Diagenesis and fractures

bound textures (DRT2, DRT3 and DRT7).

Porosity-permeability measurements on core plugs have been plotted according to the dif-ferent DRTs (Figure 6.10). No specific trend has been highlighted, except for DRT4. The latter is clearly related to the highest values of porosity (13%-14%), and permeability reach-ing about 1mD, which is amongst the highest measured. DRT2 shows values close to DRT4, but display a higher scattered distribution. Other DRTs show dispersed values, which testify that porosity-permeability relationships are commonly not only controlled by the deposi-tional environment, except for reef and mud mound facies. Consequently, other factors have to be developed further to characterize properly the reservoir (i.e., effect of diagenesis, fracturation).

6.4 Diagenesis and fractures

Diagenesis can strongly affect carbonate rocks and their initial fabric, which has a potential positive or negative impact on reservoir properties. Such impact is mostly the consequence of poronecrosis stages (e.g. cementation, compaction) or porogenesis (e.g. dissolution, dolomi-tization). Assessing such features requires the characterisation of the diagenetic phases in terms of origin and timing. Such study can also give potential clues to the geometry of reservoir bodies when not only dominated by depositional properties (Moore, 2001 ; Ahr, 2008).

The origin of diagenesis can be related to different factors such as depositional environment (including water and environmental conditions at the time of sedimentation), sedimentary sequences (unconformities), burial, fluid migration or fractures (Tucker and Bathurst, 2009).

Type, morphology and fabric of diagenetic features (cement, dissolution, physical and chem-ical compaction features) have been used to evaluate the diagenetic environment, even if these criteria are not always precisely diagnostic. Nevertheless, several diagenetic phases are often superimposed, and recognizing the succession and timing of each diagenetic feature has helped deciphering their origins. Coupled with facies analysis, this study has allowed a bet-ter understanding of the conditions sediments have been exposed to since their deposition, and the related evolution of petrophysical properties.

Space-relative information is also often indicative of specific diagenetic origins, especially in the case of fluid migration and fractures. Fluids follow preferentially permeable pathways.

Fractures can act as conduits and carry large quantities of fluid, depending on their hydraulic behaviour (Cacace et al., 2013). In that case, the volume of rock diagenetically altered de-pends on the fracture network, the extension of the damage zone and host rock porosity and permeability prior to fluid migration, which in turn vary according to reservoir and mechanical properties of the different units fractured. Therefore, recognising the influence of fractures in fluid migration and related diagenesis, coupled with facies analysis, provide constraints on the geometry of diagenetic alteration. Using discontinuous well samples, such influence can be difficult to distinguish, and remains hypothetical in this study, because no direct observations on cores could be directly connected to the occurrence of faults and/or fractures. This issue can be addressed with geochemical measurements, which provide

infor-142 Chapter 6. Rock typing of the Kimmeridgian - Tithonian Reef Complex unit mation on the nature of diagenetic fluids. Such measurements are being currently performed in the context of ongoing research activities, and have not been integrated in this study.

Diagenesis influences rock mechanical properties, and therefore the propensity of a given rock interval to respond to brittle deformation (fracture) (Vandeginste et al., 2013). In carbonates, mechanical behaviour depends originally on initial texture, particularly on the mud-content, pore types and the amount of porosity. This evolves then in line with the diagenetic impact, which affects the pore network and stiffness of the different facies (mechanical stratigraphy, Laubach et al. (2009)). The depositional environment, which influences both textures and diagenetic susceptibility, plays a particular role in the distribution, characteristics and density of structural features (Lavenu et al., 2014). This issue is discussed below as a separated paragraph.

6.4.1 Paragenesis

As part of the microfacies analysis, a first description of diagenetical characteristics has been systematically performed on the well and outcrop material. It describes fabric, mineralogy, and proportion of cement, as well as burial and structural features visible on cores and thin sections. A first assessment of successive diagenetic sequences (paragenesis) has been realized, using also cathodoluminescence and SEM observations. Specific investigations on dolomitization have been performed in a parallel study, in which diagenetic phases have been completed (Makhloufi et al., 2018). Results of this study have been integrated to the present work.

Parageneses of the Reef Complex have been assessed for the three main units (Calcaires de Tabalcon, Calcaires récifaux and Calcaires de Landaize), which are compiled in Figure 6.11 (Makhloufi et al., 2018). A full account on the diagenetic processes recorded by the Reef Complex rocks is reported in Makhloufi et al. (2018). In the following paragraphs only the most important processes and related timing impacting reservoir properties are reported.

Eogenetic processes

First eogenetic processes, i.e., micritization of bioclasts, degradation of organic matter and mouldic dissolution of particles (aragonitic and high Mg-calcite precursor only) had a pos-itive effect on reservoir properties. The automicrite fraction created microporosity, and secondary larger pores were formed by dissolution, i.e., moulds and vugs. However, sub-sequent isopachous and syntaxial cement growth led to poronecrosis, while stiffening the sediment. Although these cementation phases slightly affected the pore space, early lithifi-cation seemed to be decisive for microporosity preservation (see §6.4.2). These diagenetic phases are known to occur commonly in phreatic marine conditions, and are not specifically diagnostic (Tucker and Wright, 1990). They have been recognized in the entire Reef Com-plex unit, independently from the depositional environment, although cement growth was obviously more developed in mud-free textures.

6.4. Diagenesis and fractures 143

Figure 6.11: Paragenesis of diagenetical sequences for the Reef Complex unit (Makhloufi et al., 2018).

Mesogenetic processes

Mesogenetic processes led first to the reduction of pore space. The different blocky calcite cementation episodes plugged secondary moulds and primary inter/intra-particle porosity almost entirely. Only few vugs and intercrystalline pore spaces have remained. The latter have been identified on thin section and SEM. Sharp crystal faces bordering intercrystalline pores testify that this porosity is residual and primary in the interparticle area, rather than enhanced by a subsequent, distinct dissolution episode (Figure 6.12). In partially cemented moulds, this porosity is considered as "secondary residual", because the pore space was developed by a previous, eogenetic dissolution episode.

Burial triggered compaction, which decreased even more the porosity, reducing particularly the interparticle pore space (Ahr, 2008). Microporosity in micritic zones (micritic envelope, thrombolite, allomicrite), when not framed and protected by rigid, cemented structures, was generally annihilated as well. However, in specific cases, microporosity has been preserved.

This situation is discussed separately in the following paragraph (§6.4.2).

Compaction formed also stylolites and solution seams, which could act as conduits for flow or permeability barriers (Figure 6.8-G), according to the quantity of sealing material collected in it, and their shape (rectangular-, pinning-, peak-type, wave-like type) (Heap et al., 2014

; Koehn et al., 2016).

Dolomitization

Superimposed eo- to mesogenetic processes of dolomitization have been also identified in each unit of the Reef Complex. According to Makhloufi et al. (2018), two main distinct mechanisms of dolomitization were involved, which are developed below: (1) dewatering of Mg-rich minerals such as clays, and (2) reflux of slightly evolved sea-water (mesosaline

144 Chapter 6. Rock typing of the Kimmeridgian - Tithonian Reef Complex unit

5mm

1 cm

A B

C D

E F

Figure 6.12: Examples of porosity remaining after diagenesis affected the original sedi-ment. (A) Microporous micrite (Humilly-2 well, sample ER-67. (B) Residual intercrytalline porosity between blocky calcite crystals (Humilly-2 well, sample HU-2-3). (C) Residual inter-crystalline mouldic porosity in partially cemented moulds of bivalve shells, SEM secondary electron image (Humilly-2 well, sample ER-68). (D) Isolated vug (Humilly-2 well, core). (E) Intercrystalline porosity in sucrosic dolomite (Reculet Nord, Jura Mountains, outcrop sample REN-3). (F) Intracrystalline porosity in de-dolomitized crystal (St-Germain-de-Joux, Jura Mountains, outcrop sample prp3b, see Figure 6.2 for outcrop location).

6.4. Diagenesis and fractures 145

Figure 6.13: Occurrence of reef-related deposits and sucrosic dolomite in wells and outcrops throughout and around the GGB. Numbers next to the lithology columns indicate the top and bottom stratigraphic depth at which this reef related deposit have been penetrated.

Outcrops are: Le Reculet, St-Germain-de-Joux, and Vuache, while the remaining columns refer to wells.

water), periodically invading the aquifer. In the first case, scattered dedolomite rhombs cross-cutting microstylolites have been observed in outcrop samples only. This relation implies that dolomite formed during burial diagenesis. Dewatering of residual sealing minerals (such as clays) infilling stylolites is thought to be the source of magnesium (Vincent et al., 2007 ; Brigaud et al., 2009) responsible for minor dolomite formation (planar-e crystals scattered in the matrix) in the Calcaires de Landaize unit (Makhloufi et al., 2018).

Larger, more pervasive sucrosic dolomite intervals have been identified in outcrops, in the Calcaires récifaux unit in St-Germain-de-Joux and in the Calcaires de Tabalcon unit in Le Reculet (see Figure 6.2 for location). Unfortunately, the latter intervals were never cored in the GGB, although their presence was notified in geological reports of several wells.

Based on this information and field observations, the repartition of dolomite occurrence has been mapped together with evidences of reef environment (Figure 6.13). It highlights that dolomitization seems to follow a similar trend to that of reef progradation, decreasing downdip to the east (§6.1.4).

In St-Germain-de-Joux, dolomitization appears in irregular patches, and was interpreted as the result of sea and meteoric water mixing zones in phreatic conditions by Fookes (1995).

This interpretation has been refuted by recent study (Makhloufi et al., 2018), which carried

146 Chapter 6. Rock typing of the Kimmeridgian - Tithonian Reef Complex unit out a new and more detailed investigation of outcrop samples in several locations in order to have a larger regional approach. Moreover, the mixing zone model has been extensively discussed in the last decade, and dismissed as a potential source of dolomitization (Warren, 2000 ; Machel, 2004).

Sucrosic dolomite is characterized by hypidio- to idiotopic fabric (planar-e (euhedral) to planar-s (subhedral) crystal shape), crystals presenting a cloudy, inclusion-rich core, zona-tions in cathodoluminescence and syntaxial overgrowth in some places (Figure 6.14). No (micro-)stylolite associated with dolomite formation has been observed in these pervasive dolomitic intervals, nor large anhedral dolomite crystal occurrence (no xenotopic fabric or saddle dolomite). These observations suggest that precursor dolomite nuclei formed during eogenetic processes, and were subsequently enlarged under shallow-burial conditions (con-tinuous eogenetic to early mesogenetic process), at least below 50^◦C ("critical roughening temperature") (Gregg and Sibley, 1984 ; Baldermann et al., 2015), and burial depths of likely less than 100 m (Choquette and Pray, 1970). Dolomitic nuclei are known to be preferentially generated in permeable lime-mud or micrite, whose fabric shows larger effective surface area, and thus, is more reactive than coarse-grained beds (Choquette et al., 1992 ; Choquette and Hiatt, 2008 ; Gabellone et al., 2016). Limpid crystal zonation highlights the evolution of the pore-fluid, which reacts either to successive phases of precipitation, or to pulses of pore-fluid replacement in the aquifer.

According to these observations and the coherence between sedimentary and diagenetic trends, dolomitizing fluids have been interpreted as slightly evolved sea-water into meso-saline water (Makhloufi et al., 2018). The mechanism responsible for meso-saline concentration and dolomitizing fluid supply is thought to be seepage reflux, caused by small-scale sea-level variations. During low sea-level, marine sea-water in partially disconnected lagoons was slightly enriched in Mg while evaporation occurred. The latter fluid invaded shallow aquifers during low-stand episodes, resulting in the formation of dolomite nuclei (Makhloufi et al., 2018). As mentioned, this process likely continued during early burial. Because no sucrosic dolomitized interval could be studied in the subsurface, the dolomitization model is assumed to be similar in both settings. However, advection of hydrothermal fluids through fractures could also be responsible for dolomitization in places. In fact, dolomitic cementation has been already recognized in fissures crossing Lower Kimmeridgian units in Thônex-1 well (Jenny et al., 1995). It testifies that fluid circulation through fractures already happened in the GGB, and Thônex-1 well might not be an isolated case. This assumption should be verified by geochemical measurements and further analysis of fracture diagenesis, both in the field and on new subsurface data.

Assuming that dolomitization was formed by seepage reflux provides constraints to frame stratigraphic sequences. If the formation of mesosaline fluid and pumping through the aquifer happened during low sea-level episodes, the (near-) top of sucrosic dolomite intervals can be considered as the sequence boundary, or the end of the low-stand, i.e., a transgressive surface.

This hypothesis has been used to help sequence stratigraphic interpretation presented in Figure 6.5, coupled with other sedimentological information available from adjacent outcrops (Meyer, 2000a). This approach has been also applied in different other studies (Rameil, 2008

6.4. Diagenesis and fractures 147

; Andrieu et al., 2017).

Geometrical and facies-relation of dolomite bodies has been evaluated. Following interpreta-tions of stratigraphic sequences (Figure 6.5), dolomitic bodies could be roughly correlated in the Calcaires de Tabalcon unit, and seem to correspond to stratabound dolomitic intervals reaching several tens of meters. In the reef unit, dolomitic bodies appear more scattered and could rather correspond to patches from meter to tens of meters thick. This could be the result of pulses of seepage reflux fluids, in accordance to the sea-level fluctuations, leading to partial dolomitization rather than a massive, dolomitized body (Makhloufi et al., 2018).

Concerning the influence of sediments on dolomitization, the degree of dolomitic alteration depends on primary permeability and reactivity properties of facies. Optimal conjunction of reactivity and flow rate of dolomitizing fluid have been recorded in intervening, fine-grain beds (Gabellone et al., 2016 ; Gabellone and Whitaker, 2016). Volume of dolomitic alteration is also controlled by brine salinity, whose increase enhances the reaction efficiency, and by the time of reaction available, which depends on sea-level fluctuations (Gabellone and Whitaker, 2016). Long-time sea-level low-stand seems to favour the development of larger dolomitized volumes.

Assembling observations on wells and fundamental principles dealing with thermodynamic and kinetic parameters of dolomitization (Gabellone et al., 2016 ; Gabellone and Whitaker, 2016) helps predicting lateral distribution of sucrosic dolomite bodies through the GGB. The latter are more susceptible to have developed in proximal facies such as lagoonal and micro-bial mounds (DRT 4 and 5), and/or beneath a sequence boundary or transgressive surface (DRT 1 mainly). Larger volumes of sucrosic dolomite are found in theCalcaires de Tabalcon unit, in which the balance between facies relation to sea-level fluctuations is apparently the best. Reef-front deposits (DRT 1) predominantly made of auto-micrite were deposited quite homogeneously on the GGB area. They were overlain by shallower reef deposits, whose development was likely promoted by a remarkable sea-level drop. Related sequence bound-aries are Kim 3 in more proximal settings and Kim 4 in the external ones (i.e., to the east, Meyer (2000a)), reflecting the diachronic characteristic of facies propagation (Figure 6.5).

According to these observations, dolomite bodies in the Calcaires de Tabalcon unit would be likely continuous within the studied area. In the overlaying Calcaires récifaux unit, fa-cies diversity and smaller-sea-level fluctuations would likely result in smaller, fafa-cies-selective dolomite bodies. Such predictions do not consider fault-related alteration, which can largely enlarge dolomite bodies and enhance their inter-connectivity. In order to properly assess the influence of fractures in dolomitization processes, new subsurface data and measurements are mandatory.

Regarding reservoir properties, secondary intercrystalline pores were created in sucrosic dolomite, where initial rock fabric was totally obliterated. Porosity and permeability in-creased in such sucrosic dolomite facies. However, intermediate, partially dolomitized facies has also been observed, in which dolomite replaced only the matrix part, and main bioclasts were preserved (Figure 6.14-C and D) Tight matrix reservoir properties have been recorded in this case.

148 Chapter 6. Rock typing of the Kimmeridgian - Tithonian Reef Complex unit

Figure 6.14: Photomicrographs from Makhloufi et al. (2018), illustrating porous sucrosic dolomite (sample YM61) A) in transmitted light, B) in cathodoluminescent light, showing cloudy core and growing rims in dolomite crystals of the Calcaires de Tabalcon unit. C) and D) shows partially dolomitized facies (replacement dolomite), recording thight reservoir properties (the Le Reculet outcrop, Calcaires récifaux unit (sample YM63)).

6.4. Diagenesis and fractures 149 Telogenetic processes

Partial de-dolomitization and calcite replacement of dolomite crystals affected both sucrosic and scattered dolomite units observed in outcrop and subsurface. Dissolution and replace-ment seem to have commenced from the inside, affecting first crystal core. This characteristic has already been identified in other outcrops of similar age in the Jura Mountains (Rameil, 2008) and named "hollow dolomite". Composition zonation (higher Ca content), growth defects and soluble inclusions in crystal core would have promoted focussed diagenetic pro-cesses of dissolution and replacement (Swart et al., 2005 ; Jones, 2007 ; Nader et al., 2008 ; Rameil, 2008).

This diagenetic phase is commonly interpreted to have been triggered by late, telogenetic processes, involving meteoric water circulation (Shearman et al., 1961 ; Ayora et al., 1998).

In the deep subsurface, it could likely happen at sequence boundaries and/or during low-stand, when conditions were close to subaerial exposure. This affirmation also supports the present interpretation of stratigraphic sequences (Figure 6.5), where the top of dolomitic intervals marks a sequence boundary, or a transgressive surface in the case of interpreted low-stand system tract. Dissolution might also have affected limestone, leading to the for-mation of moulds and vugs, fissures and fractures enlargement, and even larger paleokarsts.

However, no direct proof of the latter phenomenon has been evidenced to date, and only dedolomitization provides us a tangible indication that dissolution processes have impacted the deep subsurface.

The timing of telogenetic processes in the GGB is still not well constrained. According to interpreted environmental conditions during the Kimmeridgian-Tithonian and stratigraphic sequences, meteoric alteration could already have occurred during small-scale sea-level vari-ations, as mentioned above. In addition, it could have been active again, since the Jura Mountains uplifted. In the shallow subsurface and at the surface, dissolution mechanisms have been recognized in theReef Complex unit, resulting in moulds, vugs, and karst develop-ment, as testified by the higher diversity of pore types observed and enhanced petrophysical properties of outcrop samples. Wide caves and tunnel networks continuously evolving in surrounding massifs demonstrate also this recent alteration. Deep water circulation in Up-per Jurassic units (> 1850 m) has been recorded in Thônex-1 well (Jenny et al., 1995 ; Muralt, 1999 ; SIG, 2011). According to these authors, geochemical analyses of the sampled borehole fluid have indicated recharge in the Jura Haute-Chaîne, and a residence time of 10’000-15’000 years. At this location in the basin and according to its trajectory through apparently low-permeable rocks, the fluid is in equilibrium with its geothermal reservoir.

At this stage, this fluid is no more corrosive, and would likely not alter the rock matrix.

However, closer to the zone of recharge, and along potential fluid flow corridors, fluid geo-chemistry is certainly closer to meteoric composition, and could still react with carbonates.

In larger pore networks, turbulent flow currents could still also enlarge rapidely such struc-tures rapidely. These assumptions have to be verified by the analysis of new data acquired in the future wells. Nevertheless, this model should not be refuted, since the extension of the current karst network and role of fractures in water circulation through the subsurface are still poorly documented and understood. The South German Molasse Basin (SGMB)

150 Chapter 6. Rock typing of the Kimmeridgian - Tithonian Reef Complex unit geothermal reservoir shows a nice example of fractured/karstic reservoir, whose dissolution features extend from the surface (Swabian Alb) to more than 5000 m southward (see Chap-ter 6.9). This testifies that karsts can widely extend within the subsurface, even at a great depth (Schulz and Thomas, 2012 ; Homuth et al., 2015b).

De-dolomitization, even partially replaced by calcite, modified the petrophysical properties.

Where dolomite rhombs appear scattered in microporous, micritic matrix (Figure 6.12f),

Where dolomite rhombs appear scattered in microporous, micritic matrix (Figure 6.12f),