Download PDF

Travertine shrub structures: origin, diagenetic modifications and petrophysical characteristics - Tivoli case (Central Italy)

Publication date: 2018-02-05

Author:

Marques Erthal, M
Swennen, R ; CAPEZZUOLI, E

Abstract:

The discovery of complex continental carbonate reservoirs in the South Atlantic (Brazil and Africa) rift-sag lacustrine basins has generated considerable industrial and scientific interest. One of the most common and promising reservoir lithotype described in the so-called "Pre-Salt" interval is characterized by shrub structures. Searching for analogues, the shrub morpho-types from Tivoli travertines (Central Italy) seems to be a candidate, since they display petrographic features and pore-morphologies remarkably similar to the Pre-Salt reservoir rocks. The shrubs from Tivoli also call attention by their laterally very flat and continuous layers, mapped over hundred meters, with local packages of more or less 40 m thick. They possess the dimensions of a small reservoir petroleum field. To better understand and characterize the shrubs from Tivoli and their pore network, a 2D and 3D multi-method and multi-scale workflow was worked out, in which the sedimentology and geochemistry is first studied and subsequently the pore network is accessed. Understanding complex variations in pore geometry within different lithofacies is the key to improve reservoir description and exploitation. In fact, variations in pore geometrical attributes define distinct flow zones (hydraulic units) with similar fluid-flow characteristics. The shrubs from Tivoli are mainly characterized by their branching texture, however, they are very variegated. By analyzing them petrographically and by using 3D micro computer tomography (µCT) images, it was possible to distinguish 6 shrub morpho-types, named: narrow dendriform shrubs, wide dendriform shrubs, fili dendriform shrubs, arborescent, arbustiform and pustular shrubs. Their textures (morphologies, size, size sorting and packing) greatly vary, however petrographic analysis showed that they are monotonous in relation to their mineralogy and basic fabrics. They are 100% calcite, and are constituted of peloidal micritic aggregates, always surrounded by a very thin coating of sparry calcite cement. Rarely, they possess a crystalline habit. The presence of these sparry calcite cements that surround the shrubs, influences the quality of the reservoir, because it shields micro-porosity, which is present in the center of the shrub, from meso- and macro-porosity surrounding the shrubs. The travertine shrub structures from Tivoli are interpreted in the present study to have developed in very shallow extensive waterlogged, slightly inclined flat areas, changing laterally into a slope system with crusts as the main lithotype. Shrub morphologies likely reflect specific (micro-)environments that are controlled by water flow rates, evaporation and microbial activity. Under high flow conditions, CO2 degassing is the main process leading to carbonate precipitation. Consequently, dense and tightly packed morphologies will precipitate, mainly consisting of the crust lithotype. In this setting, microbes are less dominant. Moreover, dendriform shrubs, with narrow, wide and fili morphologies are interpreted to occur in settings with moderate- to low-energy water flows. Narrow dendriform shrubs reflect faster flowing conditions, with decreasing impact of flow on the morphological characteristics toward wide dendriform shrubs to fili dendriform shrubs. Slow to stagnant flowing waters are more characteristic for the arborescent, arbustiform and pustular shrubs that are possibly highly influenced by evaporation. Besides, the shrubs in the study area make up three depositional sequences that are limited by erosive surfaces. The stable C and O isotopes are also in agreement with the proposed sedimentological interpretation. The stable carbon isotope signature showed high values varying between +8.71 and +11.32‰VPDB, and stable oxygen isotope values varied between -4.97 and -8.25‰VPDB, indicating that precipitation was mainly influenced by degassing and evaporation processes. The very high C stable isotope signatures and the shrub fabrics that are mainly composed of micritic peloidal aggregates suggest that microbes mediated also the carbonate precipitation. The 87Sr/86Sr ratio signatures pointed out that the Mesozoic limestones of Central Italy served as the main source rock. Besides, the presence of Sr, S, Na and Ba obtained by elemental analysis, suggest that the fluids also percolated the Triassic evaporites. The main pore types observed in all the shrub samples consist of intershrub and interdigit growth framework pores. Plugs yield porosities from 0.8 to 20.9 % and permeabilities from 0.001 to 5255 mD. No relationship was observed between the shrub morphologies and porosity and permeability lab measurements. However, it was observed that the shrub packing, which corresponds to the ratio between shrub width and adjacent spacing, controls primarily the porosity and secondarily permeability. In addition, shrub size sorting, defined by the complexity of the shrub sizes within a sample, controls primarily permeability and secondarily porosity. Micro-Computer Tomography imaging was used to render and evaluate the 3D arrangement of shrub morphologies, pore connectivity and porous framework. The results pointed out that shrub morpho-types can be distinguished based on their pore shape volume and possess high pore network connectivity. Therefore, an integrated methodology, including Nuclear Magnetic Resonance (NMR), Mercury Intrusion Porosity (MIP), Micro-Computer Tomography (μCT), porosity and permeability measurements, petrography and SEM analyses, was introduced to gain insight into the complex variations of the pore network within the Tivoli shrub facies. It was observed that micropores have a significant impact on the permeability by increasing the fluid pathways and tortuosity, and consequently negatively affecting the permeability. This workflow allowed observing that shrub size sorting not only controls permeability, but also pore entrances. Besides, they secondarily influence pore body sizes and porosity, while shrub packing influence primarily the pore body sizes and porosities. The study of the Reservoir Quality Index (RQI) and Flow Zone Indicator (FZI) displayed a relationship with pore parameters, and also evidenced that the amount of microporosity in the pore network negatively affects the quality of the Tivoli travertine shrub reservoir analogue. The relationship between RQI and FZI with MIP/NMR groups represents the heterogeneities of the complex travertine reservoir, and helped understanding the complexity of the pore network. The application of the Lattice Boltzmann method for permeability simulation derived from μCT images pointed out that shrub morpho-types display a relationship with porosity (µCT results), permeability (Ksim) and tortuosity, which varies according to the different depositional shrub architectures (morphology, arrangement of the branches, packing and size sorting). Furthermore, by the use of a portable hand held air mini-permeameter it was possible to quantify the permeability distribution within a reservoir with different shrub morpho-types in detail. Vertical permeability variation was very high, as well as high lateral permeability variation is observed at cm-scale. High anisotropy occurs within laminae, with larger permeability continuity in the direction parallel to the laminae. Textural heterogeneities are responsible for the high permeability variation values within laminae. The study of acoustic wave velocities showed a relationship between shrub morpho-types, pore-types and porosity. Samples with frame-forming pore-types as mouldic pores or encrusted bubbles show higher velocity values and lower porosities than samples with no frame-forming pores as vugular pores, resulting in lower velocity values and higher porosities. Besides, the analyzed samples yield lower velocities with higher porosities. In addition, the comparison of Tivoli travertine samples with Turkish and Hungarian ones showed very similar acoustic wave velocity behavior. On the other hand, the comparison with marine carbonate indicates very different compressional-wave velocity relationships with porosity.