How diagenetic rock composition and mineralogical changes associated with water injection affect geomechanics and fluid behaviour in chalk i.e. producibility (IOR)

Abstract

Geo-mechanical responses associated with seawater injection for improved oil recovery (IOR) processes have been thoroughly investigated during the last few decades. It has been suggested that the presence of non-carbonate minerals will impact brine-rock interaction processes as water weakening chalk. Moreover, mineralogical processes and geo-mechanical responses associated with seawater injection into fractured chalk and how these processes impact aperture modification and fluid flow have not previously been completely understood. The main objectives of the thesis are to characterize impurities in chalk, which always exist to some extent (quartz, dolomite, pyrite, etc.), and to map brine-rock interaction processes that may be linked with impurities. The second objective is focused on how artificially fractured chalk alters properties during brine injection, in terms of water weakening and mineralogical alteration processes. To reach these objectives, a wide range of methods has been used. A toolbox was developed for providing a guide for selecting an adequate collection of methods for characterization of chalk (Paper I).

The newly developed Helium Ion Microscopy combined with Secondary Ion Mass Spectroscopy (HIM-SIMS) was used due to the exceptionally high resolution for imaging and chemical identification in-situ. This was the first time that HIM-SIMS was used for IOR research on chalk (Paper II). The HIM-SIMS showed that it is possible to discriminate noncarbonate minerals from the calcite surface of micron-sized fossils. Calculations of surface mineral distribution in a synthetic seawater (SSW) flooded outcrop sample, indicated that the clay covered 6.3 % in addition to 39.8 % mixed compound of clay and calcite. On the other hand, the bulk geochemistry showed only a concentration of 5 weight % (Wt.%) non-carbonates (Paper III).

Pure dolomite and pure manufactured calcite powder, in addition to three mixes of the two minerals with different ratios were manually compressed to regular cores. Laboratory experiments were performed on these samples using triaxial cells which can mimic reservoir conditions such as temperature and stress while flooding samples with different brines (Paper IV). Flooding powdered samples in mixed ratios of dolomite and pure manufactured calcite showed complex chemical processes, with precipitation of new minerals that were not identified when flooding dolomite and manufactured calcite separately. Precipitation of new minerals was dependent on the composition of the injected brine and the ratio of dolomite versus manufactured calcite. These complex chemical processes were for some samples associated with higher total axial strain than observed for the two minerals flooded separately. The total axial strain was dependent also on the porosity and mineral composition present in the sample. A sample with a realistic composition representing the chalk from the Ekofisk oil field on the Norwegian Continental Shelf (NCS) with 5 Wt.% dolomite injected with SSW, showed dissolution of calcite and precipitation of abundant aragonite crystals with sizes ranging up to 20 μm. On the other hand, pure manufactured calcite and dolomite powder showed no significant chemical alteration when SSW was injected in these samples separately. Despite significant dissolution of calcite and precipitation of aragonite in the mixed compound (5 Wt.% dolomite), the precipitation of aragonite did not impact the rate of compaction in the powder samples.

Five Upper Cretaceous outcrop chalk samples from the Obourg quarry, St. Vaast Formation (OBSV), had been flooded in a previous study under comparable conditions as used for the powder experiments. Three of these samples had an artificial fracture of 2.25 (± 0.05) mm that was drilled before the flooding. Here, characterization of mineralogy, chemical, and textural properties on flooded and unflooded samples facilitated interpretation of changes that took place during the flooding experiments. Altered properties as aperture modification were further associated with the composition of the injected brine, present mineralogy, and geo-mechanical response during flooding (Paper V). Three artificially fractured Upper Cretaceous chalk samples showed lower mechanical resilience than unfractured samples flooded with inert brine in the creep phase (above yield stress). During hydrostatic loading there was no significant difference in strength. The water weakening effect was identified after the brine composition changed to reactive brines such as SSW and magnesium chloride. The artificially fractured samples showed a delayed and less significant water weakening effect than observed for unfractured chalk samples for a test duration of two months. This was associated with brine dominantly flowing through the fracture, thus delaying chemical interactions further away from the fracture. An extrapolation of the creep indicated that fractured samples may experience a higher compaction over a prolonged period of 200 days. All apertures were reduced during the flooding experiment. This was dominantly associated with the total axial strain. Matrix (core material) had been forced into void spaces. Precipitation reduced the aperture additionally. The aperture reduction was controlled by the mineralogy present in the matrix as well as the injected brine composition (Papers V, VI).

The laboratory experiments demonstrated how complex the mineralogical system of chalk can be, by just adding one single mineral such as dolomite to pure calcite. This highlights the necessity to detect and characterize impurities in chalk, even at very low concentrations. The HIM-SIMS results indicated that impurities can be distributed over a significantly larger area than anticipated based on bulk geochemistry. A systematic approach to characterizing impurities in reservoir samples can provide more accurate models for wettability properties, fluid flow and oil recovery. This may improve precision in modelling fluid transportation in the reservoir, as well as a better understanding of rockfluid interactions with an improved understanding of aperture modification and compaction. Details within this study (Figure 1) also demonstrate that mineralogical composition and microscopic analyses are paramount to develop IOR applications.