Pile driving in difficult Solar terrain: what to spec

Pile driving is the single biggest cost overrun risk on a utility-scale Solar project. We've seen 15–25% schedule slippage on jobs where the ground turned out to be more difficult than the desktop study assumed. Here's what we've learned about specifying the work properly in the first place.

Refusal depth and what to do about it

Refusal — the point at which a pile stops advancing despite continued energy input — is the contractor's nightmare. The standard mitigations are well-known but consistently under-specified in tenders:

• Pre-drilling rig on standby: not "available", actually on site from day one. Difficult terrain (rocky inclusions, frozen ground, hardpan layers) will demand pre-drilling on 5–15% of piles. Without the rig already mobilised you lose days waiting.
• Helical pile fallback: specify the brand and supply-chain contingency upfront. Late substitution of pile type triggers re-engineering of the tracker rows and is uniformly painful.
• Pre-cast concrete foundation as a last resort on sub-1% of pile positions, with the cost agreed in the tender. Don't negotiate this mid-construction.

Ramming energy

The default ramming energy specified in many RFPs is too low — bidders size for typical European soils (sandy loam, clay) and then face penalty days when the actual ground demands higher impact energy. We prefer to see:

• Minimum ramming energy: 35 kNm for sites where the geotech shows any consolidated horizon below 1.5 m.
• Variable-energy hydraulic hammers preferred over fixed-energy pneumatic — the marginal cost is small, the schedule flexibility is large.

The pre-driving survey that pays for itself

Spec test piles at 0.5% of total pile count, minimum 8 piles, distributed across the site rather than clustered. Pull-out and load tests on the test piles before the production driving starts. This catches three things in advance:

1. Variable refusal depths across the site, so the EPC sizes pile stock correctly.
2. Pile rotation under load — early indicator of pile-to-soil bond issues that will affect tracker performance over 25 years.
3. Aggressive groundwater triggering early coating-spec changes rather than mid-project rework.

The test-pile campaign costs a few thousand euros per MWp. We have yet to see a site where it didn't return the investment many times over.

Frozen ground

The forgotten edge case. Sites in northern Poland, the Baltics, and Romania's mountain regions can have frozen subsurface layers in November through March that defeat standard ramming. If your construction schedule overlaps a winter season:

• Spec a thermal allowance in the soil-class definition — contractors should be pricing for cold-weather methodology, not retrofitting it later.
• Consider steam injection or heated drill equipment as part of the tender's mobilisation kit. Cheap insurance.

Our view

Pile-driving issues are almost always survey-and-spec issues, not contractor issues. The EPCs we see consistently delivering on time are the ones who walk the site themselves before pricing, demand a proper geotech (not a marketing-grade desktop study) from the developer, and price the contingencies explicitly rather than burying them in margin. Developers who treat the pile-driving spec as a checkbox get exactly what they pay for.