Modelling facies heterogeneity in carbonate ramp systems. From petrophysical characteristics to forward modelling

In the last decades digital modelling applied to geological research is getting increasing attention (Alaei, 2012; Tomassetti et al., 2018; Trippetta et al., 2020; De Franco et al., 2019; Mascolo and Lecomte, 2021). Indeed, relevant implications both in scientific and economic terms could be inferred by using this technique. In particular, the application of digital models in complex geologic scenarios is critical for the understanding of potentially exploitable systems from multiple perspectives. Starting from the most classical model application for the exploitation of oil and gas fields passing through the implementation of extraction strategies - by reducing uncertainties (Macgregor & Moody, 1998; Racey 2001) - digital models find new place in latest applications such as natural gas storage. Recently, models are also applied for the study of geological bodies, potential reservoirs for the CO2 or hydrogen injection (Dockrill and Shipton, 2010; Trippetta et al., 2013; Aminu et al., 2017; Heinemann et al., 2018). Modelling contribute and facilitate to capture and store gases in the subsurface, balancing their release into the atmosphere. Digital modelling represents one of the major innovative strategies in the control of greenhouse gases concentration in atmosphere, a currently trending topic from media, public opinion, and political points of view. Another possible application of digital models for subsurface gas storage involves the monitoring of reservoirs in order to ascertain and quantify gas leakage through fault or fracture systems (Wang et al., 2018). Moreover, radioactive waste storage could be integrated as current and powerful employment of digital models (Malvić et al., 2020). In particular, the technological tools used for these purposes are called forward models since their outcomes gives predictive results on the processes happened in the past and protracted towards the future. They appear extremely suitable for the study of geological subsurface formations that can be also applied to an emerging field such as the development of geothermal energy power plants (De Franco et al., 2019). All these are topics of great actuality since world governments' plans are1 directed towards the total replacement of classic energy sources from hydrocarbons with green energies. However, digital modelling needs input data such as geometries and rock properties that should be well constrained. Seismic exploration is probably the most powerful tool for investigating subsurface rock formations (Avseth et al., 2010). Important progress has been made in recent years, but significant problems remain in the geologic interpretation of seismic data. The reflections that can be read in seismic data depend on the Acoustic Impedance (AI) contrast in the transit of the P-wave between layers in the subsurface. AI depends on the density (ϼ) and the P-wave velocity (Vp) of the medium through which wave propagates (AI= ϼ Vp). These petrophysical characteristics, in turn, are controlled by structure, texture, porosity, and boundary conditions of the rocks (Dvorkin et al., 2014; Tomassetti et al., 2018; Trippetta et al., 2020; Brandano et al., 2020). These two links, one between rock structure and its elasticity and the other between elasticity and signal propagation, form the physical basis of seismic interpretation (Anselmetti and Eberli, 1993; Eberli et al. 2003; Weger et al. 2009; Hairabian et al. 2014; Dvorkin et al., 2014). Dealing with these relationships, we are facing the so- called inverse problem. We see from seismic sections the resulting seismic images of rock formations where the same signal can be the result of a combination of different features. It should be, thus, very useful to well understand what are the features that lead to a certain seismic image. Synthetic seismic modelling (or forward modelling) is a fundamental prospecting method for understanding the features leading to the corresponding seismic images of subsurface structures and reservoir architectures (Alaei, 2012). Forward modelling methodology, as approach to the interpretation of seismic data, involves the detailed characterization of lithology, density, porosity, seismic velocity and fluid in the rock, as well as the reservoir geometry. As a result, the corresponding seismic properties are calculated, and then synthetic seismic traces are generated. These issues became essential for lithologies characterized by a complex seismic interpretation (Al-Salmi et al., 2019). In addition, synthetic seismic forward models allow accurate analysis of fault zones. The study of seismic response in fault zones is crucial since the2 fracturing or compaction that faults create strongly modifies the petrophysical characteristics of rocks by affecting their properties (Botter et al., 2017; Kolyukhin et al., 2017). Synthetic seismic forward models are, therefore, mandatory for the comprehension of faults behaviour through seismic imaging. Faults play a key role in reservoirs by increasing or limiting fluid flow. Even if interpretation of seismic data is a pivotal method for studying the subsurface, the internal structure and properties of fault zones are often below the limit imposed by seismic resolution (Botter et al., 2017). Despite the impact of faults on reservoir permeability, modelling tools and workflows still lack for realistic representation of fault zones in models (Tveranger et al., 2005; Braathen et al., 2009; Manzocchi et al., 2010). With facies analysis and petrophysical data it is possible to build field-based digital models fundamental in understanding architectures of carbonate sedimentary bodies which often constitute reservoir surface analogues of buried world-wide petroleum systems, CO2, hydrogen, radioactive waste storage sites and geothermal fields. Surface analogues are rocks with depositional, textural, and petrophysical characteristics similar to those constituting the petroleum system, but they outcrop on the surface. Starting from petrophysical characteristics of facies, forward models can be built. In this thesis, as a case study for the development of a forward model, rocks belonging to the carbonate realm, more specifically carbonate ramps, were analyzed. Carbonate ramps constitute important hydrocarbon deposits in North Africa (Macgregor & Moody, 1998), Venezuela, and many other regions of the World (Racey, 2001) due to their excellent porosity and permeability characteristics. However, the depositional model that is the basis for a proper interpretation produces many uncertainties arising from the difficulty in attributing different facies to a depositional environment and process due to the poor occurrence of sedimentary structures (Buxton and Pedley, 1989; Pomar and Kendall, 2008; Burchette, 2012; Bassi et al., 2013; Tomassetti et al., 2018; Tomassetti et al., 2022). In addition, strong lateral heterogeneity in terms of petrophysical characteristics, components, structure, and texture leads to complex distinction of facies belts (Tomassetti et al., 2018; Trippetta et al., 2020; Brandano et al., 2020). To overcome these issues, quantification of3 petrophysical characteristics can be crucial in understanding facies heterogeneity from a physical perspective to be incorporated in synthetic seismic forward models building. Carbonate rocks are often difficult to interpret seismically because the slight acoustic impedance contrast at the interface between carbonate facies in subsurface does not allow a clear resolution of major reflectors and reservoir formations. Strong constraints are often imposed by geophysical survey techniques characterized by low resolution especially in carbonates and interpretation capabilities that depend on the interpreter skill (Tomassetti et al., 2018; Trippetta and Geremia, 2019; Faleide et al., 2021). These constraints can be overtaken through the modelling of surface analogues allowing a detailed analysis on the facies association but also their petrophysical characteristics and seismic properties such as acoustic impedance (Tomassetti et al., 2018; Lipparini et al., 2018; Trippetta and Geremia, 2019; Brandano et al., 2020). In order to analyse the petrophysical characteristics and seismic response of the carbonate realm through modelling two carbonate ramps both belonging to the Adria plate were considered as case studies. The first is the Chattian carbonate ramp of the Porto Badisco calcarenite outcropping in the southern Salento peninsula, the southernmost portion of the Apulian carbonate platform. The Porto Badisco carbonate ramp is an excellent surface analogue of exploited oil and gas field in the offshore Venezuela, Philippine and South China Sea (Zampetti et al., 2005; Sattler et al.,2004; Fournier and Borgomano, 2007; Lallier et al., 2012; Marini and Spadafora, 2014; Pomar et al., 2015; Valencia and Laya, 2020) as well as fields in offshore Adriatic Sea such as Ombrina Mare field (Campagnoni et al., 2013). In this carbonate system firstly the analysis of outcropping facies was carried out observing over 100 thin sections produced. Consequently facies association modelling was performed through Petrel software (mark of Schlumberger) using TGSim stochastic approach algorithm adopting the depositional model based on field data. This model is useful for qualitatively understand the broad facies spacial distribution which reflects the abrupt heterogeneity from a sedimentary point of view. To physically quantify the lateral facies heterogeneity the petrophysical characteristics such as porosity, density and seismic velocity were measured and analyzed through a multi-analytical approach. Density4 measurements were carried out with the helium pycnometer. Porosity was firstly calculated from the density data and then was additionally measured through image analysis and point counting to cross-correlate the values. Seismic velocity was measured by using an ultrasonic generator connected to piezoelectic transducers and to an oscilloscope. The analysis performed on the carbonate ramp outcropping in Porto Badisco offers the opportunity to analyze facies heterogeneity, modeling its distribution and physically quantifying it through petrophysical characterization. From the petrophysical data, it was possible to construct 2D models of the distribution of porosity and P-wave seismic velocity along the depositional model. This study, which can be applied globally to carbonate platforms, emphasizes with the modelling exercise how facies heterogeneity is an intrinsic feature of these systems. The petrophysical characterization which provides quantitative values to the heterogeneity allow to build more complex models such as seismic forward models discussed in the second chapter. The other case study is represented by the Cenozoic carbonate ramp outcropping on the Majella Massif in Abruzzi, the northernmost portion of the Apulian carbonate platform which gives the opportunity to study a carbonate ramp surface analogue of a buried reservoir. Also in Majella the Oligo- Miocene stratigraphic interval represented by the Bolognano Formation which is the reservoir of the system is considered an excellent surface analogue of the productive fields in the Adriatic Sea, offshore Venezuela, Philippines and many others worldwide (Tomassetti et al., 2021). Specifically, this system offers the opportunity to integrate the facies heterogeneity in the synthetic seismic forward modelling and understand its seismic response without the introduction of artificial noise to obtain additional information. On the Majella Massif a model of the facies heterogeneity to understand their seismic response was performed. After analyzing the facies and measuring their petrophysical characteristics, the data obtained were used as input for build a 3D property modelling in Petrel software representing the entire carbonate ramp from the topographic surface to the Upper Cretaceous from the platform top going towards the basin located northward from the Majella Massif. From the 3D model was cut a section whose data were used as input in Matlab (mark of Mathworks) in order to perform the synthetic seismic forward model5 with the geophysical codes provided by the CREWES consortium. The resulting forward model represent the seismic response of the facies heterogeneity of carbonate rocks. In addition, from the obtained seismic images it is possible to evaluate the presence of hydrocarbons and to identify how the presence of important bituminous impregnations – that can be appreciated in the field in Majella – modify the seismic response. The workflow developed to quantify the signature of the facies heterogeneity of carbonate rocks and the presence of infilling hydrocarbons is applicable to other systems worldwide, which is a large issue that is still open and can help in the problems relative to seismic interpretation associated with these systems. Given the presence of a buried normal fault system in the study area, a forward modelling in the fault zones was performed as well. By measuring the petrophysical characteristics of the fault rocks characterized by both fracturing or compaction, fault zones were modeled. Two end member scenarios with two opposite behaviors of the rocks belonging to the damage zone were modeled in Matlab. A scenario in which the damage zone is characterized by fracturing and therefore rocks affected by greater porosity than the host rock. In the other scenario was modeled a damage zone with lower porosity than the host rock caused by the presence of compaction bands. Consequently, the seismic response of these end members was compared to understand how faults affect the seismic response of carbonate ramp systems. Notoriously, fault systems globally characterize carbonate ramps, and understanding their seismic response facilitates interpretation of the deformation behavior that a fault can assume under different boundary conditions. This can lead to an understanding of whether faults behave as barriers or conduits for fluids with the important implications for the study of fluid leakage from reservoirs.