-grained sandstones from the Devonian Imperial Formation).
Furthermore, Rock-Eval parameters indicate severe contamination of the upper part of the Imperial Formation by organic-rich shale of the overlying Cretaceous Slater River Formation, with substantially decreased contamination toward the base of the
unit where the AFT sample was collected. Finally, discussions with persons involved with the drilling indicate that contamination was more serious than if it was just the result of borehole caving. Cretaceous drill cuttings were actually recirculated
back into the mud system due to problems with the mud filters (A. Stirrett, pers. comm., 2006) and used to increase mud density, resulting in extensive sample contamination over the Imperial Formation interval that is not evident in the well history
report. In order to confirm the degree of sample contamination and attempt to ameliorate its consequences, cuttings samples were collected for AFT analysis from the overlying Upper Cretaceous Little Bear and Slater River formations, and the upper
part of the Imperial Formation. AFT results for these samples confirm the interpretation that Cretaceous apatite grains are contaminating the Imperial Formation AFT sample and support the interpretation of Figure 1b. In Figure 1b, rmro is
recalculated in terms of effective Cl content and it is apparent that many of the Cretaceous apatite grains have similar annealing kinetics to those of the higher Cl Devonian grains due to increased amounts of other cations (mainly Fe, Na and
Figure 2 shows the distribution of AFT grain ages with respect to the effective Cl content for the Little Bear (approximately 990 m depth) and Slater River (approximately 1290 m depth) formations. The Campanian (~83 Ma) Little Bear sample
shows two AFT populations based on Binomfit analysis: a minor population of older grains interpreted as detrital (derived from Albian or older units) and a younger dominant population interpreted to be of volcanic-type (rapidly exhumed or
contemporaneous volcanism). The Turonian (~90 Ma) Slater River sample shows a single AFT population interpreted to be of volcanic-type that is slightly younger than the Little Bear sample due to increased AFT annealing. Both samples are slightly
younger than their stratigraphic ages and successful models that are compatible with thermal maturity and kinetic parameter values require initial AFT ages for the volcanic-type populations to be zero at the time of deposition. Comparison of Figure 2
with Figure 1b shows that the suspected contaminant Cretaceous apatite grains in the Imperial sample are very similar to those in the overlying Cretaceous stratigraphic units. As further evidence of contamination, another sample collected in the
upper part of the Imperial Formation (approximately 1615 m depth) yielded 13 out of 14 grains with AFT ages and effective Cl values very similar to those of the Slater River sample, implying more than 90% contamination. Fortunately, the effects of
contamination can be minimized by removing the lower Cl content Cretaceous grains from the analysis and modelling the remaining population of high Cl, silt-sized Devonian apatite grains. A plot of AFT grain ages versus Dpar for the Imperial Formation
sample is similar to Figure 1b with both Cretaceous and Devonian apatite grains having similar Dpar values. Without probe data, a single parameter such as Dpar cannot be used to try and remove the effects of contamination if different age populations
overlap in Dpar space.
The following conclusions can be derived from this study:
1. It can be misleading to rely on Cl content alone as a parameter for characterizing different kinetic populations for AFT annealing, even if
seems to produce good results. Apparently good model results may yield incorrect thermal histories because the effect of cation substitutions in apatite has been ignored. For the case of samples from western and northern Canadian sedimentary basins,
cation substitutions are common and can cause track retentivity to be higher than that inferred from Cl content alone and this supports the use of the rmro parameter.
2. With full suite probe data, it may be possible to correct for the
effects of sample contamination and extract useful information from samples where no other material is available. Such is the case for the Imperial Formation sample where in situ grains have a higher Cl content than caved Cretaceous grains.
Cretaceous sediments of the central Mackenzie Valley appear to be dominated by volcanic-type apatite that does not carry a record of the pre-depositional thermal history. Such samples may give improved thermal history resolution because they were
deposited as a uniform age population and any subsequent deviations from uniformity during burial heating may be attributed to variation in annealing kinetics.