Ground-mounted solar power plants: Why static loads aren’t enough

In a paper presented at the 2015 convention of the Structural Engineers Association of California (SEAOC), Dr. David Banks of CPP and Joseph Cain, P.E., of SunEdison explain why it’s not good enough to design most ground-mounted solar installations for static wind loads alone. Dynamic loads, it turns out, are just as important, and in some cases, even more so.

Structural engineers responsible for calculating design wind loads for large-scale solar installations have long turned to the monoslope free roof calculations in ASCE 7-10 and its predecessors for coefficients that estimate the uplift forces and overturning moments on racks of solar modules. These calculations offer a reasonable place to start, though they don’t necessarily produce conservative results.

Wind tunnel studies of numerous ground-mounted solar arrays have repeatedly demonstrated that such arrays suffer from significant edge effects, the most common of which is vortex shedding. As wind encounters the first row of panels before impinging upon the array, it sets up a flow pattern that creates alternatingly high and low pressures, very similar to the wake behind a cylinder. The effect of such first-row vortex shedding is to generate dynamic loads that fluctuate at a specific frequency. That frequency is related to the array’s tilt angle, the size and shape of the racking systems, and the wind speed.

As it happens, the energy contained within these fluctuating vortices is often concentrated near the natural frequencies of the solar racking structures themselves. Much like science demonstrations that break wine glasses with an increasingly high-pitched (high-frequency) sound, wind and vortices that introduce energy to the solar array at just the right frequencies can create dynamic loads that quickly overwhelm the system, causing catastrophic failure.

The provisions in ASCE 7-10 and earlier editions draw a distinction between rigid and flexible structures, marking the boundary between the two at a natural frequency of 1 hertz (Hz). However, this 1 Hz limit assumes that the structures under consideration are the size and shape of buildings. When it comes to something as small and flexible as solar module racking structures, the dividing frequency should be more like 4–5 Hz in order to avoid amplifying the wind inputs to the structure.

As manufacturers of racking structures rely on static-load wind analysis to justify removing material from their systems, these effects only become more apparent: The lighter a structure becomes, the more susceptible it is to dynamic wind forces. The best approach, therefore, is to conduct a comprehensive wind-tunnel-based analysis that captures both static and dynamic effects. This helps designers create a system that strikes the right balance between material savings and structural integrity.

The paper, entitled”Wind Loads on Utility Scale Solar PV Power Plants” is part of the 2015 SEAOC Convention Proceedings and may be downloaded from CPP’s website at