The quantity of light hydrocarbon and natural gas in tight-oil and gas shales respectively is one of the primary indicators of reservoir quality (RQ). The measurement of RQ therefore depends on the ability to distinguish the quantity of the light oil or gas from other fractions of the total organic carbon, namely the immobile hydrocarbons such as kerogen, bitumen, heavy oil and formation water. Additionally, the separation of the oil into fractions hosted in organic versus inorganic porosity is important for determining the potentially producible fraction. Although multidimensional diffusion-relaxation correlation experiments can distinguish hydrocarbons from other fluids in conventional reservoirs, their use is restricted owing to the presence of short relaxation times in most tight oil and gas shale plays. We demonstrate the feasibility of nuclear magnetic resonance (NMR) relaxometry for determining the different constituents in shale based on the frequency dependence of their relaxation times. Two-dimensional NMR T1-T2 experiments take advantage of this frequency dependence to provide a robust method for the clear identification of the different fractions.
We determined the 2D T1-T2 relaxation distributions of the different fluid fractions in unconventional shale plays, namely the heavy oil fraction, light oil in organic versus inorganic porosity (and contrast it to gas in organic versus inorganic porosity), and bound versus free water. We discuss the origin of the T1-T2 ratios as measured by low-field NMR based on the intrinsic relaxation mechanisms in each component: spin rotation for the gas molecules, intramolecular and intermolecular dipolar relaxation for heavy oil and bitumen, and, dipolar relaxation modulated by long residence times for oil in organic versus inorganic porosity. We also discuss the practical implications of these relaxation mechanisms by determining which fractions are identifiable in low-field NMR logs, thereby setting limits to fluid typing and saturation determinations, downhole.
A high-pressure fluid re-saturation methodology was used to mimic formation properties in native-state samples. We reveal how experiments on these re-saturated samples together with log analysis enable in-situ estimates of the potentially producible fluid volumes, thereby aiding well placement, completion methodologies and production predictions. This technique uniquely enables the segregation of the total organic porosity from the inorganic porosity based on the NMR T1-T2 maps of the re-saturated samples. We discuss some of the strengths of this technique, especially identification of the fraction of producible oil, and also the limitations for downhole fluid-typing applications.
