Electrical Resistivity Tomography (ERT) for Soil, Concrete, and Rock: Principles, Electrode Arrays, and Engineering Applications

TL;DR

Soil, concrete, and rock are all porous conductive media: ions in the pore water carry the current while the solid skeleton is nearly insulating. Anything that changes pore-water content, salinity, or connectivity — rainfall infiltration, contaminant migration, concrete curing, chloride ingress, crack growth in rock — leaves a trace in the resistivity distribution. Electrical Resistivity Tomography (ERT) exploits exactly this: it inverts the internal resistivity distribution from boundary electrode measurements, giving non-destructive, planar, repeatable imaging of all three media.

1. Why soil, concrete, and rock are natural fits for resistivity imaging

The three media share the same electrical structure:

Two-phase conduction:
├─ Solid skeleton (soil grains / cement paste / rock minerals): ρ ≈ 10³–10⁶ Ω·m, nearly insulating
├─ Pore water (dissolved ions): ρ ≈ 1–10² Ω·m, the main conduction path
└─ Bulk resistivity is jointly controlled by porosity, water content,
   pore-water salinity, and temperature

Because conduction is carried entirely by the pore liquid phase, the electrical state tracks moisture and salt movement directly. ERT does not need to “see” the solid itself — it characterizes the engineering quantity of interest through the liquid phase:

• Soil: moisture content, seepage paths, contaminant plumes, grout spreading
• Rock: crack growth, loosened zones, leachate advance in in-situ leaching
• Concrete: moisture migration and curing uniformity, chloride fronts, internal defects

Archie (1942) put this relationship into quantitative form: bulk resistivity relates to porosity, water saturation, and pore-water conductivity through the cementation and saturation exponents — the starting point for quantitative interpretation in soil and rock. Concrete is more complex because the cement pore solution is inherently alkaline and conductive, but the “moisture and salt dominate conduction” framework still holds.

2. The three media at a glance

Dimension	Soil	Rock	Concrete
Typical scale	m–100 m (surface profiles) / cm–m (cross-borehole)	cm (specimens) – 10 m (tunnels)	cm–m (members)
Resistivity magnitude	1–10⁴ Ω·m (varies strongly with moisture/salinity)	10²–10⁶ Ω·m (high-resistivity background)	10²–10⁵ Ω·m (varies with w/c ratio and age)
Electrode form	Steel stakes at the surface, cross-borehole arrays, probes	Specimen clamps, borehole/tunnel-wall electrodes	Surface pads / embedded electrode nets
Excitation frequency	Low (< 10 Hz–kHz, to reduce EM coupling and IP effects)	Low + attention to high-voltage insulation	mHz–kHz (suited to time-lapse)
Typical inversion target	Moisture field, seepage front, contaminant distribution	Damage / loosened zone, grout, leachate extent	Moisture migration, defects, corrosion regions

All three share the same forward problem (electrode model + FEM) and inversion framework; the differences lie mainly in electrode technology, measurement scale, and prior information — meaning a well-designed measurement system can be reused across scenarios.

3. Soil: from profile mapping to time-lapse monitoring

3.1 What it can measure

• Moisture and seepage monitoring: rainfall infiltration fronts, irrigation percolation, dam seepage paths. With a saline tracer injection and time-lapse imaging, water movement can be tracked directly (Daily et al., 1992)
• Contaminants and salinization: landfill leachate, seawater intrusion, saline-soil extent — high-ion regions appear as clear low-resistivity anomalies
• Foundations and underground structures: auxiliary delineation of voids, loose zones, and grout spread

3.2 Electrode arrangements

Choose geometry by target:
├─ Surface profile (2D): Wenner / Schlumberger arrays scanned along a line
│   └─ Fast deployment; resolution decays with depth
├─ Full array (3D): electrodes on a grid, full dataset per epoch
│   └─ For areal targets (seepage zones, contaminant plumes)
├─ Cross-borehole CT (2D/3D): electrode arrays in cased boreholes
│   └─ Much better deep resolution than surface profiles; cost is the drilling
└─ Probe/rod arrays: root zone, laboratory soil columns
    └─ Small electrode spacing for root water uptake and column tests

3.3 Quantitative interpretation

Moisture inversion depends on calibrated Archie parameters (a, m, n): calibrate on soil samples in the lab, or convert the unknown spatial prior into a change-detection problem with time-lapse differencing. A common pitfall is reading absolute resistivity as moisture directly — spatial variability of temperature and pore-water conductivity distorts absolute interpretation. Time-lapse differencing plus a few calibration boreholes is the more robust combination (Samouëlian et al., 2005 reviews the soil-science application landscape).

4. Rock: imaging against a high-resistivity background

4.1 What it can measure

• Specimen scale: electrical response during crack initiation–propagation–coalescence in uniaxial/triaxial loading; freeze–thaw and chemical-dissolution damage evolution
• Engineering scale: tunnel surrounding-rock loosened-zone thickness, grout distribution, leachate advance in in-situ leaching mines
• Hydrogeology: preferential flow channels in fractured rock masses

4.2 Key differences from soil surveys

The difficulty in rock ERT is not the algorithm but the electrical magnitude:

1. High-resistivity background: dry rock is 2–4 orders of magnitude more resistive than soil, so the useful signal (ΔV) is weaker, placing higher demands on input impedance, common-mode rejection, and electrode contact
2. Contact impedance: hard-rock surface electrodes have small contact areas and large contact resistance; conductive coupling compound or light moistening is needed, and the complete electrode model should be used in inversion to account for it
3. Injection voltage: maintaining sufficient current through a high-resistance loop raises the drive voltage — mind electrode polarization and safety isolation

In specimen tests, the crack mechanism is “conductive cross-section cut off”: in saturated specimens the damaged zone appears high-resistivity, and returns to low-resistivity after grouting — so ERT can track the spatial evolution of damage without sectioning the specimen, complementing the point-wise localization of acoustic emission with planar imaging.

5. Concrete: moisture, chlorides, and internal defects

5.1 Three typical tasks

Task	Electrical mechanism	Typical array
Moisture migration / curing	Hydration consumes free water → resistivity rises with age; surfaces dry faster than cores	Surface pad array, time-lapse
Chloride ingress / corrosion risk	Cl⁻ lowers pore-solution resistivity; the low-resistivity zone marks the ingress front	Near-surface array or cross-rebar plane
Internal defects (honeycombing / cracks / voids)	Defects change the conductive cross-section: saturated cracks are conductive, dry cracks resistive	Ring / double-sided arrays

5.2 A reference research paradigm

Karhunen et al. (2010) used a 16-electrode ERT setup to locate polyurethane blocks, a rebar bar, and a plastic plate simulating a crack inside concrete specimens, demonstrating the feasibility of ERT for imaging building structures; our article on ERT imaging of concrete covers this study in detail, including array design and resolution discussion. In practice, concrete ERT is often combined with ultrasound and ground-penetrating radar (GPR): ultrasound is sensitive to mechanical defects, GPR to interfaces and metal, and ERT to moisture and salts — complementary rather than interchangeable.

6. Data acquisition and inversion practice

The three media share the same technical stack; the critical points concentrate in four places:

Acquisition → inversion pipeline:
├─ Excitation & measurement
│   ├─ Low-frequency square/sine excitation to suppress EM coupling and polarization
│   ├─ Current reversal + averaging removes DC drift; for high-resistivity targets
│   │   check input impedance on the order of ≥ 10⁸ Ω
│   └─ Four-terminal measurement (independent current injection and voltage sensing) is standard
├─ Forward problem
│   ├─ Complete electrode model (CEM) includes contact impedance
│   └─ FEM mesh refined near electrodes
├─ Inverse problem
│   ├─ Tikhonov regularization against ill-posedness; structural priors
│   │   (reference model / boundaries) improve deep recovery
│   └─ For time-lapse, prefer difference imaging to cancel systematic error
└─ Quality control
    ├─ Periodic contact-impedance checks; a bad electrode produces radial artifacts
    └─ Reciprocity checks (swap injection–measurement pairs) to verify data consistency

The ill-posedness and soft-field nature of the inverse problem, and how to read the sensitivity matrix, are covered in our articles on soft-field ill-posedness and on sensitivity matrices.

7. Equipment selection and implementation notes

When selecting an ERT system for soil / rock / concrete:

1. Channel count and electrodes: 32–64 electrodes are a practical starting point for shallow profiles; cross-borehole and 3D monitoring scale with target size, while 16–32 electrodes suffice for structural members
2. Input impedance and measurement resolution: for high-resistivity targets (dry soil, rock, lean concrete), examine input impedance and voltage resolution before frame rate
3. Electrode technology: soil stake electrodes need insertion stiffness; rock/concrete pad electrodes need coupling compound and long-term stability; long-term monitoring must account for polarization drift and corrosion
4. Environmental robustness: field deployments need temperature-drift management and lightning protection; structural monitoring needs to manage electrical interaction between embedded electrodes and rebar
5. Data pipeline: time-lapse monitoring needs stable synchronization and storage; raw data export (.bin/.csv) enables MATLAB / Python post-processing

Items 3 and 4 often decide project success yet have no standardized answer — they call for customization to the medium and site. Our article on custom ET sensor design describes the full engineering workflow, and our industrial multi-plane ERT systems can serve as a platform for laboratory-scale method validation.

8. FAQ

Q: How deep can ERT see, and what resolution? Resolution decays with electrode spacing and depth: for surface profiles, the resolution cell at target depth is typically on the order of 10–30% of depth; cross-borehole CT reaches 5–10% of the borehole spacing. Any concrete number requires the geometry and noise level first — one more reason we recommend a forward-problem simulation before fielding.

Q: What happens with poor electrode contact? At best larger data variance; at worst radial artifacts centered on the bad electrode. For long-term monitoring, make contact-impedance checks part of the maintenance plan.

Q: How large is the temperature effect? Pore-water conductivity changes roughly 2%/°C, so seasonal temperature swings create “false changes” unrelated to seepage in time-lapse data. A deep, thermally stable region can serve as a reference baseline for correction.

Q: How does it compare with GPR and ultrasound? GPR suits interfaces and metallic objects via dielectric contrast; ultrasound suits mechanical defects given coupling; ERT is sensitive to moisture, salts, and connectivity and fits long-term unattended monitoring. Water-rich clay strongly attenuates GPR — exactly where ERT performs comparatively well.

Q: How long can it monitor? Electrode and cable lifetime is the main limit. Soil stakes last months to a few years; embedded electrode nets with anti-corrosion design can support multi-year campaigns — accepting that electrode drift is part of the data, managed with difference imaging and periodic calibration.

9. Further reading

• Resistivity Tomography, Resistance Tomography, and Electrical Impedance Tomography
• ERT Technology Guide
• ERT Imaging of Concrete
• ERT Testing of Porous Media
• Soft Fields and Ill-Posedness in ET
• Custom Design of ET Sensors

Key references: G. E. Archie. The electrical resistivity log as an aid in determining some reservoir characteristics. Trans. AIME 146 (1942) 54–62. A. Samouëlian, I. Cousin, A. Tabbagh, A. Bruand, G. Richard. Electrical resistivity survey in soil science: a review. Soil and Tillage Research 83 (2005) 173–193. W. Daily, A. Ramirez, D. LaBrecque, J. Nitao. Electrical resistivity tomography of vadose water movement. Water Resources Research 28 (1992) 1429–1442. M. H. Loke, R. D. Barker. Rapid least-squares inversion of apparent resistivity pseudosections by a quasi-Newton method. Geophysical Prospecting 44 (1996) 131–152. K. Karhunen, A. Seppänen, A. Lehikoinen, P. J. M. Monteiro, J. P. Kaipio. Electrical Resistance Tomography imaging of concrete. Cement and Concrete Research 40 (2010) 137–145. doi:10.1016/j.cemconres.2009.08.023 A. Binley, S. S. Hubbard, J. A. Huisman, A. Revil, D. Robinson, V. Singha, L. D. Slater. The emergence of hydrogeophysics for improved understanding of subsurface processes over multiple scales. Water Resources Research 51 (2015) 3837–3866.