Mineral replacement in long-term flooded porous carbonate rocksMona Wetrhus Mindea,b,⇑, Udo Zimmermanna,b, Merete Vadla Madlanda,bReidar Inge Korsnesa,b, Bernhard Schulzc, Sabine GilbrichtcaUniversity of Stavanger, P.O. Box 8600 Forus, N-4036 Stavanger, NorwaybThe National IOR Centre of Norway, University of Stavanger, P.O. Box 8600 Forus, N-4036 Stavanger, NorwaycTU Bergakademie Freiberg, Institute of Mineralogy, Brennhausgasse 14, D-09596 Freiberg, Saxony, GermanyReceived 21 August 2018; accepted in revised form 9 September 2019; Available online 18 September 2019AbstractThis study reports mineralogical and physical property changes linked to geo-chemical alterations processes during threeultra-long-term tri-axial tests on outcrop-chalk from Lie`ge (Belgium). The test core plugs were flooded with MgCl2-brines forapproximately one and a half, two and three years, mimicking effective reservoir stresses (9.5–12.5 MPa) and temperature(130°C) of important hydrocarbon deposits at the Norwegian Continental Shelf. The flooded cores were studied using elec-tron microscopy, whole-rock and stable isotope geochemical analyses, and ion chromatography of the effluent water.All tests show altered textures and mineralogy at the flow-inlet side of the approximately 7 cm long cores. With longerduration of flooding, these alterations moved further into the cores, and for the three-year-test, the entire core was altered.When studied at nano-scale, the newly formed crystals were found to be magnesite containing minor calcium impurities,together with clay-minerals. On the outlet side of the alteration-fronts in the two shorter tests, the mineralogy still mainlyconsists of calcite and primary clay-minerals, together with newly formed magnesite and secondary clay-minerals. Dolomiteor low- and high-Mg-calcite are not observed. The textures of larger micro-fossils are often preserved, but the mineralogy oftheir shells is altered.A sharp, only 4 mm narrow transition zone at the border of the alteration front towards the less altered area for the twoshorter tests, shows the highest porosity in the cores. This pattern resembles what is observed in single-crystal experiments,where the alterations are driven by phase dissolution and subsequent precipitation, the progression of high porosity zones andthe state of equilibrium at the boundary between the primary and new mineral phase. This is also in line with observations innature and models for transport driven mineral replacement in porous media, where differences in dissolution and precipita-tion rates may cause high porosity transitions zones.During the experiments, all cores underwent severe overall compaction between 10.1% and 18.2%. However, in the two-and three-year long test-cores, the permeability, and calculated porosity, started to increase after a primary phase of reduc-tion. As magnesite precipitates at the expense of calcite, the density increase, but the solid volume decrease. As the bulk vol-ume is constant, porosity and permeability are increased. The changes in ion-concentration of effluents, monitoredthroughout the experiments, balance the changes in mineralogy, compaction and permeability within the cores. Composi-tional variations of the injection fluid effectively control the amount of chemical reaction in chalk. This allows for controland predicting changes in geo-mechanical parameters induced by mineralogical replacement, which has significant impacton reservoir conditions.Ó2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).Keywords:Chalk; MLA; Chemo-mechanical testing; Mineral replacement; Long-term testing; Alteration frontshttps://doi.org/10.1016/j.gca.2019.09.0170016-7037/Ó2019 The Authors. Published by Elsevier Ltd.This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).⇑Corresponding author at: University of Stavanger, P.O. Box 8600 Forus, N-4036 Stavanger, Norway.E-mail address:mona.w.minde@uis.no(M.W. Minde).www.elsevier.com/locate/gcaAvailable online at www.sciencedirect.comScienceDirectGeochimica et Cosmochimica Acta 268 (2020) 485–508
1. INTRODUCTIONStudies of mineral replacement encompass a wide rangeof disciplines. A considerable amount of research has beencarried out to understand the kinetics and reactions associ-ated with mineral replacement, whether these are linked tometamorphism, metasomatism, diagenesis and/or weather-ing. Understanding these basic mineral replacement reac-tions is of importance not only as basic knowledge ofrock-fluid interactions, but also to ‘‘(...) quantify and pre-dict the response of Earth’s surface and crust to the disequi-libria caused by the various natural and anthropic input ofenergy to our planet.”(Oelkers and Schott, 2009). Carbon-ate mineral kinetics dictate a wide range of processes in ourworld like preservation of large monuments and buildings,climate models, as well as the characteristics of petroleumreservoirs (Morse and Arvidson, 2002).Magnesite itself is often found as an alteration productof other carbonate minerals, such as calcite (CaCO3) anddolomite (CaMg(CO3)2), during e.g. hydrothermal alter-ation. Formation of magnesite through alteration of calciteby non-equilibrium solutions, has been studied in severalcontributions, both as batch reactions and at nano-scale(e.g. by AFM, review inMorse and Arvidson, 2002). Theseexperiments also show that other compounds most likelyalso play significant roles in the alteration process interplay.The majority of experiments in the field of mineral replace-ment is carried out on single crystals, on powder and on sin-gle grains from natural samples. A thorough review of suchexperiments is given inPutnis (2009), concluding that influid-induced mineral replacement, porosity generation isthe main driver for the progression of the alteration-front.However, observations of convection driven mineralreplacement are observed in geological systems, where theprocess similarly seems to proceed in a front-like mannerwith a porous transition zone (e.g.Merino and Banerjee,2008).Kondratiuk et al. (2015, 2017)propose chemicalmechanisms and model such synchronized dissolution andprecipitation fronts in porous rocks based on the chemicalkinetics of the involved minerals species. They argue thatvolume-preserving replacement can be caused by shifts inthe equilibrium for the dissolving and precipitated phases,rather than driven by the force of crystallization (Malivaand Siever, 1988). The model Kondratiuk et al. presentallows for a stable system with self-regulating fronts topropagate through the system, preserving the volume. Theprocesses involved differ slightly based on the state of equi-librium of the solution with regards to the primary and thesecondary minerals in the transformation.As carbonate reservoirs hold over 50% of the world’shydrocarbon reserves (Roehl and Choquette, 1985;Flu ̈gel, 2004), research on carbonate mineralogy is impor-tant. In the southern part of the North Sea, particularlyin Norwegian, British and Danish sectors, one of the majorreservoir rocks is chalk, containing large hydrocarbondeposits like those in the carbonate-rich Ekofisk (Danian),Tor (Campanian to Maastrichtian) and Hod (Turonian toCampanian) Formations.More than a thousand carbonate EOR experiments havebeen performed at the University of Stavanger, along withother petroleum laboratories world-wide. The time-span ofsuch experiments commonly varies from days or weeks toseveral months, but rarely for years e.g. (Hellmann et al.,2002; Nermoen et al., 2015; Zimmermann et al., 2015). Inthis study, we present the results from experiments onUpper Cretaceous chalk from Lie`ge in Belgium with regardto its mineralogical alterations from the three mentionedultra-long-term tests (516, 718 and 1072 days) flooded withMgCl2at conditions matching important North Sea reser-voirs. The length of these triaxial-cell experiments is noveland paramount to be able to better understand field-scalewater-injection, which is commonly scheduled for years,not weeks or months.Oil-production from the chalk reservoir at the Ekofiskfield started in 1971. A declining production curve alongwith severe compaction in the reservoir and seabed subsi-dence, initiated injection of seawater in the late 1980s(Teufel et al., 1991; Maury et al., 1996; Nagel, 1998;Hermansen et al., 1997, 2000). The rate of subsidence wasreduced; but, further compaction was still observed. Thisindicates, together with decades of experimental work, thatthe interplay between injected fluids and the chalk itselfplays an important role in the mechanical behaviour ofchalk (Hellmann et al., 2002; Risnes et al., 2003;Heggheim et al., 2005; Risnes et al., 2005; Madland et al.,2006, 2008, 2011; Korsnes et al., 2008; Wang et al., 2016).This effect is commonly referred to as water weakening ofchalk. The weakening may have a positive impact on theproduction of oil through reservoir compaction, and is alsofound to play an important role in erosion of carbonatecoastal cliff formations occupying large areas of e.g. the Bri-tish and French coastlines (Lawrence et al., 2013).Water-injection has been used successfully world-wideto (i) maintain reservoir pore-pressure and (ii) increasethe recovery of hydrocarbons. Carbonate reservoirs areprone to be very reactive towards fluids and especially sea-water e.g. (Tucker et al., 1990; Flu ̈gel, 2004; Heggheimet al., 2005; Austad et al., 2008; Korsnes et al., 2008;Strand et al., 2006; Madland et al., 2011). Chalk is a car-bonate rock, which due to its very fine-grained character,has a high specific surface area (often between 1.5 and7m2/g,Hjuler and Fabricius (2009)). Several parameterslike the type of chalk, the composition of the fluid as wellas the pressure and temperature conditions have been var-ied in numerous experiments to understand how theseparameters impact fluid flow, rock-fluid interactions andcompaction. Experiments reveal an extraordinary complex-ity of reactions even though the mineralogy of the rockitself is rather simple. These reactions have an effect onoil recovery through chemical and mineralogical alterationsas well as changes in mineral surface complexes, surface-charge and -potential of the rock (Borchardt et al., 1989;Zhang et al., 2007; Hiorth et al., 2010, 2013; Megawatiet al., 2012; Jackson et al., 2016; Wang et al., 2016;Nermoen et al., 2018).Sakuma et al. (2014)showed thation substitution in calcite, between Ca2+and Mg2+, mayalter the surface tension of calcite where Mg2+incorpora-tion renders the calcite surface more water wet, enough toincrease the oil-production at macro-scale (Sakuma et al.,2014).486M.W. Minde et al. / Geochimica et Cosmochimica Acta 268 (2020) 485–508