Sedimentological, diagenetical and petrophysical characterization of travertine lithofacies in the east of Denizli basin (SW Turkey)

The Ballık subaqueous travertine lithofacies in Denizli Basin were studied as reservoir analogues for the Cretaceous lacustrine Pre-salt carbonate reservoirs offshore Brazil and Angola. Despite that these subsurface reservoirs display significant differences in terms of depositional fabrics when compared to travertines, the latter show some similarity to lacustrine carbonate reservoirs, with regard to microporosity, vuggy and calcitic dendrites-related macroporosity, abundant cementation and dissolution overprints. In this thesis the spatial distribution of Quaternary depositional environments with diverse travertine lithofacies in the Ballık area was reconstructed to unravel reservoir architecture in geobody-scale. In addition, travertine lithofacies and travertine geobody-based reservoir heterogeneity in terms of diagenetic and petrophysical properties were determined. Primary fabrics of travertines are frequently modified by diagenesis, leading to heterogeneous pore networks. Therefore, it is very hard to define the porosity and permeability characteristics of travertine lithofacies formed in slope and sub-horizontal travertine geobodies. In travertine lithofacies, an integrated approach of computer tomography (CT), mercury injection porosimetry and imaging techniques has been applied to sort out the distribution of pore networks linked to their petrophysical parameters. Within the scope of the study, it is also explained how microporosity and macroporosity affect petrophysical parameters. In this thesis three main research targets were identified in determining reservoir heterogeneity of lacutrine travertines, namely: 1) how to reconstruct travertine lithofacies-based paleoenvironmental reconstructions, based on line-drawings from different travertine quarries in the Ballık area. The Faber-West, İlik, Alimoğlu, Best-Abandoned and Demmer-Başaranlar travertine quarries reflect a west to east lateral transition in the Ballık area. These travertine quarries, furthermore, occurred in a similar topographic level and formed a domal structure which was named as Lower Domain. Some terrestial clastic inputs were observed adjacent to travertine lithofacies in the Upper Tuna and Kömürcüoğlu travertine quarries, which are located at a higher topographic level in the Ballık area. Since these two travertine quarries in the so-called Upper Domain display different travertine lithofacies and depositional environments from those of the Lower Domain, paleoenvironmental reconstruction of Lower and Upper Domain was studied separately to delinate the spatial distribution of depositional environments, associated with depositional geometry of travertines. 2) addressing the diagenetic overprints and pore type properties and their impact on petrophysical parameters acquired from the Ballık, Aşağıdağdere, Gürlek, Akköy and Tivoli travertine lithofacies. Based on a large dataset related to travertine lithofacies-based reservoir heterogeneity, the controlling factors with regard to the relationship of porosity and permeability as well as porosity and Vp was investigated. 3) to study the Aşağıdağdere travertine quarry including slope and sub-horizontal travertine geobodies for travertine geobody-based reservoir heterogeneity. The following research steps were undertaken. Two hundred sixty-nine (269) plugs were sampled for helium porosimetry and nitrogen gas permeability analyses by using a hand-drilling machine. In addition compressional acoustic velocity measurements were performed on 117 plugs under low pressure conditions (5 MPa). To determine the dominant diagenetic overprints and pore types formed in different travertine lithofacies fluorescent-dye resin impregnated thin sections were made of 58 samples. High-pressure mercury injection capillary pressure (MICP) analyses (up to 60.000 psia) were carried out to quantify pore throat size distribution and to acquire information on capillary pressure plateau indicating displacement pressures on 7 broken fragments. Medical-CT analysis was performed on 4 inch cores representing different travertine lithofacies in order to render their pore networks in 3D. In Aşağıdağdere travertine quarry δ13C and δ18O isotope analysis was applied to 58 samples to figure out the origin of fluids giving rise to the slope and sub-horizontal travertines and to investigate the meteoric and vadose diagenetic products that occurred within these travertine deposits. Besides, sixteen samples were dated by the U/Th method to acquire information on the age of deposition of the slope and sub-horizontal travertine geobodies. Statistical methods were applied to reinforce the integrated study of diagenetic and pore type properties observed in the six differentiated travertine lithofacies and their petrophysical parameters such as porosity and permeability variables. These methods consisted of the application of linear regression model associated with multiple-way analysis of variance (ANOVA) analyses, homoscedasticity, principal component analysis (PCA) and Kruskal-Wallis test. The first statistical method was applied to determine how travertine lithofacies controls porosity and permeability as well as porosity and Vp relationships. In this study data (from 269 samples) from Turkey and Italy were compared with those (from 283 samples) from lacustrine Pre-Salt carbonates in terms of porosity and permeability variables by using Kruskal-Wallis test. In relation to the 1st research goal, it should be mentioned that the "Lower Domain" consists from west to east of a laterally complex amalgamation of extended pool, marsh pool and flood plain environments that formed within a mixture of spring and ground waters. The extended pond environment characterized by a stromatolitic boundstone lithofacies in the west evolves eastward into a marsh pool and flood plain environments. This change is interpreted in terms of decreasing CO2 degassing and water temperature as the water depth of the Lower Domain diminished towards the east. The marsh pool environment includes pack- to grainstone lithofacies and abundant wackestone lithofacies made of phytoclasts, whose crusts exhibit pustular fabrics. Moreover, the flood plains along with the marsh pool consist dominantly of pack- to grainstone lithofacies with many gastropods and intraclasts, interfingered with wackestone lithofacies made of phytoclasts. Irregular clotted fabrics, along with coated grains with radial fibres and high lime mud content with bioturbation are also characteristic features. The "Upper Domain" displays a laterally less heterogeneous paleo-environmental distribution with flooded slope and flood plain deposits. The eastern part of the "Upper Domain" is characterized by a systematic alternation of these environments, with intercalations of wackestone lithofacies made of phytoclasts, packstone lithofacies made of intraclasts and lime muds as well as coated grains. The flood plain consists of coated grains having peloidal nuclei and coatings of sparry laminations as well as clotted fabrics of peloids, whereas, the flood slope possesses coated grains with coatings of dendrites and nuclei of peloids, boundstone of stromatolites which have flat-laminated and columnar-laminated fabrics indicating laminar fluid discharge away from the spring(s). Noteworthy is also that alluvial fan and palustrine deposits with abundant bryophytes and reeds frequently interfinger with marsh pool deposits in the "Lower Domain", and with flood plain and flood slope deposits in the "Upper Domain". These results illustrate well how environmental changes identified in the two different domains control heterogeneity in reservoir-based depositional architecture. With regard to travertine lithofacies-based reservoir heterogeneity, six main travertine lithofacies were defined, namely: massive packstone (MP), pack- to grainstone (PG), phytoclastic packstone (PP), dendrite crusts (DC), stromatolite boundstone (SB) and phytoherm boundstone (PB). It was found that travertine lithofacies showed a moderate porosity to permeability relationship (R2 = 0.5), and it exhibited very weak porosity to Vp relationship (R2 = 0.12). Removing 43 outlier samples out of the total dataset (269) the porosity to permeability correlation increased to a R2 value of 0.7. These outlier samples contain abundant microporosity, vuggy macroporosity, microcrystalline calcite and blocky calcite cements and, to a lesser extent, geopetal micrite. After being created dummy variables for each of six travertine lithofacies in multiple regression analyses, ANOVA analyses show that phytoherm boundstone, pack- to grainstone and phytoclastic packstone travertine lithofacies are very distinctive in porosity to permeability, while phytoherm boundstone, stromatolite boundstone lithofacies are distinguished in terms of porosity to Vp relationship. Furthermore, the emprical equation representing pack- to grainstone travertine lithofacies can be used for calculating the permeability from porosity values. According to principal component and correlation analysis, variables showing a positive trend against PC1 was cementation, while the negative contributing variables to PC1 were permeability, dissolution and phyto-mould macroporosity. Similary, for PC2 the positively oriented variables consist of porosity and vuggy macroporosity, the negatively oriented variable was represented by the only microporosity.The Aşağıdağdere travertine quarry including slope and sub-horizontal travertine geobodies was selected as a case study to determine travertine geobody-based reservoir heterogeneity. The slope travertine geobody (deposited between 29.6 to 17 ka) is characterized by terrace, sloping fan and waterfall deposits displaying a progradational to aggradational pattern. The floodplain deposits (deposited between 17 to 10 ka ago) contain extraclasts, intraclasts and lime muds. They overlie a horizon reflecting less travertine precipitation, which truncates the slope travertines likely due to the high discharge of meteoric waters within the travertine system. The overlying sub-horizontal travertine geobody (deposited between 10 to 1.5 ka) contains intercalations of palustrine and marsh pool deposits displaying an aggradational pattern. The influx of meteoric waters caused an inverse relationship between δ13C and δ18O values which slightly decrease from terrace slope deposits towards the marsh pool and paludal deposits, reflecting abundant vadose diagenetic overprinting. The dissolution, recrystallization and cementation features observed in the terrace and sloping fan deposits created microporosity, inter-laminar, inter-peloidal, intershrub growth framework, mouldic and vuggy porosity. They are characterised by micro- and meso-sized pore throat abundances of 58 to 65% and 31 to 36%, respectively. In the sub-horizontal travertine geobody the marsh pool and palustrine deposits exhibit abundant micritization associated with microborings and dissolution features. The latter diagenetic processes can generate additional microporosity, inter-peloidal as well as vuggy porosity, represented by 58 to 75% micro- and 25 to 38% meso-sized pore throat abundances. In both travertine geobodies the diagenetic overprint resulted in (i) extensive micro- and macroporosity, (ii) patchy poroperm distribution, (iii) unimodal to bimodal pore throat size distribution, (iv) steep capillary pressure plateau levels and different displacement pressures, from 32 to 59 psia and 3 to 122 psia in slope and sub-horizontal travertines, respectively. In addition, the slope travertine in contrast to sub-horizontal travertine was characterized by a more homogeneous pore network and wider poroperm range. Our findings illustrate that their reservoir quality and heterogeneity can be linked to sedimentary and diagenetic processes. In terms of porosity, Pre-salt calcite crusts and grainstone/rudstone lithologies are similar to stromatolitic boundstone and packstone to grainstone travertine lithofacies, respectively. Moreover, mean permeability value of the first reservoir lithology is very close to those of dendrite crusts and stromatolitic boundstone travertine lithofacies. This confirms that porosity and permeability are controlled by diagenetic overprints and pore type properties since travertines contain diverse diagenetic fabrics yielding a variety of macro- and microporosity. Reconstructing the spatial distribution of porosity, cementation and dissolution zones in sub-horizontal and slope travertine geobodies allowed to figure out reservoir heterogeneity at geobody scale. This can serve as a good analogue study to improve exploration and production targets in subsurface lacustrine carbonate reservoirs as in the Pre-Salt. It is recommended that their spatial distribution should be merged with seismic data reflecting domal or slope morphology with aggradational and progradational depositional geometries.