PV Pile Foundations in Corrosive Soil: Assessment and Design

Driven steel piles are the default foundation for utility-scale PV mounting structures, and in most soils a hot-dip galvanized section outlasts the plant. In highly corrosive soil that assumption fails quietly: the zinc is used up years before the design life ends, and the steel loss concentrates where nobody inspects it, just below the ground line and at boundaries between soil layers. The bottom line is that soil corrosivity has to be measured, classified and turned into a durability calculation before the pile section is frozen. Depending on the result, the answer is a thicker pile with a heavier coating, a protected ground-line zone, or a different foundation type.

Why corrosive soil is a design-stage risk for PV piles

For driven PV piles, wind-induced uplift usually governs the design ^[13], and the pile must also carry bending from lateral wind load. Steel lost to corrosion reduces the tension and bending resistance that the structural check relied on, so corrosion is a capacity question, not only a maintenance question. There is no dedicated standard for corrosion design of solar piles; the industry’s primary approach is galvanizing plus a calculated sacrificial steel thickness, and each owner decides how conservative to be ^[9].

Corrosion on a PV pile is rarely uniform, and the locations that matter are predictable:

• Layer boundaries. A driven pile typically crosses several soil units. Aerated sandy or gravelly layers act as cathodes and wetter clays or silts as anodes, setting up macro-cells that localize attack even in moderate soils ^[1].
• Disturbed ground and fills. Disturbed soil holds more oxygen and is more corrosive than undisturbed ground ^[4]. The EN 1993-5 loss allowances for steel piles reflect this: non-compacted fills lose more steel than undisturbed natural soils, and aggressive fills such as ash or slag lose the most ^[4].
• The ground line. The soil-air interface carries a risk of accelerated corrosion, with higher microbial activity in the first 200-300 mm below natural ground level ^[6].
• Electrical coupling. Galvanized steel connected to an extensive copper earthing system suffers high corrosion effects, because the copper acts as a large cathode ^[11].
• Poor drainage. In a two-year field exposure across several unsaturated soils, the most damaging site was not the most saline one but a poorly drained one, where heavy rain drove complete penetration of both carbon steel and zinc samples ^[2].

Hungarian sites are not exempt. Salt-affected soils cover about 5,340 km², or 5.8% of the country, with high alkalinity and sodicity at shallow depth, saline topsoil, clay-rich textures and gleyic conditions deeper down ^[12]. Soluble salts, clay and waterlogging score badly in any corrosivity assessment, so my recommendation for Hungarian EPC projects is to include corrosion testing as standard scope rather than wait for a warning in the geotechnical report.

How to assess soil corrosivity for PV pile foundations

The parameters that matter

Soil corrosivity is the combined effect of moisture and aeration, electrical resistivity, pH and buffer capacity, dissolved ions (chlorides and sulfates in particular), microbial activity such as sulfate-reducing bacteria, hydro-climatic variability and geological heterogeneity ^[1]. Practical test lists from engineering firms and the US Department of Energy cover the same ground ^{[9, 10]}.

Some indicators are visible in the field. Gray soil signals poor aeration and is generally more corrosive to zinc than red, yellow or brown soil. Very acidic soils are associated with extreme corrosion rates and deep pitting, and one study found the corrosion rate doubled between 4 °C and 20 °C ^[5].

DIN 50929-3 classes

The most complete method in European practice is DIN 50929-3, now in its 2024-05 edition. It estimates the likelihood of external corrosion of buried pipelines and structural components and is used alongside the European standards EN 12501-1 and EN 12501-2 on corrosion likelihood in soil ^[8]. Each measured parameter receives a weighted rating; the 2024 edition uses 15 parameters ^[1]. Summing the soil ratings gives B0 for a homogeneous soil. Adding penalties for horizontal and vertical heterogeneity gives B1, which captures the macro-cell risk described above ^[4]. The sum maps to a class:

The zinc rates are a compilation used by a tracker manufacturer in an EU research project, not values from the standard itself ^[4]. They are useful for a first estimate, not as design values.

A faster screen is the AASHTO “mildly corrosive” check used in the PV durability framework of Jares Salguero et al.: resistivity above 3,000 Ω·cm, pH between 5 and 10, organic matter of 1% or less, chlorides below 100 ppm and sulfates below 200 ppm. Soils that fail it need the full assessment ^[1].

What “active” and “passive” aggressiveness should mean

Site teams sometimes describe a soil as actively or passively aggressive. Neither term is defined in DIN 50929-3 or EN 12501, so it is worth pinning down. The useful reading follows the DIN logic. The intrinsic aggressiveness of the soil (its chemistry, resistivity and moisture, expressed in B0) is the passive load that is always present. The conditions that switch localized attack on (heterogeneity expressed in B1, disturbance, wet-dry cycling near the surface, poor drainage and electrical coupling to other metals) are the active drivers. A soil with a moderate B0 can still destroy piles if the active drivers are strong, so the design has to address both.

The limits of any classification

Classification is a probability estimate, not a prediction ^[8]. In the field exposure cited above, the poorly drained site defeated the DIN estimate, and the authors propose adding soil drainage capability to the method ^[2]. Measured annual rates across those sites ranged from 20 to 120 µm for carbon steel and 3 to 8 µm for zinc, with steel corroding up to 15 times faster than zinc at one site ^[2]. The DOE advice is sound: a corrosion study is different from a geotechnical study and should involve an engineer experienced in corrosion ^[10].

From corrosivity class to pile specification

The zinc budget

The life of a galvanized coating in soil is simply its thickness divided by the zinc loss rate ^[1]. EN ISO 1461 sets minimum coating thicknesses by steel thickness: for steel over 6 mm, 85 µm mean and 70 µm local; for steel over 3 mm up to 6 mm, 70 µm mean and 55 µm local ^[7]. Heavier coatings are possible, for example by grit blasting before galvanizing, and should be agreed with the galvanizer in advance ^[7].

Applying published rates to an 85 µm coating shows how wide the uncertainty is:

• Class Ib at 1.25 µm/yr: about 68 years. Class II at 2.4-3.2 µm/yr: about 27-35 years. Class III at 5.0 µm/yr: about 17 years ^[4].
• A conservative AASHTO-based path, with up to 12 µm/yr in the first two years and about 4 µm/yr after, gives roughly 17 years even in mildly corrosive soil ^[1].
• The 3-8 µm/yr zinc rates measured in the field exposure give roughly 11-28 years ^[2].

The practical conclusion is uncomfortable but clear. Unless the soil is clearly class Ia or Ib, the minimum EN ISO 1461 coating alone is unlikely to carry a pile through a 30-35 year design life on a conservative basis.

Sacrificial steel

When the coating life falls short, the remaining years are covered by extra steel, calculated from bare-steel corrosion rates for the relevant class ^[1]. Using the indicative rates in the table, a class III soil with 17 years of zinc and a 30-year design life leaves a 13-year deficit, or about 0.8 mm of steel from each exposed face at 60 µm/yr. The same calculation in class II gives well under 0.1 mm. The class, not the design life, dominates the answer.

EN 1993-5 gives a cross-check for bare steel piles: 0.30 mm of thickness loss at 25 years in undisturbed natural soils, 1.00 mm in aggressive natural soils such as swamp or peat, and 2.00 mm in non-compacted aggressive fills. Values for compacted fills are halved, and only the 5- and 25-year values are based on measurements; longer periods are extrapolated ^[4].

My recommendation is to show this calculation explicitly: coating thickness, zinc rate and its source, years of coating life, deficit, steel rate, allowance per face, and the reduced section used in the end-of-life structural check. It costs little and gives the technical advisor a traceable basis ^[1].

Heavier and alloy coatings

Zinc-magnesium-aluminum (Zn-Mg-Al) coatings are often proposed for corrosive sites. Available studies suggest a durability factor of 1.5 to 2.5 relative to galvanized coatings in soil ^[1]. A 2025 study of a Zn-Mg-Al coating in highly aggressive, heterogeneous soil found low corrosion rates but warned that the advantage over galvanizing is not universal in such soils, and observed preferential attack of the magnesium- and aluminum-rich phases ^[3]. My position is to accept an alloy-coating credit only when soil-specific test data back it, and to compare products on the coating mass actually applied per side, not on the alloy name.

Protecting the ground line

The ground line combines oxygen, moisture cycling, microbial activity, debris and, on many PV sites, vegetation control and grazing. The Galvanizers Association of Australia lists herbicides, fertilizers and animal urine among the aggravating factors and recommends considering additional protection for the first 250 mm below and above the natural surface on partially buried steel ^[6].

My position is to protect a band around the ground line of roughly 30 cm on each side with a barrier coating over the galvanizing, creating a local duplex system. Suitable options include epoxy mastic, high-build epoxy, powder coating, elastomeric polyurethane or polyurea, corrosion-inhibiting tapes and, particularly for piling, high-density polyethylene; bituminous paint is often unsuitable where it is exposed to UV ^[6]. The system should seal against moisture wicking between coating and zinc, and the site should be graded so water does not pond at the pile ^[6].

Three details decide whether the band works in practice:

1. It must straddle the ground line. Paint applied only above grade misses the zone just below grade where microbial activity is highest ^[6]. Position the band relative to the design embedment, with allowance for driving-depth tolerance and final grading levels.
2. It must survive installation. Coatings can be damaged as piles are driven ^[9]. Specify an abrasion-resistant system, inspect after driving and define a touch-up procedure in the QA plan.
3. It is not a substitute for the durability calculation. A ground-line band does nothing for macro-cells at depth or for full-depth attack in class III soil.

When to change the foundation type

For most class Ib and II soils, a thicker section with a heavier coating and a protected ground line is the economical answer. The case for a different foundation grows with corrosivity, design life and geotechnical difficulty.

Micropiles are small-diameter grouted piles, typically under about 300 mm, with internal steel reinforcement, and can resist compression, tension and lateral load ^[14]. In my view they are justified mainly by geotechnical reasons such as rock, obstructions or refusal, not by corrosion. Caltrans goes further and does not recommend micropiles in corrosive environments without specialist mitigation ^[14]. In aggressive soil the reinforcement still needs its own corrosion design: grout cover, coated or encapsulated bars, or sacrificial thickness.

Bored, concrete-cast foundations set a steel post in a cast-in-place concrete pier, so the buried steel is protected by concrete rather than zinc. The concrete must be specified for the soil’s chemical exposure class under EN 206, from XA1 (slightly aggressive) to XA3 (highly aggressive) ^[15]. The point where the steel leaves the concrete becomes the new ground line and needs the same barrier detailing.

Screw piles are still buried steel, so the same zinc-budget and allowance calculation applies to both the shaft and the helix.

Ballasted concrete foundations avoid buried steel entirely. They suit sites where penetration is not allowed or the soil is too aggressive for any reasonable steel allowance, at the cost of more concrete, heavier logistics and tighter grading.

Sacrificial anodes are rare in solar installations ^[9]. Whatever the option, compare costs over the full design life: alternatives add drilling, concrete, curing time and slower installation, but in my judgment piles that need strengthening or replacement mid-life are the most expensive outcome of all.

Practical tools: decision table and survey checklist

The matrix below is my starting point for selecting foundation and protection options. It assumes the inputs come from site-specific testing, and it does not replace the durability calculation.

Survey and testing checklist

1. Desk study. Geology, mapped salt-affected soils, former land use (fills, industrial ground), groundwater levels and drainage.
2. Geotechnical investigation. Space boreholes or CPTs by geological zone, since capacity uncertainty grows between test points and where geology varies ^[13]. Run uplift (pull-out) and lateral tests on representative piles in each zone.
3. Corrosion sampling by layer. Sample each soil unit the pile crosses, not one composite per borehole. Measure in-situ resistivity with the four-pin Wenner method, mainly between 0.5 and 3.0 m depth, and soil-box resistivity across the expected moisture range ^[1].
4. Laboratory analysis. Soil type, water content, pH, buffer capacity, chlorides, sulfates, sulfides, neutral salts, redox potential and organic content, prepared to a protocol such as DVGW GW 9 ^[1]. Add microbial screening where redox results indicate anaerobic conditions.
5. Classification. Calculate B0 and B1, assign the DIN 50929-3 class and run the AASHTO screen ^{[1, 4]}.
6. Electrical checks. Avoid copper earthing coupled to galvanized piles; high-alloy stainless steel (316Ti or 316L) is the corrosion-resistant option for buried earthing ^[11]. Measure structure-to-soil potential before and after commissioning ^[1].
7. Durability report. Zinc budget, steel allowance, ground-line detail and reduced-section structural check, with traceable inputs.
8. Construction QA. Verify coating thickness on delivery against EN ISO 1461 ^[7], inspect for driving damage, avoid over-excavation and compact backfill in layers with low-corrosivity material ^[1].
9. Long-term verification. Install spare, non-structural test piles and extract them at intervals to measure actual zinc and steel loss ^[9].

Conclusion

Corrosive soil is a solvable problem when it is treated as a design input. Measure the soil layer by layer, classify it, and convert the class into a zinc budget and a steel allowance that the structural check actually uses. Thicker steel with a heavier coating and a protected ground line will cover most sites. Where the calculation shows that no reasonable allowance survives the design life, the decision to switch to micropiles or concrete is made on evidence rather than habit.