Bulk Relaxation in NMR Logging: Free-Fluid Viscosity Interpretation and Fluid Typing in WCSB Heavy Oil, Bitumen, and Montney Tight Siltstone Formations

Key Takeaways

• Viscosity-T2 bulk relaxation relationship and in-situ heavy oil viscosity estimation in WCSB Athabasca and Cold Lake oil sands: The empirical relationship between crude oil T2,bulk and viscosity (T2,bulk = C × T / eta, where T is absolute temperature in Kelvin, eta is oil viscosity in cP, and C is approximately 0.004 cP × s/K for crude oils) provides a direct method for estimating in-situ viscosity from NMR log measurements. This is significant for WCSB Athabasca oil sands evaluation because in-situ oil viscosity at 10-15 degrees C reservoir temperature (500,000-5,000,000 cP) is far higher than API gravity alone indicates, and accurate viscosity estimation is required for SAGD steam chamber design, CSS (cyclic steam stimulation) injectivity modeling, and calculation of diluent volumes for bitumen pipeline transport. NMR T2 distributions measured in WCSB Athabasca core at reservoir temperature show T2 peaks at 1-5 ms for the bitumen phase and 30-100 ms for the pore water phase; the bitumen T2,bulk estimates in-situ bitumen viscosity at 12 degrees C of approximately 1.5-5 × 10^6 cP, validated against ASTM viscosity measurements on core extracts. As steam heats the reservoir in SAGD operations, bitumen T2,bulk lengthens as viscosity decreases: at 200 degrees C, bitumen viscosity drops to 2-10 cP and T2,bulk rises to 50-200 ms, approaching the water T2,bulk value and enabling steam-chamber monitoring by time-lapse NMR logging.
• Separation of bulk relaxation from surface relaxation in WCSB sandstone NMR log interpretation and its implication for porosity in clay-rich formations: In a porous medium, the total measured relaxation rate (1/T2,measured) is the sum of all three contributions: 1/T2,surface + 1/T2,bulk + 1/T2,diffusion. For large pores with abundant free fluid far from grain walls (WCSB Viking and Cardium clean sandstone at porosity above 20%), the surface relaxation rate (1/T2,surface = rho2 × S/V, where rho2 is the surface relaxivity and S/V is the pore surface-to-volume ratio) is small relative to bulk relaxation, so T2,measured approximately equals T2,bulk for the fluids present. For small pores (WCSB Montney tight siltstone at porosity below 8%) or clay-mineral-bound microporosity (Cardium shaly sandstone), S/V is very high and surface relaxation dominates, with T2,measured much shorter than T2,bulk. Distinguishing these regimes is essential for accurate NMR porosity calculation: clay micropore water (T2 approximately 0.5-3 ms) must be separated from capillary-bound water (T2 approximately 3-33 ms) and free fluid (T2 above 33 ms) using cutoff values calibrated to WCSB core capillary pressure data rather than assumed from bulk relaxation theory alone.
• Gas bulk relaxation in WCSB Montney tight formations and NMR fluid typing using Diffusion-T2 crossplots: Free methane gas has a bulk relaxation T2 that depends on pressure, temperature, and gas composition through the molecular collision rate governing spin-spin interactions. At WCSB Montney reservoir conditions (35 MPa, 80 degrees C), pure methane has T2,bulk approximately 50-80 ms; at lower pressure (10 MPa), T2,bulk,gas increases to 100-200 ms. The diffusion coefficient of gas (D_gas approximately 1-3 × 10^-8 m2/s in porous media) is much larger than for water (D_water approximately 1-3 × 10^-9 m2/s) or oil (D_oil approximately 1-10 × 10^-10 m2/s), so diffusion relaxation dominates gas T2 in NMR tools with significant magnetic field gradients. This contrast in diffusion coefficient between gas, oil, and water is exploited in the Diffusion-T2 (D-T2) crossplot technique: gas appears at very high D and moderate T2, oil at intermediate D and viscosity-dependent T2, and water at low D and short T2, enabling trilinear fluid volume calculation from a single WCSB Montney NMR log run with multi-TE acquisition that varies echo spacing to accentuate diffusion contrast between fluid types.
• T1 bulk relaxation and its use in WCSB NMR logging for improved heavy oil volume calculation: T1 (longitudinal or spin-lattice relaxation) describes the recovery of the nuclear magnetization toward thermal equilibrium after a radio-frequency pulse, and T1,bulk also scales approximately inversely with fluid viscosity: T1,bulk is approximately 1.0-1.3 × T2,bulk for crude oils and water at low field strength, while T1 is much longer than T2 for methane gas at reservoir conditions (T1,gas at 35 MPa approximately 3,000-5,000 ms, T2,gas approximately 50-80 ms). This T1/T2 ratio contrast between gas (T1/T2 much greater than 1) and liquids (T1/T2 approximately 1.0-1.3) provides a secondary fluid typing criterion: acquiring a T1-T2 map (using magnetization recovery followed by Carr-Purcell-Meiboom-Gill decay) allows gas to be identified by its very long T1 at short T2, independently of diffusion measurements. For WCSB Cold Lake heavy oil wells where T2,bulk of the oil overlaps the T2 of clay-bound water (both near 1-5 ms), switching to T1-based logging (saturation-recovery sequence) at the correct T1 recovery time allows the heavy oil signal (T1,bulk approximately 5-50 ms at 30 degrees C) to be separated from clay water (T1,surface approximately 0.5-5 ms), improving heavy oil porosity and saturation estimates in SAGD production monitoring where accurate oil volume is commercially critical.
• Temperature correction for bulk relaxation in WCSB NMR logging and the effect of downhole temperature on T2 interpretation: Bulk relaxation times increase with temperature for both liquids and gases (since molecular motion increases, reducing the correlation time for magnetic field fluctuations and thereby lengthening the relaxation time). For WCSB Cardium oil (viscosity approximately 8 cP at 20 degrees C surface conditions), the in-situ bulk T2 at 65 degrees C reservoir temperature is approximately 2-3 times longer than at surface temperature because viscosity decreases from 8 cP to approximately 2.5-4 cP at 65 degrees C, shifting the T2,bulk from approximately 50 ms (at surface) to 130-200 ms (at reservoir conditions). NMR logs run in situ at reservoir temperature therefore show longer T2 distributions for oil-bearing intervals than core NMR measurements at laboratory temperature, and T2 cutoffs calibrated from core capillary pressure experiments at room temperature must be temperature-corrected before applying them to in-situ WCSB log data. For WCSB oil sands (Athabasca bitumen), the temperature correction to T2,bulk is especially large: heating bitumen from 12 degrees C (reservoir) to 25 degrees C (NMR laboratory temperature) decreases viscosity by a factor of 10-100, lengthening T2,bulk by the same factor, meaning that lab NMR on bitumen core at 25 degrees C severely overestimates the in-situ NMR signal that would be recorded by a downhole tool at reservoir temperature and may lead to substantially wrong bitumen saturation calculations if not corrected.