Surface Impedance Measurements Enable Non-Destructive Assessment of Steel Corrosion in Concrete

Steel reinforcement gives concrete structures much of their strength, but the steel can corrode over time and cause damage that may require expensive repairs. Checking the condition of the rebar is difficult because it is hidden inside the concrete. Some electrochemical methods use sensors embedded in the concrete, and others need a direct electrical connection to the steel. Embedded sensors are also limited to the locations where they were installed. Surface sensors are more flexible, but making a direct electrical connection to the rebar usually means drilling through the concrete. Indirect impedance spectroscopy offers another possibility. A set of electrodes placed on the concrete surface applies an alternating electrical signal and measures the resulting voltage without making direct electrical contact with the rebar. The measured impedance contains information about the steel–concrete interface and, in principle, about the corrosion state. The difficulty is separating that information from the effects of concrete resistivity, electrode arrangement and structural geometry. Existing indirect methods have shown that corrosion affects the measured response, but converting those measurements into reliable quantitative corrosion parameters remains challenging.

In a recent research paper published in Electrochimica Acta, Dr. Chenghao Hang and Professor Guofu Qiao from Harbin Institute of Technology, working with senior engineer Song Chen from Fujian Special Equipment Inspection and Research Institute, developed a complete system for assessing steel corrosion in reinforced concrete. The system combines an optimized four-electrode surface sensor, an electrical model for interpreting impedance measurements, and a portable measurement instrument. It was designed to estimate conditions at the steel–concrete interface directly from the concrete surface, without drilling into the concrete or making an electrical connection to the reinforcing steel.

The first issue was electrode geometry. In a conventional four-electrode arrangement, the potential-measuring electrodes lie between the two electrodes that inject current. The researchers examined how changing both the separation of the current electrodes and the position of the potential electrodes affected sensitivity to corrosion. The underlying behavior comes from two competing current paths. Some current travels through the concrete between the surface electrodes without reaching the steel, and some passes through the concrete cover, enters the rebar and later returns to the surface. At high frequency, the steel–concrete interface has relatively low impedance because its response is dominated by capacitive behavior. At low frequency, charge-transfer resistance becomes much more influential and changes strongly with corrosion state. The difference between high- and low-frequency impedance therefore carries information about the condition of the steel.

Increasing the spacing of the current electrodes caused a larger fraction of the current to interact with the reinforcement and improved corrosion sensitivity. The researchers selected a 22 cm separation after finding that, under the geometric configuration they studied, further increases offered diminishing benefit. They also found that placing the two potential electrodes outside the current electrodes gave better sensitivity than the conventional arrangement in which they sit inside them. A spacing of 5 cm outside the current electrodes was selected for subsequent measurements. High-frequency impedance was also found to depend almost linearly on concrete resistivity while being relatively insensitive to charge-transfer resistance, providing a way to separate the influence of the concrete from that of corrosion.

Interpreting the complete impedance spectrum required a second step and a full finite-element model can reproduce current flow through the concrete and rebar, but repeatedly running such simulations during parameter fitting would be computationally inefficient. The authors instead used finite-element calculations to capture the geometric relationships of the structure once, then incorporated those relationships into an equivalent circuit model. Their model represents the rebar as a series of segments connected through the concrete and includes the impedance of the steel–concrete interface. Once the geometry-dependent quantities have been established, different corrosion parameters can be tested without performing another complete finite-element calculation for every trial.

The team performed comparison with detailed finite-element simulations showed impedance differences below 3 Ω over the investigated conditions. When the model was used in reverse to recover known interface parameters from simulated spectra, charge-transfer resistance could also be identified accurately. Across the uniform-corrosion test cases, the maximum reported error in recovered charge-transfer resistance was about 6%, occurring when low concrete resistivity reduced the sensitivity of the measurement.

The team then built a dedicated instrument to measure the indirect impedance response over a wide frequency range. They performed tests using circuits with known impedance and showed that the instrument closely matched the expected values, with most measurements having less than 5% error. Finally, the complete system was tested on reinforced concrete specimens with different corrosion conditions. In the specimen without added chloride, the embedded sensors showed that the steel was still passive. The indirect method reached the same conclusion, and it also classified the steel as passive. For the highly chloride-contaminated specimen, the indirect method identified a high corrosion rate and produced charge-transfer resistance values close to those obtained from embedded sensors. A deliberately non-uniform specimen provided a more demanding test. Surface potential measurements first identified the approximate location of the anodic region. Fitting the indirect impedance spectra then distinguished this corroding region from surrounding passive areas. The method estimated a corrosion current density above 4.29 μA/cm² in the anodic region, while locations farther away were classified as passive, broadly matching the embedded sensors.

The new work shows that corrosion information can be obtained from measurements made entirely at the concrete surface. By combining sensor design with electrical modelling, the system can separate the influence of the concrete from changes associated with the steel–concrete interface. The geometry-based equivalent circuit also makes the analysis more practical. Finite-element calculations are used to describe the effect of the structure and electrode arrangement, but they do not need to be repeated every time the impedance data are fitted. This reduces the computational effort needed to estimate corrosion-related parameters.