Solar foundation testing, Sweden.

The challenge

The developer needed foundation parameters they could design and quote against. The desk study showed competent ground in some areas and soft material in others, but not where the boundaries ran and not how soft the soft zones really got.

Over-design adds steel everywhere. Under-design fails on the worst rows. The way to draw the line is to install test piles and measure what they do.

What we found in the field

Conditions varied widely across the site. In one part of the area, surface layers were thin but bearing: a usable crust that softened rapidly with depth. In other zones, piles sank under their own weight, with water at the surface, organic material below, and very soft deep clay underneath that. Elsewhere on the same site we hit rock close to surface. Surface water, boulders, and weak bearing made the work demanding, but never unworkable.

The test programme

Three working days, 58 GPS-tagged measurements:

• 12 tension and lateral tests for tracker uplift and wind-load resistance.
• 6 compression tests for vertical capacity in the softest zones.
• 40 torque measurements to map installed-capacity behaviour pile by pile.

The result: three foundation zones

The data didn't support a single foundation type for the whole site. Forcing one would have meant either over-buying steel for soft areas or under-spec'ing the rock zones. The site was divided into three zones, each with its own foundation:

• Helical-pile zone (majority of the site). Standard helical-pile foundation, geometry sized to measured soil capacity row by row.
• Helical pile + diagonal bracing zone (deep soft clay). Helical piles still carry the load, with a diagonal-bracing detail engineered for the soft profile.
• Alternative ground-bearing zone (rocky areas). Where the rock was close to surface, a ground-bearing foundation was the right answer, not a helical pile.

The detailed design now uses measured data. Pile geometry, pile count, and foundation type all change between zones. On-site QA during installation will use the same measurement framework as the testing programme.