From geotechnical report to foundation decision

A geotechnical report is not a foundation specification. It is evidence. The jump from borehole logs to a pile type, length, and installation method is a chain of engineering judgments. Skipping a link in that chain is how solar projects buy the wrong steel and discover refusal with a crew already mobilized.

This article walks the chain conceptually. It is not a design procedure.

Infrenta provides decision-support tools and does not replace project-specific engineering review or sealed professional design where required.

The investigation

A useful investigation for ground-mount solar typically includes borings or CPTs on a grid that reflects both area and suspected variability, plus laboratory tests that match the failure modes you actually care about: strength, density, corrosivity, frost-related properties, and rock characterization if refusal is plausible.

If the investigation was scoped for a building pad and then reused for a 2-mile array, the first decision is whether the data even applies. Software cannot repair an investigation that was never performed.

Soil profile

Capacity methods are profile methods. A single “allowable 50 kPa” note in a summary table is not a profile. Layer thickness, strength with depth, and whether the profile is sand, clay, fill, or mixed control shaft friction, end bearing, and whether a helix can advance.

When teams skip the profile and jump to a typical pile, they are betting that the site is the typical site. Utility-scale sites are often not.

Groundwater

Groundwater changes driving, drilling, corrosion, and sometimes buoyancy and construction sequencing. A water table that sits in the frost zone is a different problem than a water table at 12 metres. If the report is silent, the correct engineering posture is reduced confidence — not a guessed depth that makes the calculation look complete.

Frost

In cold regions, frost depth and adfreeze can govern uplift more loudly than wind in a spreadsheet that forgot them. Embedment that stops in frost-susceptible soil because “that’s what we used last year” is not a method. It is a habit.

Frost also affects construction calendars. A system that requires wet concrete in January is not equivalent to a steel pile that can be driven in frost, even if both “work” in a summer capacity table.

Bedrock

Bedrock is not a yes/no flag. Depth, quality, slope, and weathering decide whether driven piles refuse early, whether screws can be installed, whether helixes can reach competent material, and whether ballast even has a surface to sit on.

Unknown bedrock should cap confidence in any evaluation. Pretending the last boring represents the ridge 400 metres away is a common bid-phase error.

Axial capacity

Compression and tension are different limit states. Solar arrays see both: gravity and equipment loads in compression, wind and frost in uplift. A system sized only for compression will surprise you at the corners and on tracker rows with high chord loads.

Capacity is also not a single number. It is a method, a factor of safety or resistance factor, a soil interpretation, and a section. Changing the section without re-running the method is how “we upsized the pile” still fails a check.

Uplift

Uplift is where installation method and geometry matter as much as soil. Helical plates, screw geometry, driven friction, and ballast weight are different physical stories. Mixing their unit rates in an estimate without mixing their mechanics in the engineering is incoherent.

Lateral considerations

Wind on a tracker or fixed-tilt table is a lateral and moment problem at the foundation. A pile that is adequate in axial but short or slender in lateral will show it in deflection and in the steel, not in a vertical capacity table.

Lateral analysis needs loads from the structural model, not a generic 1.5 kN “wind” cell.

Constructability

A foundation that cannot be installed is not a foundation. Access, slope, cobbles, refusal, torque equipment, driving equipment, grout supply, and winter constraints belong in the same conversation as capacity.

Constructability also feeds cost: mobilization of a larger driver, pre-drilling, or a change to screws after refusal is not a small delta.

Corrosion

Aggressive soils, stray current, and coating damage during driving change remaining section over the project life. Galvanizing specification is an engineering and commercial input. It should be visible in both GeoLab context and steel cost, not added as a round number after the bid.

Installation equipment

The equipment spread is part of the design. Helical installation torque equipment, pile driving hammers, and screw drivers are not interchangeable. If the recommended system requires a machine that cannot reach the back of the site, the recommendation is incomplete.

Cost

Cost is a legitimate comparison axis: steel mass, coating, installation hours, testing, and risk of change. It is not a substitute for capacity. The cheapest pile that refuses is not cheap.

Decision-support software can help keep these axes on one page. It cannot choose for the engineer of record.

Alternatives, conceptually

Driven piles, helical piles, ground screws, and ballasted foundations occupy different regions of this chain. A structured comparison is useful; a ranking widget is not a professional opinion. See Helical piles vs driven piles vs ground screws and Foundation evaluation.

The report is the start of the decision. The decision is finished when a licensed professional has accepted the loads, the ground, the installation method, and the commercial consequences together.