How to Specify an Earth Anchor That Lasts 25 Years: Soil Class, Galvanizing and the Real Cost of Under-Specifying

Specifying an earth anchor that survives 25 years comes down to four decisions made before the first unit is driven: matching the anchor type to the soil class, sizing the coating to the actual corrosivity category at the site, keeping the working load below the soil's real capacity, and checking the details that fail first - the rod-to-hub connection, the galvanized thread, and the backfill at the entry point. Most premature anchor failures are not material failures. They are specification failures, and they are cheaper to fix on paper than on site.

Why anchors fail while the structure above them is still fine

An earth anchor is the cheapest component in almost any structure it holds, and the hardest one to inspect afterwards. It sits below grade from the day it is installed. If it fails at year six, the repair cost is rarely the price of a new anchor - it is the cost of dismantling the structure around it: releasing tension on a trellis row, excavating a footing, or taking a pole out of service.

That asymmetry is the whole argument for specifying carefully:

What fails	What it costs
Anchor body (corrosion)	Full replacement plus structure rework
Rod / thread (galvanic attack)	Re-tensioning, possible anchor loss
Soil capacity (under-designed)	Progressive movement, structure distortion
Field details (backfill, alignment)	Reduced capacity, water ingress, slow corrosion

A specification that only names a model number and a load rating does not address any of these four. Below is a specification process that does.

Step 1 - Classify the soil class before choosing the anchor

Different anchor families mobilise soil in fundamentally different ways, and the same nominal capacity means different things in sand and in clay.

Percussion-driven anchors - including our CM® and percussion type style units - are driven on a rod, then tipped or rotated into undisturbed soil and loaded by pulling against the deadman. Capacity comes from a plate bearing against soil that was never excavated, which is why they hold well in stiff cohesive soils and can be verified by a tension test immediately after tipping. They are a poor choice in loose, saturated fill, where the plate simply drags.

Helical / ground screw anchors - power driver screw anchors and helical ground screws - advance on a helix and gain capacity in granular soils by bearing on the plate plus friction along the shaft. They are torque-verified, which makes every unit traceable, and they can be extended to reach competent strata below weak surface layers.

A practical field proxy is a probe or penetrometer reading. Our own soil class table for driven anchors runs from Class 1 (solid bedrock, probe value above 600 in-lb) through Class 4 (gravel and compact clay mixtures, 300–400) down to Class 8 (marsh and saturated silt, under 100). Specifying a single anchor model across that whole range is choosing to be either wasteful in good ground or unsafe in poor ground.

The rule worth writing into your specification: never accept a single anchor type for a whole site until the soil class has been recorded at the actual anchor locations. A trellis row crossing a former waterway is not one soil.

Step 2 - Size the coating to the corrosivity category, not to a habit

This is the single most under-specified item in earth anchor procurement, and the easiest to cost-justify.

Many buyers still ask for "Z450" or "275 g/m²" as a blanket galvanizing requirement. Both are routinely misapplied. AS 1397 Z-coatings describe continuously galvanized sheet and strip, not fabricated anchors. For fabricated, hot-dip galvanized articles the governing specification across Australia and New Zealand is AS/NZS 4680, and it sets coating thickness by article thickness, not by a single number.

Minimum average coating, articles not centrifuged, to AS/NZS 4680:

Article thickness	Min. local	Min. average	Coating mass
≤ 1.5 mm	35 µm	45 µm	320 g/m²
> 1.5 – 3.0 mm	45 µm	55 µm	390 g/m²
> 3.0 – 6.0 mm	55 µm	70 µm	500 g/m²
> 6.0 mm	70 µm	85 µm	600 g/m²
> 6.0 mm (HDG900)	85 µm	125 µm	900 g/m²

 

Now match that to the exposure. Using widely published ISO 9223 corrosivity categories and typical zinc corrosion rates:

Category	Typical environment	Zinc loss	Years to first maintenance at 70 µm
C2	Arid / inland, >50 km from sea	0.1–0.7 µm/yr	100+
C3	Low-salinity coastal, urban, light industrial	0.7–2.1 µm/yr	33–100
C4	Calm sea-shore, 0.2–1 km inland	2.1–4.2 µm/yr	17–33
C5	Surf coast, beachfront, aggressive industrial	4.2–8.4 µm/yr	8–17
CX	High salt deposition, severe acid industrial	8.4–25 µm/yr	3–8

 

The conclusion is uncomfortable for anyone buying on price: a 55 µm earth anchor that gives 50+ years inland can be due for first maintenance in under a decade on a surf-beaten coastal block. If your project sits in C4 or worse, specifying the heavy coating tier is not a premium - it is the only way the anchor outlives the structure. Ask for coating thickness test results, not coating weight claims.

Two related details get missed in the same conversation: cut ends and welds at the helix must be zinc-rich repaired, and threads should follow AS/NZS 1214. Corrosion failures concentrate at threads and at the rod entry point, not in the middle of the shaft.

Step 3 - Separate the structural rating from the ground's rating

A product page that says "95KN" tells you what the hardware survived in a test rig. It tells you nothing about what your soil will give. Ultimate anchor capacity is the lower of the anchor's structural capacity and the soil's pull-out resistance, and in most non-bedrock sites the soil is the limiting term.

For helical units the industry handles this with the torque correlation method, Q_ult = Kt × T, where T is the installation torque averaged over the final few plate diameters of embedment. Published Kt ranges from ADSC and ICC-ES AC358 guidance run roughly 3–5 ft⁻¹ for soft clay, 5–8 for stiff clay, 6–10 for loose sand, 8–12 for dense sand and 10–15 for gravel. These are preliminary screening values only - a thin dense lens inside soft clay can produce a torque spike that overpredicts real capacity if it is folded into the average, and a saturated sand can read low and then drain.

For driven plate anchors there is no such formula; verify by pulling. A tipped percussion anchor can be proof-loaded immediately, before the structure is built onto it. Build that test into the programme for any anchor whose failure would be expensive, and record it.

Step 4 - Price the cost of doing it twice

The honest comparison is not galvanizing tier against galvanizing tier. It is installed cost against replacement cost.

Take a trellis or a line of poles in a C4 coastal block. Upgrading from a 55 µm to an 85–125 µm coating adds a modest amount per unit. Replacing those anchors at year ten - releasing tension, extracting, re-driving, re-tensioning, losing a season of growth or a span of line - costs multiples of the entire original anchor purchase. This is why the UNI EN ISO 12944 / ISO 9223 durability tiers exist: to make "design life" a specification you can write down, not a hope.

Specification checklist

Hand this list to procurement and it becomes hard to buy the wrong anchor:

Anchor type matched to recorded soil class at each anchor location - not one model across the site

Coating specified to AS/NZS 4680 by article thickness, not as a single Z or g/m² figure

Coating tier matched to the site's ISO 9223 corrosivity category, with test certificates

Cut ends and welds zinc-repaired; threads to AS/NZS 1214

Verified torque or proof-load record per anchor, retained as QA documentation

Rod-to-hub connection inspected for weld continuity and for crevices that trap water

Entry point detailed so backfill is compacted - this is where water gets in first

Working load set at a defensible fraction of soil capacity, not of the structural rating