By Michal Bartko, PhD, and Bas A. Baskaran, PhD, P.Eng.
Photovoltaic (PV) systems are becoming an increasingly important element of renewable energy strategies across Canada. For building owners wishing to take advantage of this technology, the roof represents an ideal location for a PV array. A rooftop installation gives added purpose to a large and otherwise unused area of the building. Rooftop solar arrays are generally safe from the threat of vandalism and theft, and by virtue of having good exposure to sunlight with minimal shading, they produce considerably more energy than, for instance, wall installations.
There are numerous choices of PV system types to suit the needs of building owners. Mechanically attached (penetrating) types, while adding more penetrations to the roofing system, provide a reasonably durable solution. Ballasted (non-penetrating) systems, on the other hand, are quick to install and leave the roofing system essentially intact. The disadvantage of the non-penetrating systems is the need for ballast. Before considering this, owners should engage a structural engineer to evaluate the bearing capacity of the roof deck.
As businesses have made investments in PV technology, they have been aided by research and advances in regulations and standards. For instance, the National Building Code of Canada’s (NBC’s) 2015 version incorporates a procedure to calculate wind loads on rooftop solar arrays. In the commentary segment of NBC, Sections 53 to 56 provide a wind load design procedure for these arrays. The Structural Engineers Association of California (SEAOC) has also prepared a design document, PV2-2012, for wind design load calculations. Both these design procedures are derived from existing ones developed for wind load calculations for roofing systems specifically, with a few additions made to account for the effects of wind on PV arrays.
The procedures are limited to low-slope roofs (up to seven degrees). As an example, Figure 1 shows a diagram with sizes of roof corner (3), edge (2), and field (1) zones. Those familiar with roof system wind design will notice significantly larger dimensions of edge width, which increased five-fold from 0.4h for roof systems to 2h (where h is a building height). Figure 1 also presents pressure co-efficient (CgCp) determination diagrams based on the inclination of the PV panel and the normalized wind area.
NRC developed Wind PRA—a simplified online tool to calculate wind loads on rooftop solar systems, based on the 2015 NBC procedure. Once the simple four-step process to input the building and solar array characteristics is completed, the online calculator provides PV wind loads for the specific location and building.
An example of the calculator output is shown in Figure 2.
Wind resistance of rooftop PV arrays
Numerous wind tunnel studies to evaluate the wind resistance of PV arrays have been carried out. Despite this extensive testing and the availability of the standard design methods previously mentioned, failures of rooftop PV systems are not exceptional. For mechanically attached systems, the failure of fastening elements (e.g. clips and bolts) tends to be the most common factor. Further, because there are more penetrations—and, hence, flashings—with these systems, there is the possibility of poor-performing flashings eventually resulting in water intrusion. For ballasted systems, PV array sliding or an individual module overturning are the most common failure modes. An example of an overturned PV array is shown in Figure 3.
To help prevent such failures, NRC is seeking supporters in order to create a consortium to develop a standard for wind resistance evaluation of photovoltaic roof assemblies (PRAs). Such assemblies are defined as a combination of a PV system with a roofing system.
This proposal is a natural progression building on NRC’s previous work on wind effects on roofing systems. For instance, NRC first responded to industry trends by developing a standard for the evaluation of wind resistance in membrane roofing systems—that is, Canadian Standards Association (CSA) A123.21-14, Standard Test Method for the Dynamic Wind Uplift Resistance of Membrane Roofing Systems. When trends changed and vegetated roofs gained popularity, NRC kept pace with another standard to evaluate wind resistance of these assemblies—CSA A123.24-15, Standard Test Method for Wind Resistance of Modular Vegetated Roof Assemblies. The specimen defined in both those standards represents the complete roof assembly, including the roof deck, thermal insulation layer, membrane, and—in the case of the latter standard—vegetation.
The effect of wind on building components is dynamic. Hence, the methods and procedures described above use a dynamic test protocol to evaluate roofing and vegetated systems. Figure 4 shows a diagram of the proposal for wind resistance evaluation of PRAs. Similar to previous NRC works, Task 1 of the proposed project would be to quantify the wind uplift resistance of PRAs, while Task 2 would be to quantify their wind flow resistance. A complete roof assembly combined with a full-scale photovoltaic system (including racking) will be tested.
Test facilities for the experiments will be the same ones used for the development of the above CSA standards—namely, a dynamic wind uplift test chamber and a wind flow test apparatus. This will provide a much more affordable evaluation method than wind tunnel testing. During both tasks, adverse effects of placing PVs on roofing systems (such as compressive strength of the roof system, PV system sliding, and membrane resistance to abrasion) will be examined and quantified.
The goal of this proposed project is to develop an evaluation standard to help the industry achieve safe and standardized installations for roof-mounted PV systems, providing users with de-risked, efficient, and durable technologies for energy production.
Michal Bartko, PhD, has served for two years as research officer with the National Research Council of Canada (NRC) and was previously research director at the National Roofing Contractors Association (NRCA). He is also a post-doctoral researcher at the University of Tokyo with more than 10 years of experience in the study of roofing technology. Bartko can be reached at firstname.lastname@example.org.
Bas A. Baskaran, PhD, P. Eng., is a group leader at NRC, where he researches the wind effects on building envelopes through experiments and computer modeling. He is an adjunct professor at the University of Ottawa, and a member of Roofing Committee on Weather Issues (RICOWI), RCI Inc., Single-Ply Roofing Industry (SPRI), and several other technical committees. Baskaran is a research advisor to various task groups of the National Building Code of Canada (NBC) and member of the wind load committee of American Society of Civil Engineers (ASCE). He has contributed hundreds of research articles and received several awards, including Frank Lander Award from Canadian Roofing Contractors Association (CRCA) and Carl G. Cash Award from ASTM. Baskaran was recognized by Her Majesty Queen Elizabeth II with a Diamond Jubilee medal for his contribution to fellow Canadians. He can be reached at email@example.com.