Diffusion

Key Takeaways

• Molecular diffusion in oil reservoirs operates on geological timescales and has produced the compositional gradients observed in the fluid columns of many large fields — in an oil column thousands of feet thick, gravity segregation causes heavy hydrocarbon components to concentrate at the base while lighter components accumulate at the top; molecular diffusion acts in opposition to this gravity segregation, tending to homogenize the composition; the equilibrium state depends on the relative rates of thermodynamic driving forces, and in thick, connected oil columns these forces sometimes produce a compositional gradient where the oil at the top is lighter (lower density, lower viscosity) than at the bottom; in some fields — particularly in the Middle East — the oil column shows a systematic API gravity increase from base to crest that reflects this thermodynamic equilibrium composition; producing the crestal light oil while leaving the denser base oil unproduced can lead to field performance that deviates from predictions based on average oil properties; understanding diffusion-controlled compositional equilibrium is relevant for PVT sample interpretation (a sample from one depth may not represent the fluid at a different depth in the same connected column) and for EOR design (CO2 minimum miscibility pressure depends on oil composition, which varies with depth).
• In dual-porosity fractured reservoirs, diffusion and imbibition drive matrix-fracture transfer that is the primary recovery mechanism beyond the fracture drainage — a naturally fractured carbonate reservoir consists of two interconnected flow systems: the fracture network (which has very high permeability but low storage capacity) and the rock matrix (which has high storage capacity but very low permeability); in a waterflood of a fractured reservoir, water preferentially flows through the fractures (bypassing the matrix) and quickly breaks through to producers, leaving most of the oil stored in the matrix uncontacted; matrix-fracture transfer (the exchange of oil and water between the matrix and the fracture) is driven by capillary imbibition (water from the fracture is spontaneously drawn into the water-wet matrix, displacing oil into the fracture by capillary pressure forces), gravity drainage (oil flows from the matrix down into the fracture by density contrast in a fracture above the matrix block), and molecular diffusion (dissolved components exchange between the matrix fluid and the fracture fluid along concentration gradients); diffusive transfer is particularly important for EOR methods involving miscible gas (CO2 or hydrocarbon gas) where the concentration gradient between the gas-saturated fracture and the residual oil in the matrix drives diffusion of gas components into the matrix and oil components into the fracture, progressively extracting matrix oil without requiring the pressure gradient that the low-permeability matrix cannot sustain at commercial flow rates.
• CO2 diffusion in EOR and storage applications determines the rate at which injected CO2 mixes with reservoir brine and oil, affecting both EOR performance and the long-term security of CO2 geological sequestration — in a CO2-EOR project, injected CO2 must mix with the crude oil to achieve the viscosity reduction and swelling that drive incremental recovery; in a water-alternating-gas (WAG) injection scheme, CO2 diffuses from the injected CO2 slugs through the water slugs that separate them from the oil, a process that is much slower than convective mixing but which becomes important over the long injection periods of multi-year EOR projects; in CO2 geological storage, the CO2 injected into a saline aquifer initially forms a discrete gas phase at the top of the storage formation (trapped by capillary forces and the overlying seal); over centuries to millennia, CO2 dissolves into the brine by diffusion from the gas-brine interface, increasing the brine density and causing convective overturn that accelerates further dissolution; this convective dissolution mechanism is an important long-term trapping mechanism for CO2 in geological storage, and modeling it requires accurate representation of the diffusion coefficient of CO2 in brine at reservoir temperature, pressure, and salinity conditions.
• NMR logging uses diffusion measurements to distinguish oil from water and estimate permeability in sandstone and carbonate reservoirs — the NMR log measures the T1 (longitudinal relaxation) and T2 (transverse relaxation) of the hydrogen nuclei in the pore fluids after applying a magnetic field pulse; the relaxation times depend on the pore size (surface relaxation dominates in small pores), the fluid viscosity, and the diffusion coefficient of the fluid in the applied magnetic field gradient; by applying a pulsed field gradient (CPMG pulse sequence with varying echo spacing) and measuring how the T2 relaxation changes with echo spacing, the NMR tool can measure the diffusion coefficient of the pore fluid directly; the diffusion coefficient of water, oil, and gas differ systematically: water has a diffusion coefficient of approximately 2 x 10-9 m2/s at reservoir conditions; live oil has diffusion coefficients 3-10 times lower depending on viscosity; gas has diffusion coefficients 100-1000 times higher than water; these differences allow the NMR diffusion measurement (using a two-dimensional NMR plot of T2 versus diffusion coefficient) to separate the water signal from the oil signal and the oil signal from the gas signal in the pore space, providing fluid typing information that complements the conventional resistivity-based water saturation calculation — particularly valuable in low-contrast, fresh-formation-water environments where resistivity contrast between oil and water is insufficient for reliable saturation calculation.
• Pressure diffusion in reservoir engineering — the propagation of a pressure disturbance through a permeable formation — obeys the same mathematical equation as Fick's molecular diffusion law, with hydraulic diffusivity replacing the molecular diffusion coefficient — the diffusion equation for pressure (del squared p = (phi x mu x Ct) / k x dp/dt) governs how a pressure pulse created by starting or stopping production spreads through the reservoir; the hydraulic diffusivity (k / phi x mu x Ct, where k is permeability, phi is porosity, mu is fluid viscosity, and Ct is total compressibility) determines the speed at which pressure information travels through the reservoir; high diffusivity formations (high permeability, low compressibility) transmit pressure pulses rapidly, causing pressure transient tests to detect boundaries quickly; low diffusivity formations (tight rock with high compressibility) transmit pressure slowly, requiring long test times to detect boundaries and achieve radial flow regimes; the analogy between pressure diffusion and molecular diffusion is more than mathematical — both reflect the same underlying physics of how information (whether concentration or pressure) spreads through a medium by random thermal motion or viscous drag, and the mathematical tools developed for one can be directly applied to the other.