NMR T2 Relaxation and Pore-Level Fluid Partitioning: Bound Fluid Volume, T2 Cutoff Calibration, and Permeability Prediction in WCSB Reservoir Characterization

Key Takeaways

• T2 relaxation mechanism: why small pores produce short T2 and appear as bound fluid: The NMR T2 relaxation rate is dominated by the surface relaxation term rho_s × (S/V) in reservoir rocks, where S/V is proportional to the inverse of pore radius for approximately spherical pores. A 1-µm radius clay-lined pore (S/V ~3×10^6 m^-1) relaxes at roughly T2 = 1/(4×10^-6 × 3×10^6) = 0.08 s = 80 ms at rho_s = 4 µm/s — already approaching the bound-fluid range — while a 10-nm clay micropore (S/V ~3×10^8 m^-1) relaxes at T2 = 0.8 ms, far below the 33 ms sandstone cutoff and contributing entirely to bound fluid volume. This scaling is why clay content is the dominant control on bound fluid volume in sandstone reservoirs: illite-coated pore throats in Viking and Cardium sandstone convert intergranular macroporosity into effectively bound-fluid-hosting microporosity, reducing the free fluid index by 30-60% compared to clean quartz sandstone with equivalent total porosity.
• T2 cutoff calibration: centrifuge capillary pressure methodology and WCSB formation-specific values: The T2 cutoff for a specific formation is determined by centrifuging core plugs at a capillary pressure corresponding to pore-entry pressure at reservoir conditions (typically 100-150 psi for WCSB sandstones at 1,000-3,000 m depth), measuring the remaining saturation as Swir, and matching Swir against the NMR T2 distribution to find the cutoff where NMR-BVI = NMR-porosity × Swir. In WCSB Viking sandstone (clean, 25-32% phi, 100-1,000 mD), the T2 cutoff is typically 25-33 ms — close to the industry default. In tight Montney siltstone (8-12% phi, 0.001-0.1 mD), the T2 distribution rarely extends above 10 ms, so a 33 ms cutoff would classify virtually all porosity as bound fluid and overestimate BVI; a formation-specific T2 cutoff of 3-5 ms is more appropriate for Montney, where capillary pressure curves from MICP show pore-throat entry pressures of 1,000-10,000 psi confirming that essentially all Montney porosity is capillary-bound at any practical reservoir drawdown.
• Three-component fluid partitioning: clay-bound versus capillary-bound versus free fluid: The T2 distribution contains three overlapping populations corresponding to distinct fluid environments. Clay-bound water occupies T2 0.5-3 ms (electrostatically held on clay mineral surfaces, zero producibility, requires a TW greater than 12 s to fully polarize with CMR). Capillary-bound water occupies T2 3-33 ms in sandstone (retained by pore-throat capillary pressure, potentially producible at very high drawdown but not under normal reservoir conditions). Free fluid occupies T2 greater than 33 ms in sandstone and represents moveable water and hydrocarbons (including gas, which appears at T2 1-200 ms and can overlap with both capillary-bound and free ranges, requiring D-T2 diffusion-relaxation maps to separate from water). In shale gas reservoirs, a fourth component — gas in organic nanopores — contributes T2 0.1-1 ms and is indistinguishable from clay-bound water without complementary diffusion measurements, complicating BVI estimation.
• Coates permeability equation: calibration constants and applicability limits: The Coates permeability model k = (phi/C)^2 × (FFI/BVI)^2 performs well for conventional sandstones where the free-fluid index is greater than 5% and the BVI-to-FFI ratio captures the pore-size distribution controlling flow. It breaks down in formations where BVI approaches zero (clean, well-sorted sandstone with essentially all free fluid) or where FFI approaches zero (tight formations with no measurable free fluid). An alternative formulation by Kenyon and the SDR model k = a × phi^4 × T2LM^2 (where T2LM is the logarithmic mean of the T2 distribution) performs better in carbonates and in situations where the BVI is poorly constrained because the T2 distribution is broad. WCSB Cardium formation Coates calibration constants range from C = 8-12 for clean reservoir facies and C = 15-20 for clay-rich facies — reflecting the fact that clay-filled microporosity contributes BVI without contributing meaningfully to permeability, so higher BVI for the same total porosity yields lower permeability than the standard C = 10 formula would predict.
• NMR bound fluid in WCSB Montney and Duvernay tight reservoirs: implications for completability assessment: In WCSB Montney wells, NMR is run as part of the logging suite specifically to characterize the total porosity (including clay-bound water not seen by density-neutron logs in gas-bearing zones), the T2 distribution shape (narrowly peaked at 0.5-3 ms in tight siltstone vs broader distribution in cleaner zones), and the permeability prediction from Coates or SDR models for selecting the highest-flow-capacity stages for perforation in the multi-stage frac completion. A Montney horizontal well with 40 logged stages uses the NMR permeability profile to rank stages by completion quality, targeting stages where the Coates permeability exceeds 0.01 mD for priority perforation clusters — stages below this threshold are bypassed or given limited perforation clusters to avoid fracture-network complexity without drainage area. Post-completion production logs confirm that NMR-permeability-ranked stages contribute proportionally more to total well production than equally-spaced uniform perforation designs.