Utility scale PV makes up the largest fraction of annually installed PV capacities in the U.S. and the first half of 2018 has seen “the largest half year of utility solar procurement ever” [SEIA, “U.S. Solar Market Insight 2018 Q3”]. As the investment for utility scale solar PV power plants is significant, its resilience should be ensured in great detail. Wind is one of the main hazards to ground-mount systems and attention should be paid regarding whether wind loads have been taken into account properly.
We will present boundary layer wind tunnel measurements of a generic array of PV trackers tested at various wind directions, tilt angles and row spacings. We will outline the procedure needed to correctly calculate the effective wind loads and show generic results with application to utility scale installations.
The static loads in question are the torque around the torsional axis along the row and the normal force acting on the panels. We have found that loads on the perimeter of the field are twice as high as in the interior, sometimes even higher. Moreover, static loads are highly dependent on tilt angle, with torque decreasing and normal force increasing with higher tilt angles. Interior rows experience shielding at higher tilt angles which result in lower loads. With increasing row spacing, on the other hand, rows behave much like stand-alone rows and experience higher loads. Dynamic effects are often neglected or not accounted for in the design. However, resonant dynamic loads caused by vortex shedding from upstream rows must never be neglected as they can lead to an increase of static loads by more than a factor of 2 depending on natural frequency and structural damping.
Aeroelastic instabilities such as torsional galloping are crucial, too. In the worst case, this instability can lead to total failure of the structure and 100% economic loss. We present video footage of catastrophic events, but also present a solution of how to counteract torsional galloping by means of numerical simulation. It will be shown that trackers deployed in multi-row arrays are more resistant to the instability than stand-alone trackers. Moreover, it will be highlighted that stowing parallel to the ground is never a good solution with small tilt angles of 10° already providing much better resilience to the instability. Increasing torsional stiffness at the array perimeter appears to be an effective strategy to protect the array interior.
Thorsten Kray– Head, Building Aerodynamics, I.F.I. Institut für Industrieaerodynamik GmbH (Institute for Industrial Aerodynamics LLC)