Pitcher, Eleanor Grace
(2021)
Novel characterisation methods for pore systems of seal rocks in reservoirs used for downhole gas production and storage.
PhD thesis, University of Nottingham.
Abstract
Seal rocks, also called cap rocks, are a crucial and sometimes overlooked factor (due to not being the primary factor in exploration, having more of a role in the resource evaluation and development) in the evaluation of a potential gas accumulation, and is critical in downhole gasification (enhanced gas recovery) and storage of other gases. Shale rocks are the most common seal rock in conventional reservoirs; currently shales are providing an unconventional oil and gas source which can act as a potential buffer to the energy industry as it transitions towards renewable energies (which are still in their formative years) whilst there is a continued rise in demand for energy globally. Over the past ten years there has been a boom in shale gas production in the United States (Barsotti et al., 2016; Li et al., 2016; Yu et al., 2016), and it is anticipated that this boom may be repeated in the UK (Andrews, 2013).
Downhole gasification (enhanced gas recovery) offers a potential way to produce from these “difficult-to-extract” (as a result of low permeability’s) reservoirs by using carbon dioxide as a displacement gas for methane. At the same time this carbon dioxide can be also be stored resulting in the environment being exposed to less greenhouse gas (Kim, Cho and Lee, 2017; D. Liu et al., 2019). However, it is erroneous to consider shales as a completely impermeable layer, and their ability to retain different fluids is variable (controlled by the capillary entry pressure and/or the permeability and the extent of diffusive loses) which could result in some/all of them being ineffective at retaining carbon dioxide. This is because shales are highly complex and anisotropic containing pores over several orders of magnitude. Typically they have a significantly low permeability and porosity, combined with structural and chemical heterogeneities of shales mean that physical processes are significantly impacted. Importantly the structure-transport relationship is complex resulting in processes such as hydrocarbon migration, methane extraction, gas storage, or carbon sequestration being poorly understand. This project proposes the development of several novel characterisation techniques and combinations of complementary techniques to characterise the multi-scale properties of shales in order to more accurately provide the information needed for secure decisions regarding gas production and storage.
In this work mercury porosimetry, together with mercury thermoporometry, and computerised x-ray tomography (CXT) were performed on post-porosimetry samples containing entrapped mercury, to characterise the pore structure of cap-rocks. However limitations were identified where mercury was trapped in pores too large to sufficiently suppress the bulk melting point (thermoporometry) such that a separate melting peak formed. However, the combined use of mercury porosimetry and computerised x-ray tomography was effective at highlighting the location of trapped mercury, but was ineffective at providing quantitative results regarding the macroporosity of the sample. Further drawbacks of mercury porosimetry based analysis are the potential destruction of experimental material where further analysis cannot be carried out unless mercury forms part of the experimental technique (i.e. thermoporometry and computerised x-ray tomography as described above).
Therefore, gas overcondensation, was proposed as an alternative technique as a bridge between micro-pore characterisation, below the limit of mercury detection, up to macro- pores which are undetected in conventional sorption experiments, with the additional benefit that the overcondensation method preserves experimental material. In previous work, gas sorption experiments typically consist of a boundary adsorption isotherm up to a restricted maximum pressure (e.g. up to 0.995 p/p0). Following this there is a pseudo-boundary desorption isotherm, which is merely a descending curve since complete pore- filling with liquid-like condensate was not achieved. As a result of this conventional gas sorption alone cannot prove the complete pore size range up to large macro-pores. Gas overcondensation experiments can be expanded with gas sorption scanning curves which have successfully revealed advanced condensation effects, allowing probing of the inter- relationship and spatial juxtaposition of multi-scale porosities. Gas overcondensation and scanning loops were successfully used for the Utica and Bowland samples to reveal where additional percolations knee develop that are characteristic of a particular pore size within the wider pore network (Utica). Work on the Bowland was able to determine that there are some large macro-pores shielded by pore necks of <4nm; complimentary adsorption calorimetry work was able to relate this shielding to pore necks by calculating the mass transfer and thermokinetic properties of the samples.
Prior to the use of gas overcondensation mineralogy was assessed with the use of conventional gas sorption where results (Marcellus and Utica) showed an inverse relationship between carbonate and illite quantities (i.e. an increasing carbonate content was associated to a decreasing illite content). Utica surface areas demonstrated a strong correlation to illite quantity, whereas Marcellus surface areas demonstrated a weaker correlation to illite. For both samples there was good correlation to the total organic carbon. With the new information gained from gas overcondesation it has allowed for additional, and more advanced correlations to be made with other physical properties of shales such as the mineralogy. It was found that for the changeover period (Utica samples), from primarily clay to carbonaceous deposits, there was an associated growth in the disorder of the pore network over particular key length-scales. These length-scales were highlighted by percolation processes in the gas overcondensation and scanning curves. This peaking in disorder was also associated to a peak in total organic carbon content and the accessible porosity was shown to be dominated by the organic carbon phase.
Following the identification of this trend with the use of gas overceondensation and mineralogy, numerical analysis techniques were used to replicate these findings with the use of the homotattic patch model. It was established that with the use of conventional gas sorption (nitrogen) isotherms and isotherms for the pure mineral phases of the sample good results can be generated indicating the associated quantity of each mineral to the sample.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
