Main results achieved during the project. 3.3.2.1 Polysaccharide-based Formulations In order to check the effectiveness of our formulations at high-salinity conditions we tested different salt solutions, by following three complementary approaches. At first, we collected the responses of our formulations to a set of thirteen different salts/co- solute, in the individual concentration of 0.5 mol/L. The salts/co-solutes examined are urea (CH4N2O), trehalose (C12H22O11), potassium chloride (KCl) and ten sodium salts (NaF, NaCl, NaBr, NaI, Na2SO4, NaSCN, NaClO4, Na3PO4, Na2HPO4, NaH2PO4). As can be seen in Figure 9, various salt/co-solute can have a strong effect on the rheological properties of the formulations, even leading to a remarkable change in viscosity. These results confirm how the presence of specific salts can strongly affect the robustness of the network and, consequently, represent a potential trigger to control and modify the fluid viscosity. So, starting from the knowledge of the salt composition and concentration of each different shale basin, we can choice the better formulation that fits with the required operative conditions. Figure 9: Influence of some specific salts on Sodium Hyaluronate viscosity. Secondly, we tested our formulations in a mixed salt solution that we called “shale water” (SW). To prepare SW we took into account data available for the flowback water salinity content of some European shale basins [5,6]. Shale water is composed by NaCl 1 M, CaCl2 0.2 M, XXx 0.000 X, XxXx0 0.01M and BaCl2 0.002 M. The assumption underpinning these experiments is that the salt composition and concentration of the flowback water should be not too dissimilar from the salinity conditions present downhole. Thus, by examining our formulations in shale water we are able to qualitatively predict how these fluids behave in salinity conditions near to the operating ones. As shown in Figure 10, both the GG and SH based formulations exhibit good salt tolerance and their rheological profile is only slightly affected; in the case of HPC the high salinity conditions strengthen the network and increase the viscosity. However, all the investigated systems show a good resistance towards high salinity environment. Finally, we performed additional tests aimed at qualitatively detect the salt toleration thresholds of the various formulations. For reaching this goal we overstressed the salt concentration up to 5 mol/L, a much higher value in comparison with the averaged total salinity of European flowback...
Main results achieved during the project. From the synthetic point of view we have examined the general synthetic pathway for the synthesis of novel zeolites as depicted in Figure 20. The ADOR method developed in our laboratory utilizes oriented chemical weakness of some bonds (in our particular case of Ge- O bonds) to selectively disassemble the parent zeolite and to form the layered material. In this project we succeeded to use zeolite UOV and to transform it into another new zeolite IPC-12. Figure 20: Schematic representation of ADOR process. ‘A’ stands for assembly of the parent UTL structure, ‘D’ – disassembly to IPC-1P zeolite precursor, ‘O’ – organizing by intercalation of organics, and ‘R’ – reassembly by calcination (in this example to IPC-4 (PCR)). Successful synthesis of IPC-12 is important step forward in the synthesis of zeolites as this was the second example how to disassemble the parent zeolite and to produce another one. It strongly evidences the general applicability of ADOR mechanism, which in the future surely will not be limited to zeolites. Using Monte Carlo and molecular dynamics simulations, we predicted adsorption isotherms and transport diffusivities of light alkanes in microporous and hierarchical ZSM-5 zeolites. The hierarchical ZSM-5 zeolite with dual micro/mesoporosities was modeled as an MFI structure where a cylindrical mesopore of 4 nm diameter was introduced. All isotherms were type I, except for propane and n-butane in the dual-porosity zeolite, which were type
Main results achieved during the project. 3.2.2.1 Water based fracturing fluids in clay mesopores New atomistic models of clays have been developed and used within the project: uncharged clay (kaolinite), lower-charge smectite (swelling clay, represented by montmorillonite), and higher charge illite (non-swelling clay, represented by muscovite). For the first time, a significant attention is paid to the structure and adsorption of aqueous species at the non- basal edges surfaces of clay nano-particles. For the adsorption on the basal surfaces of kaolinite, a slit type mesopore 40 Å wide was constructed by 4 kaolinite layers (each composed of a gibbsite and a siloxane sheet) created by a 9x9x4 arrangement of 324 crystallographic unit cells. In this way one gibbsite and one siloxane surfaces are exposed to the fracturing fluid on opposite sides of the pore. It was loaded with a model composition of a water based fracturing fluid - a mixture of water, a salt and an organic additive. For comparison reasons, simpler water solutions with either the salt or the additive were also considered. The salts examined were NaCl, CsCl, SrCl2, and RaCl2. Methanol and citric acid, both in its fully protonated and fully dissociated form, were the three model organic additives studied. Classical equilibrium MD techniques were used to simulate these systems. Preferential distribution of the various fluid components as function of their distance from the walls of the pore (along the z-axis) was calculated based on their atomic density profiles. The pore was divided into three regions (layers) based on the density profile of water and lateral diffusion coefficients along with residence times were calculated for each layer. The formation of citric acid aggregates and number of hydrogen bonds formed between the fluid species and the pore walls were also qualitatively analysed. Figure 5: Atomic density profiles along the axis normal to the kaolinite (001) surfaces for the systems with and without the protonated citric acid (H3A). a) NaCl, b) CsCl, c) SrCl2 and d) RaCl2. The RI and RIII regions of each diagram correspond to the gibbsite and siloxane surfaces, respectively. Water density profile at kaolinite basal surfaces is generally symmetric with two peaks close to each wall. Cations prefer the siloxane side and anions the gibbsite side of the pore (Figure 5). Methanol and protonated citric acid are concentrated towards the siloxane pore wall, while the fully dissociated prefers the gibbsite side. Additives do not affect...
Main results achieved during the project. 3.4.2.1 Microscopic transport modelling SXT developed a lattice-based kinetic Monte Carlo (KMC) model to study fluid transport through slit-shaped pores with different chemical composition. The substrates analysed represent the main components of the inorganic material found in shale rocks. The proposed stochastic model was validated against the analytical solution of the diffusion equation and molecular dynamics (MD) simulations on slit pores. It was found that the chemistry of the pores affects the transport behaviour of methane molecules, as reported in previous studies. The hydrated silica nanopores exhibit the highest permeability, followed by the permeability observed in hydrated magnesium and alumina pores. The agreement between the KMC model and MD simulations is quantitative, however, the computational times are significantly reduced when using the KMC model. The model was then extended in 2-dimensions and was used to provide insights regarding the contribution of the pore network characteristics in the transport behaviour. Two deterministic methods were also used for these investigations. Based on our analysis, among the three approaches considered here the KMC method is considered to be the most sensitive and reliable method. This method responds to changes in both the low and high permeability values. The most valuable feature of the KMC method, compared to the deterministic methods considered, was its sensitivity to anisotropy. For broad distributions, the EMT always over-estimated the network’s permeability, the Simplified Renormalisation method provided low estimates, due to the zero cross-flow assumption, while the KMC predictions were in-between the two. Our model can be considered as a bottom-up approach for mesoscopic studies. Any type of designed or natural network can be simulated, as long as the kinetic (diffusion constants) and thermodynamic (barriers due to the interfaces) properties are provided. The KMC approach as developed by SXT responds to changes in both the low and high permeability values. For dual-permeability networks, KMC detected changes proportional to the components and provided an estimate that captured the matrix structural features. The most valuable feature of the KMC method, compared to the deterministic methods considered, was its sensitivity to anisotropy. KMC could be applied to low-connectivity networks and could also quantify the effect of small-scale heterogeneities (e.g., local low connectivity). When KMC ...
Main results achieved during the project. The models developed enable full and accurate characterisation of safety hazards and evaluation of safe distances to people and structures as a part of risk assessment of a shale gas exploration/production site. We have demonstrated effectiveness of the model in a case study based on realistic design of a shale gas exploration site. The simulation model predictions are presented in the form of 2D plots of thermal radiation and explosion over- pressure contours as a function of distance and time following well blowout. This data in turn forms the basis for determining the minimum safety distances taking into account defined thresholds for deferent severity harm scenarios. Figure 24 shows the instantaneous incident heat flux radiation contours at the ground level within +/- 200 m from the jet flame at 0.5, 2, 10 and 50 seconds after the well blowout. The results correspond to zero wind speed and 200 bar formation pressure. It can be clearly seen that the incident heat flux decreases with the distance from the centre of the jet and also decays with the time, reaching its maximum of ca 3 kW/m2 at ca. 20 m distance from the well at time interval of 30 s. Figure 24: Incident heat flux contours at the ground level around vertical flame formed from the wellhead, predicted at 0.5, 2, 10 and 50 seconds following blowout under no wind conditions.
Main results achieved during the project. During the project, coupled reservoir flow and geomechanical models for computer simulation of sub-surface fluid injections and fault reactivation for a given site characterisation were developed. The reservoir models for fluid flow are used to predict subsurface pressure changes during and after injection. The reservoir flow models are based on the numerical solution of sets of finite-difference equations that describe transient, multi-phase flow in heterogeneous porous media. Analytical solutions were obtained for different types of reservoir and injection conditions. With a prediction of pore pressure changes from the reservoir model, the potential to induce seismicity is then assessed using geomechanical models of faults, which can predict their propensity for slip. The model was applied to European case studies of induced seismicity for Xxxxxxx Shale and also to an Enhanced Geothermal System in Basel, Switzerland. For a given set of model input parameters, the model was able to reproduce the observed seismicity satisfactorily.
