Why PV Structures Collapse: Five Contributing Factors

It’s the destiny of structures to eventually decay or collapse. Take the Seven Wonders of the Ancient World. Once the most enduring structures known to mankind, all but one collapsed within hundreds of years and the Great Pyramid at Giza is a shadow of its former grandeur. The PV sector isn’t aiming to compete with the Wonders of the World but it’s true that a structure’s useful life depends on both design and build quality.

Attention to design and safety margins makes the difference between a one-year life or a 100-year life, and in our experience occasionally PV structures do collapse after one year. How confidently can we claim that a PV structure will survive its design life?

In other sectors, structural reliability levels are hard-coded in to standards through prescribed loads and safety factors. However, PV structures don’t have dedicated design standards and the practical consequence is that confidence and reliability vary significantly from project to project.

So why do PV structures collapse?

Simon Hughes who is a Principal Engineer within DNV GL – Energy’s UK Project Engineering team has answered this question with five aspects which can lead to problems.

1. Site wind conditions

Site conditions are covered by standards but errors can be made in applying them, especially when it comes to regional differences. We frequently find differences in approaches to terrain and topography effects, return period for extreme winds, wind directionality, and choice of safety factors. Even a 1-in-50-year wind has an unacceptably high chance of occurring when compounded over a normal project life (about 4 in 10), so as well as understanding the site conditions, it’s critical that proper factors of safety are applied.

2. PV structure wind loading

Tilted modules exhibit wing-like behaviour: wind causes lift from one direction and downforce from the other. This much is agreed but, go deeper and the range of understanding and approaches between designers is surprising. Wind pressure depends on the tilt angle and comprises a mean component and an unsteady component which constantly fluctuates over the PV surface. ‘Pressure coefficients’ convert wind air pressure to loads on structures and can be obtained from design codes or from wind tunnel testing but often the unsteady component is overlooked, or poorly scoped testing gives overly-optimistic results.

Then there are ‘aeroelastic effects’ – with phenomena like vortex shedding, galloping, wake buffeting, flutter – which are not particular to solar PV but do require specialist knowledge and are routinely ignored. It’s important to know that they can lead to spectacular failures under only moderately high winds and imperative to know which designs warrant further study. Add other poorly-codified aspects like directional effects, edge effects, wake effects and sheltering, and perhaps the varying approaches and quality of design aren’t surprising.

3. Forgotten lessons from structural engineering

Sometimes well-known structural failure modes, such as prying (a kind of levering open of connections in tension) or various types of buckling (a sudden sideways failure under compression), are ignored for convenience of calculation. These should be understood and are well codified, so there are no excuses for learning the same lessons again when it comes to PV projects.

4. Ground conditions

There is a saying that “you pay for a ground investigation, whether you have one or not” and with budgets in the region of 0.1% CAPEX, limiting the scope of ground investigations could be a false economy considering the premium for transferring unknown ground risk or paying for un-optimised foundations. Obviously rather than the price, it’s the scope and quality of investigations which are key, so it pays to be an informed buyer. Next, a recognised geotechnical design approach with codified safety factors should be used to assess pull-out strength, lateral strength, downforce and settlement. Finally, soil chemistry should be considered against corrosion protection specifications. Unfortunately, geotechnical design is another area where appropriate checks are sometimes overlooked or where valuable ground investigation results are ignored.

5. Build quality

Solar PV has an admirable reputation for rapid deployment; per-megawatt build times are impressive, especially compared to conventional generation. But with large sites, price pressure and cliff-edge deadlines, quality can sometimes suffer. Overall, build quality is reasonably good but it’s not uncommon to encounter issues like uncontrolled modifications to foundations, systematic under-torqueing of bolts, and occasionally even incorrect material types.

For more information about DNV GL