Wind Tunnel Testing
Safety is fundamental, but how do we define safety in the context of a commercial or utility-scale PV system?
We believe that ASCE 7 standard “Minimum Design Loads for Buildings and Other Structures,” which forms the heart of nearly all building codes in the U.S., provides the appropriate benchmark for safety.
This ASCE standard defines design loads for building structures, rooftops, rooftop equipment and building cladding in such a way that an “acceptable” level of life safety is attained when the design standard is followed. Virtually all of the design curves for wind pressure included in ASCE 7 are the result of wind tunnel test programs.
With that in mind, in 2006, SunLink approached renowned wind expert Dr. Greg Kopp at the Boundary Layer Wind Tunnel Laboratory (BLWTL) at the University of Western Ontario about developing an “ASCE 7-type” design curve for rooftop PV arrays.
Over the next five years, SunLink and the team at the BLWTL embarked upon what is, to our knowledge, the most extensive wind tunnel test and analysis program ever performed for any type of rooftop equipment. As Dr. Kopp noted in late 2011, “The extent of the test data set generated as part of this program is impressive. It likely exceeds that used as the basis for much of the wind design values in the current building code”
Over the course of SunLink’s relationship with Dr. Kopp and the BLWTL, we have performed over 1000 tests on more than 75 different array and building models. This extensive program has provided insight and design parameters for PV systems deployed in many configurations. Specific issues tested and addressed include:
- Effect of module tilt angle: 2°,5°,10°,15°,20°,30°,35° systems have been tested.
- Effect of height off the roof or ground: various tests from 0” to 4’ clearance.
- Effect of spacing between rows: various tests with spacing between 2 and 4 times height.
- Effect of set-backs from roof edge: various tests from very close to the edge to distances of nearly twice the building height from the side of building.
- Effect of deflectors / shrouds: various deflector designs and configurations applied to the arrays.
- Effect of roof height: from 24’ to 72’ buildings.
- Effect of panel height (chord length) from 3’ to 15’.
- Effect of arrays on roof surface loads.
- Effect of combinations of the above.
While SunLink has used aeroelastic fly-away, and force-balance type models as part of its wind tunnel testing, it has relied almost exclusively on pressure tapped models to generate design parameters. The reasons for this are many, and are best described in the publication, Rooftop Solar Arrays and Wind Loading: A Primer on Using Wind Tunnel Testing as a Basis for Code Compliant Design per ASCE 7, but, fundamentally, pressure tapped models provide the best and most reliable data, and have been the basis of most of the design wind loads used in the building code. They do, however, generate extremely large data sets which must be analyzed.
Over the course of collecting and analyzing terrabytes of data from these tests, SunLink developed proprietary software to explore and visualize the test data.
Software that allows us to see the location and circumstances behind peak pressures on the design envelope.
Time series videos of pressures on the modules.
Developing appropriate design curves for PV arrays, however, turned out to only be part of the challenge. How the design curves should be applied to specific arrays which inevitably differ in some way from the array that was tested (e.g. a different building shape, array configuration or location on the building) is a particularly difficult issue, especially for the case of ballasted arrays.
Structural Analysis Informed by R&D
For guidance in determining the appropriate way to apply our wind tunnel test data to project-level structural calculations, SunLink looked to another expert ― Dr. Joe Maffei at structural engineering firm Rutherford & Chekene. Dr. Maffei is recognized as an expert on “performance-based” design and has been instrumental in helping write building codes and standards for various organizations including the U.S. Building Seismic Safety Council and the Structural Engineers Association of California. SunLink worked with Dr. Maffei, other experts at R&C and Dr. Kopp to create a design methodology for rooftop PV array design that is consistent with the intent of the ASCE Standard.
At the heart of this methodology is the need to run each PV array in a given installation through a non-linear structural analysis program using multiple load cases. Starting in 2008, our R&D partners began developing software that could perform the requisite calculations. This development effort was broken into two stages.
In the first, we built, tested and tuned analysis models of each of our system components. We also determined how to select and apply the appropriate wind load cases. In order to meet the intent of the building code, we configured the software to apply multiple size wind gusts to different sections of the array and check the strength of each component. The process of building models for each sub-array, setting up the load cases and running the analysis proved to be extremely time consuming, taking weeks of effort for each sub-array.
Measuring stiffness and deflection of components and assemblies of components.
More stiffness and deflection testing.
Measuring coefficient of friction between components and various roof surfaces.
SAP model components for a particular array.
The second phase of development was therefore a large software development effort aimed at automating the process. We were able to build software that automatically generates a non-linear model for every sub-array in a given project directly from our AutoCAD layouts. We also developed software that automatically sets up and runs the various required design load cases for each sub-array.
As a result of this extensive software development effort, SunLink can now analyze thousands of PV arrays each year with a relatively small team of engineers. In addition, because of the software, SunLink engineers can optimize array design by efficiently adding or removing connectors and ballast as well as confirming that design standards are met.
Analysis model showing wind gust cases.
Codes and Standards
SunLink and our R&D team are actively working with various groups in the creation of appropriate wind design standards that will assure safety of PV installations. Dr. Greg Kopp and Dr. Maffei are both members of the ASCE 7-16 Task Committee on Wind Loads for the ASCE 7 building standard.