Soil Corrosivity and Metal Loss in Earth Retaining and Foundation Systems
Last updated August 31, 2026
MSE wall reinforcement and driven piles are usually protected from corrosion by extra steel alone, sized to corrode away over the service life. How much extra steel to provide is the subject here. Built on the National Academies' 2023 study Corrosion of Buried Steel at New and In-Service Infrastructure, FHWA and Caltrans guidance, and two Wisconsin forensic studies, the course explains how soil corrosivity is measured and where the standard tests mislead, how agency acceptance criteria differ, and how the AASHTO and Darbin metal loss models size sacrificial thickness along with the conditions those models depend on. It covers the mechanisms uniform rate models miss, including pitting, macrocells, stray current, and microbial attack, and the protection options when added steel is not enough. The final sections turn to structures already in service: monitoring hardware, electrochemical condition assessment, a forensic case study of an MSE wall destroyed by de-icing salt, and a risk scoring framework for prioritizing inspection across a wall inventory.
Learning Objectives
Plan a soil corrosivity investigation that matches test methods to site conditions and reports results with their limits stated.
Apply the AASHTO and Darbin metal loss models to size sacrificial steel within each model's validity conditions.
Identify the corrosion mechanisms that uniform rate models cannot represent.
Prioritize in-service walls for inspection using monitoring data, salt exposure indicators, and risk scoring.
BasePDH | Course 024 | 3 Professional Development Hours
Soil Corrosivity and Metal Loss in Earth Retaining and Foundation Systems
1. The Corrosion Allowance Commitment
Every design that puts steel in the ground makes a commitment about corrosion, whether or not the drawings say so. The National Academies' 2023 consensus study Corrosion of Buried Steel at New and In-Service Infrastructure describes two philosophies that divide buried-steel practice, and the division is in what gets characterized before construction and what gets monitored after [1].
Under the corrosion allowance approach, the designer characterizes the corrosivity of the earth materials, laterally and vertically, in order to predict a corrosion rate, then adds cross-sectional steel thickness to compensate for the expected metal loss over the designated performance period, sometimes alongside a protective coating such as galvanizing [1]. Once the steel is buried, little is monitored. Safety rests on structural redundancy: failure of one component does not cause failure of the system. This is the practice of what the report calls the geo-civil industries, covering structural foundations, earth retaining structures, dams, and tunnels [1].
Under the corrosion avoidance approach, the designer prevents or reduces the corrosion rate with coatings plus cathodic protection, and then monitors the effectiveness of that protection continuously for the life of the asset [1]. Site characterization is minimal, because a transmission pipeline runs hundreds of kilometres and characterizing the whole alignment is impractical. This is oil and gas pipeline practice. Water utilities follow a third path that is really an economic judgment: they generally focus little on external corrosion, tolerate leaks, address breakages as they happen, and spend their attention on internal corrosion and drinking water compliance, because controlling the environment is not economical given the spatial distribution of a pipe network [1].
The report is careful to say that not every industry or application fits cleanly into one approach, and that hybrids exist [1]. What it states without qualification is that the same corrosion mechanisms act on the steel regardless of which approach was chosen [1]. The philosophy changes what the designer knows about the site, not what the site does to the steel. Choosing allowance therefore commits the design to two things at once: a subsurface characterization good enough to support a rate prediction, and an acceptance that nothing will look at the steel again. Most of this course is about what each half of that commitment actually requires.
Geometry sets how much is at stake. The amount of steel exposed to the subsurface follows from the surface area in contact with it: circular bars carry a low surface-area-to-volume ratio, thin sheet piles a high one [1]. For pipelines, the contact area is dominated by installation method and quality rather than by section geometry. Trenched pipe contacts backfill according to the uniformity of compaction around it and the presence of air gaps larger than the soil pore spaces, while a bore placed by horizontal directional drilling may sit in non-native drilling mud [1].
Geo-civil practice tolerates a slow corrosion rate because most of its structures have a design life of less than 100 years [1]. The common approach is physical barriers and controlled environments combined with increased cross-sectional thickness; cathodic protection appears occasionally, on some bridges, but it is not the default [1]. The record justifies the tolerance and simultaneously bounds it. Out of tens of thousands of mechanically stabilized earth walls in service, approximately a dozen have been documented as failures caused by corrosion of the reinforcement, and all of them were correlated with corrosive environments (Gladstone et al., 2006, as reported in [1]). The report treats that record as the product of a deliberately conservative upper-bound design method, and it states both horns plainly: conservative design results in large economic cost, and it still does not preclude failure in every case [1].
1.1 The words are not shared
The committee that wrote the seed report was drawn from geotechnical engineering, structural engineering, materials science, corrosion engineering, microbiology, and several other fields, and it found that its own members sometimes spoke using the same words with different meanings [1]. Until the committee agreed on terminology, its discussions were confusing and its draft text was ambiguous and even contradictory [1]. That experience is worth taking seriously, because the same collisions occur on any multidisciplinary project team, and they produce real errors rather than pedantic ones.
Four documented collisions matter to practice [1]:
- "Corrosion potential" is an electrochemical term with a specific measurable meaning, the potential at which anodic and cathodic reaction rates are equal. Geotechnical professionals commonly use the same phrase to mean the likelihood that corrosion will occur. A geotechnical report that says a fill has "low corrosion potential" and a corrosion engineer's survey reporting measured corrosion potentials, typically -500 to -750 mV for unprotected steel in soil against a copper-copper sulfate electrode [1], are using unrelated quantities under one name.
- "Pitting" and its scale. Corrosion engineers and metallurgists on the committee defined pitting corrosion and its scales differently than the geotechnical and civil engineers did [1]. Two specialists can agree that a surface is pitted and disagree by orders of magnitude about what that implies.
- "Well graded" inverts across disciplines. A well-graded soil in engineering usage is "poorly sorted" in the geosciences, and a poorly graded soil is "well sorted" [1].
- "Soil" itself was a source of confusion early in the study, as was the recognition that soil is a multiphase electrolyte of solids, liquids, and gases rather than a simple solid [1].
The report also names the competence gap that makes the vocabulary problem consequential: engineers who design and conduct site characterization investigations are rarely knowledgeable about corrosion mechanisms, corrosion engineers are often unfamiliar with the complexities of the soil, groundwater, and gas electrolyte, and fewer still are familiar with subsurface microorganisms [1]. Its conclusion is that corrosion protection will be more effective with improved understanding than with routine application of higher factors of safety [1]. In the meantime, practitioners compensate for high uncertainty with conservative practice, which regularly results in safe infrastructure at a higher economic cost than necessary [1].