Array Induction: Definition, Resistivity Logging, and Invasion Profiling
An array induction tool is an electromagnetic wireline logging instrument that simultaneously acquires resistivity measurements at five to seven distinct radial depths of investigation — typically ranging from 10 inches to 90 inches into the formation — by transmitting a high-frequency alternating current through transmitter coils and measuring the induced secondary magnetic field at multiple receiver coil arrays spaced along the mandrel. Unlike a conventional dual-induction tool that yields only two curves, an array induction tool processes all sensor responses simultaneously through a software inversion algorithm to generate a continuous radial resistivity profile from the borehole wall to the undisturbed formation, resolving the flushed zone resistivity (Rxo), the transition zone, and the true formation resistivity (Rt) in a single pass. The technique is most effective in non-conductive or low-salinity drilling fluids — fresh-water-based muds, oil-based muds, and synthetic-based muds — where the borehole fluid does not short-circuit the electromagnetic signal, and it forms the standard resistivity method for fluid identification, hydrocarbon saturation calculations, and invasion diagnostics across thousands of wells in the Western Canada Sedimentary Basin drilled with freshwater or polymer-based fluid systems. The processed output is presented as a suite of focused pseudo-curves at fixed depths of investigation (commonly labelled 10-inch, 20-inch, 30-inch, 60-inch, and 90-inch), and the separation and convergence pattern of those curves on a log track is the primary field indicator of invasion geometry, moveable hydrocarbons, and mud-filtrate contamination extent.
Key Takeaways
- Simultaneous multi-depth acquisition: A modern array induction tool carries one or two transmitter coils and an array of five to seven receiver coil pairs spaced from roughly 6 inches to 72 inches along the mandrel, each pair operating at a slightly different frequency (typically 10 kHz to 200 kHz) to maximise sensitivity at the corresponding radial depth. All channels are recorded simultaneously on every depth sample, typically at 0.1-foot resolution, so the complete radial resistivity profile is captured in one logging pass without requiring multiple runs or tool repositioning. The Schlumberger Array Induction Tool (AIT), introduced commercially in 1991, established the industry benchmark with six depths of investigation (AHO10, AHO20, AHO30, AHO60, AHO90 nomenclature) and has been followed by equivalent platforms from Halliburton (High Definition Induction Log, HDIL), Baker Hughes (High-Definition Formation Evaluation Tool, HDFET), and other service companies. The raw voltage responses are not directly interpretable as resistivity curves; they are processed through Backus-Gilbert or Tikhonov-regularised inversion algorithms that honour formation physics to convert raw signals into the resistivity-at-depth suite presented on the log.
- Operating environment — freshwater and oil-based muds: Induction tools transmit and detect electromagnetic fields propagating through formation rock, not through the borehole fluid. In air-filled boreholes or boreholes with low-conductivity fluids (resistivity above roughly 1 ohm-m), the tool response is dominated by formation current paths and the multi-depth suite is reliable. In saline water-based muds with resistivity below about 0.3 ohm-m, the highly conductive mud column creates a short-circuit path inside the borehole that overwhelms the weak formation signal at deep depths of investigation, causing the deep curves to read falsely low and the tornado chart inversion to diverge. In the WCSB, most Cardium and Viking wells drilled with freshwater or low-salinity KCl-polymer muds yield excellent array induction quality, whereas wells drilled into Devonian carbonates with saturated KCl or CaCl2 brines to suppress swelling shales have routinely switched to array laterolog tools. The minimum borehole fluid resistivity threshold varies with tool design but most array induction specifications cite 0.03 ohm-m as the practical lower bound below which significant borehole correction is required.
- Tornado chart inversion for Rt, Rxo, and di: The standard interpretation workflow reads the deep-reading curve (90-inch) as the initial Rt estimate and cross-plots the ratio of a shallow curve to a deep curve (typically 10-inch/90-inch or 20-inch/90-inch) against the absolute value of the deep curve on a bilinear or log-log grid called a tornado chart. Each tornado chart contour represents a constant true resistivity (Rt), a constant flushed-zone resistivity (Rxo), or a constant invasion diameter (di), and the unique intersection of the two ratios resolves all three unknowns simultaneously. Positive separation (shallow reads higher than deep) indicates resistive invasion by oil-based mud filtrate displacing saline formation water — the classic signature of a water-wet hydrocarbon pay zone. Negative separation (deep reads higher than shallow) indicates conductive invasion, commonly seen in water zones where saline mud filtrate is more resistive than the pore fluid and the deep resistivity is depressed by invaded formation water. Parallel curves (no separation) indicate negligible invasion, either because the formation is tight and impermeable or because the filtrate resistivity matches the formation water resistivity.
- Dip and thin-bed effects on array induction response: Induction tools are designed for horizontal or near-horizontal current flow in thick, horizontally bedded formations. When the relative dip between the tool axis and bedding exceeds roughly 30 degrees — which occurs routinely in directional and horizontal wells — the transmitter-receiver geometry cuts across multiple beds simultaneously, and the shoulder-bed effect causes each depth-of-investigation curve to respond to a weighted average of resistivities above and below the current bed boundary. The shoulder-bed effect is most severe for the deeply-reading 90-inch curve (which averages 1.5 to 2 feet above and below the tool position) and least severe for the 10-inch curve (less than 0.5-foot vertical resolution). In horizontal Montney or Cardium wells where dip angles routinely exceed 85 to 89 degrees from vertical, the apparent separation between shallow and deep induction curves is partly a bed-boundary artefact rather than a pure invasion signal, and 2D or 3D borehole-image-guided inversion must be applied to deconvolve the true radial profile from the dip-induced response. Newer tools incorporate triaxial coils and tensor induction to provide azimuthal anisotropy and corrected logs in deviated wells.
- Integration with petrophysical workflows and saturation calculations: The 90-inch deep array induction curve is the primary Rt input to the Archie equation (Sw = [(a x Rw) / (phim x Rt)]1/n) for water saturation calculation, while the 10-inch curve provides Rxo for the flushed-zone saturation (Sxo) used in the bulk volume water (BVW) and moveable hydrocarbon index (MHI = 1 - Sw/Sxo) crossplots. In the WCSB Viking play at Redwater, Provost, and Lloydminster, array induction logs routinely detect 10 to 15 per cent water saturation in 15 to 22 per cent porosity sands and yield Rt values of 50 to 800 ohm-m in pay, contrasted with 2 to 10 ohm-m in adjacent water-saturated sands. The flushed-zone analysis is particularly valuable in Viking wells with short-radius perforations where the engineer needs to confirm that the resistivity contrast is driven by true hydrocarbon saturation rather than by shallow invasion of conductive mud filtrate — a distinction that the single-curve dual-induction tool cannot resolve without additional microresistivity measurements.
How Array Induction Tools Work: Coil Geometry and Signal Processing
The fundamental physics of induction logging rests on Faraday's law and Biot-Savart relations. A transmitter coil energised with an alternating current at frequency f (typically 20 kHz for the main transmitter on a modern array induction tool) generates a primary magnetic field that diffuses into the surrounding formation. Where formation conductivity is non-zero, the primary magnetic field induces circulating eddy currents in horizontal planes (in a horizontal borehole, these planes are perpendicular to the tool axis), and those eddy currents generate a secondary magnetic field. The receiver coils detect both the in-phase (real) component of the secondary field, proportional to formation conductivity at shallow investigation depths, and the out-of-phase (quadrature) component, which contains geometrical focusing information. Unlike conventional induction tools that subtracted a fixed "bucking coil" response to deepen the investigation window, array induction tools record all receiver responses individually and apply digital focusing in software — a technique called "software focusing" or "data-driven focusing" — allowing the petrophysicist to request multiple different radial depth profiles from a single raw data set.
Each depth-of-investigation channel on an array induction tool has a defined geometric factor (also called the radial response function or sensitivity kernel) that describes what fraction of the total measured conductivity originates at each radial distance from the borehole. For a 10-inch channel, roughly 80 per cent of the signal comes from within 10 inches of the borehole wall — predominantly the flushed and invaded zone. For a 90-inch channel, the signal is spread over a much wider radial range, with meaningful sensitivity extending to 4 to 5 feet from the wall, and the borehole contribution is typically less than 5 per cent provided the mud resistivity is above 0.1 ohm-m. The Backus-Gilbert inversion used in commercial processing optimises a set of linear weights applied to the raw array responses so that the composite response approximates a target geometric factor (sharp, well-focused, centred at the desired depth of investigation) while simultaneously suppressing borehole and shoulder-bed contributions. The result is a set of pseudo-depth-focused curves that behave as if the tool had a single, ideal coil system at each depth of investigation, even though no single physical coil achieves those focusing properties.
Borehole correction is a mandatory processing step before any quantitative interpretation. The borehole signal arises because the mud column is part of the electromagnetic measurement circuit; its contribution is computed from the logged caliper diameter, mud resistivity from the drilling fluid sample or from the shallow curve at a known mud-saturated formation, and the tool's theoretically derived borehole geometric factor. In 8.5-inch boreholes drilled with freshwater polymer mud (Rmud = 2 to 10 ohm-m), borehole corrections to the deep 90-inch curve are typically less than 3 per cent and are negligible for practical purposes. In rugose or washed-out boreholes (caliper above 12 inches) or in high-salinity mud, borehole corrections can reach 20 to 40 per cent on the 10-inch curve and the raw log is unreliable for Rxo estimation. Some WCSB operators running array induction in the Colorado Group above Cardium pay require full borehole correction at every depth step because the swelling shale above the reservoir washes out to 11 to 14 inches even when drilled with inhibitive fluid, and the variable caliper makes the uncorrected 10-inch curve track the borehole enlargement rather than the formation.