Acquiring the NMR Bound Fluid Log: CPMG Pulse Sequences, CMR and MRIL Tool Design, T2 Distribution Inversion, and Permeability Calibration in WCSB Formation Evaluation
The bound fluid log is the nuclear magnetic resonance (NMR) log product that displays the spatial variation of bound fluid volume index (BVI) and free fluid index (FFI) as continuous depth functions derived from the T2 relaxation time distributions measured by an NMR logging tool during a wireline or LWD logging run, providing a direct pore-fluid partitioning measurement — distinguishing non-producible from potentially producible porosity — that is unavailable from any combination of conventional density, neutron, resistivity, or sonic logs, which measure total porosity and fluid saturation but cannot resolve whether the pore fluid is held by capillary or electrostatic forces beyond the reach of reservoir pressure drawdown. The NMR tool creates a static magnetic field B0 using a permanent magnet assembly, which aligns the magnetic moments of hydrogen protons in pore fluids; a radio-frequency (RF) antenna surrounding the tool then transmits a 90-degree tip pulse at the Larmor resonance frequency f = gamma × B0/(2pi) — for a typical CMR (Combinable Magnetic Resonance, Schlumberger) or MRIL (Magnetic Resonance Imaging Log, Halliburton) operating at 0.048-0.052 Tesla field strength, this frequency is approximately 2 MHz — rotating the proton moments into the transverse plane where they precess and generate a detectable RF signal. The CPMG (Carr-Purcell-Meiboom-Gill) pulse sequence then applies a series of 180-degree refocusing pulses at intervals of TE/2 (inter-echo spacing TE typically 0.2-1.2 ms for CMR), generating a train of spin echoes whose amplitudes decay exponentially as proton moments dephase due to surface relaxation and diffusion relaxation; the inter-echo spacing TE controls sensitivity to fast-relaxing components (shorter TE resolves clay-bound water at T2 0.5-3 ms, longer TE improves signal-to-noise for free fluid but aliases short-T2 clay water), and the polarization wait time TW (1-12 seconds between successive CPMG trains) controls how completely the long-T2 free fluid populations are polarized before measurement (insufficient TW causes underestimation of free fluid and total porosity). The echo-train amplitude decay is numerically inverted using constrained regularization algorithms (the industry-standard method is the Butler-Reeds-Dawson algorithm implemented with Tikhonov regularization) to produce the T2 distribution, from which BVI (integral of T2 distribution below T2cut) and FFI (integral above T2cut) are computed at each depth; the resulting bound fluid log typically displays BVI and FFI as separate tracks alongside total CMR porosity, with a permeability log computed from the Coates model (k = (phi/C)^2 × (FFI/BVI)^2) or the SDR model (k = a × phi^4 × T2LM^2) for comparison with core plug data. In WCSB exploration, the bound fluid log is run on Montney horizontal wells as part of a formation evaluation quad combo (GR, resistivity, density-neutron, CMR) specifically to characterize the distribution of submicron matrix porosity (all capillary-bound, T2 less than 5 ms) versus fracture-enhanced zones (where natural open fractures add free-fluid NMR porosity at T2 greater than 100 ms) that cannot be distinguished from conventional density-neutron logs, enabling stage-by-stage completion ranking without the cost of continuous coring through a 3,000-metre horizontal lateral.
Key Takeaways
- CMR vs MRIL tool design and depth of investigation: The Schlumberger CMR tool and the Halliburton MRIL tool operate on the same CPMG physics but differ in magnet design, depth of investigation, and logging configuration. CMR uses a permanent magnet that creates a homogeneous sensitive volume at approximately 25-38 mm from the tool face, with a thin cylindrical shell geometry that provides a depth of investigation close enough to exclude mudcake (avoiding contamination of the NMR signal by invaded-zone fluids in highly permeable formations). MRIL uses a different magnet design operating at slightly lower field strength and offers multiple investigation depths selectable by frequency — allowing separate measurement of the flushed zone (near borehole, invaded by mud filtrate) and the virgin zone (beyond invasion) in high-permeability permeable formations where invasion depth exceeds 20-30 cm. For WCSB tight gas wells where invasion is negligible (permeability less than 0.1 mD), both tools measure essentially the same uninvaded formation, and CMR is the dominant tool choice due to its combination run compatibility with the sonic and density tools without a dedicated pass.
- CPMG sequence parameters: inter-echo spacing TE, wait time TW, and their effect on measurement quality: The TE controls the minimum detectable T2 value: to avoid aliasing short-T2 clay-bound water components, TE must be shorter than approximately one-third of the shortest T2 of interest; for WCSB Montney siltstone where clay-bound water at T2 = 0.5 ms is significant, TE = 0.2 ms is required (CMR standard minimum), while for clean Viking sandstone where the shortest component is 3-5 ms, TE = 0.5 ms is adequate with better signal-to-noise. The TW determines whether long-T2 free fluid components are fully polarized; for oil at T2 = 500-1,000 ms, a TW of at least 3T2 = 1,500-3,000 ms (1.5-3 seconds) is required for full signal recovery, and for gas at T2 = 1,000-3,000 ms, a TW of 6-12 seconds may be needed — this is why CMR logging speed is typically limited to 3-5 m/min, much slower than conventional triple-combo logging at 15-25 m/min, and why CMR is often the time-limiting tool in a multi-tool logging run.
- T2 distribution inversion: regularization, resolution, and uncertainty: The inversion of the echo train decay to a continuous T2 distribution is mathematically ill-posed — many different T2 distributions can fit the measured echo train within the noise level — and regularization (Tikhonov regularization or maximum entropy) is applied to produce the smoothest T2 distribution consistent with the data. The resolution of the T2 distribution (minimum separation between two T2 components that can be individually resolved) is approximately half a decade on the logarithmic T2 axis (e.g., components at 10 ms and 100 ms can be resolved, but components at 20 ms and 30 ms may not be). This limited resolution means that overlapping clay-bound and capillary-bound components (both below 33 ms) cannot always be separated without complementary mineralogy logs (spectroscopy or element capture scintillation) to estimate clay volume independently. The propagated uncertainty in BVI from inversion noise is typically 1-3 porosity units, which is acceptable for permeability prediction but must be considered when the bound fluid log is used for material balance calculations of producible porosity in a tight reservoir.
- Permeability calibration from CMR to core and well test data: The Coates permeability transform uses formation-specific calibration constants (C for the Coates model, a for the SDR model) that must be determined from core plug measurements in each formation. For WCSB Cardium formation, the Coates C typically ranges from 8-12 for clean reservoir facies and increases to 15-20 in clay-rich facies. Calibration requires at minimum 10-20 paired core plug permeability and NMR T2 distribution measurements from the same formation to establish the regression. If no core is available, default constants are used with acknowledged uncertainty of one to two orders of magnitude in predicted permeability. Formation-scale permeability from well test (DST pressure buildup or DFIT falloff) is the gold standard for calibration and should be compared against the depth-averaged CMR permeability over the tested interval; systematic over- or under-prediction identifies whether the T2 cutoff or the Coates constant needs adjustment for local calibration.
- Diffusion-relaxation (D-T2) NMR for fluid typing beyond the bound fluid log: Standard CPMG measurements cannot distinguish gas from clay-bound water (both appear at T2 less than 5 ms) or light oil from water (overlapping T2 ranges), limiting the bound fluid log to pore-size partitioning rather than fluid typing. Advanced NMR acquisition modes — two-dimensional D-T2 measurements that vary the encoding gradient strength alongside the CPMG sequence — separate fluids by their diffusion coefficient as well as T2, enabling gas (high diffusivity, D ~10^-5 m^2/s), water (intermediate, D ~2×10^-9 m^2/s), and oil (low diffusivity, D ~10^-10 to 10^-11 m^2/s) to be identified on a 2D map. In WCSB Duvernay tight oil wells where the oil T2 and water T2 overlap in the 5-50 ms range, D-T2 NMR from MRIL or CMR multi-wait-time acquisition resolves the fluid typing without relying solely on resistivity-based Sw interpretation, providing a direct fluid volume measurement for reserves estimation in the pre-completion formation evaluation suite.
CMR Bound Fluid Log Integration in a Montney Formation Evaluation Well
A northeast BC Montney horizontal well runs a CMR log (TE = 0.3 ms, TW = 8 s, 10 stacks per depth level) through the 280-m vertical Montney section before kicking off for the horizontal lateral. CMR total porosity averages 9.2% across the Montney — 2.0% higher than the density-neutron crossplot porosity of 7.2% because the density log sees gas-corrected porosity while CMR sees total hydrogen (including clay-bound water). T2 distribution: 78% of NMR porosity below 3 ms (clay-bound), 18% between 3-33 ms (capillary-bound), 4% above 33 ms (free fluid). BVI = 8.8%, FFI = 0.4%. Coates permeability (using locally calibrated C = 5 for Montney): k = (0.092/5)^2 × (0.4/8.8)^2 = 3.4×10^-4 × 2.1×10^-3 = 7×10^-7 Darcy = 0.0007 mD. DFIT interpretation from offset well: matrix permeability 0.00015 mD — within the expected uncertainty of the Coates prediction in a submicrodarcy formation where the model approaches its valid lower limit. The bound fluid log identifies three 15-25 m intervals where FFI is 1-2% (vs 0.4% average) corresponding to natural fracture corridors on borehole image — these intervals are prioritized for perforation cluster placement in the 35-stage Montney frac design.
Fast Facts
The MRIL (Magnetic Resonance Imaging Log) tool introduced by NUMAR Corporation in 1991 was the first commercially successful pulsed NMR logging tool designed for wireline formation evaluation — displacing the older continuous-wave NMR tools (the Free Fluid Index log of the 1960s-1970s) that could only measure total proton density and had no ability to resolve T2 distributions. NUMAR was acquired by Halliburton in 1997, and Schlumberger introduced the CMR tool in the same era. The availability of commercial pulsed NMR tools through the 1990s enabled the Coates permeability equation and T2 cutoff methodology to move from laboratory concept to routine wireline log interpretation, and by the early 2000s the CMR and MRIL bound fluid logs were standard formation evaluation components in WCSB Devonian carbonate and Cretaceous sandstone exploration programs.
Related Terms
The physical concept of bound fluid — the T2 relaxation mechanism, T2 cutoff calibration against capillary pressure, and the partitioning of clay-bound versus capillary-bound versus free fluid — is described under bound fluid, where the Coates and SDR permeability equations and their formation-specific calibration constants are explained alongside WCSB Montney and Viking application examples. The clay-bound water component of total bound fluid, held by electrostatic forces rather than capillary pressure and described by the cation exchange capacity framework, is covered under bound water, where the Waxman-Smits shaly sand resistivity correction that accounts for clay-water conductance in WCSB Viking and Cardium formation Sw determination is detailed. The broader NMR measurement principle and how NMR porosity compares with density and neutron porosity in gas-bearing and clay-rich formations is described under nuclear magnetic resonance.