Evaluating OSB for Coastal Roofs

After Hurricane Andrew, Florida code advisers ruled OSB sheathing inferior to plywood. Was the judgment too hasty? Evidence suggests that OSB performs well when fastened properly, but the jury is still out on impact resistance.

by Paul Fisette – © 2005

Most building contractors have made the switch. They now use oriented strand board (osb) as a substitute for plywood roof sheathing. Initially, the decision was driven by the low introductory price of osb. The honeymoon is over. As of this writing, a sheet of osb sells for a few dollars less than plywood, but the marketplace conversion to osb continues.

Osb’s acceptance has been slow in coming. Early on, osb got a bad rap because of its association with waferboard a sheathing material with an amorphous blend of wood flakes that became commercially available in 1963. The first true osb with layers of aligned fibers was introduced byt the Elmendorf Manufacturing Company of Clairmont, NH in 1982. Among builders, its initial reception was chilly. Response warmed gradually as osb established a reliable track record and now osb dominates the national market for residential sheathing, accounting for 60% or more of the structural sheathing sold nationwide. While many builders have become comfortable using osb, new suspicions surfaced when code advisors in Florida’s Dade and Broward counties charged that osb fails as structural roof sheathing in wet, windy exposures, such as those caused by Hurricane Andrew’s rampage across southern Florida in August 1992.

A History of Roof Failure

Hurricane Andrew hit south Florida with a vengeance. The damage was formidable and has only recently been surpassed by the quadruple whammy of 2004’s record-breaking storms. A public outcry moved the government to action. It created an advisory committee to investigate the level and nature of structural damage, determine why the structural damage resulted, and make recommendations to prevent any future reoccurrence. Jose Mitrani, a civil engineer and professor at Florida International University was involved from the beginning.

Mitrani was appointed to a task force charged with investigating the failures caused by Andrew. According to Mitrani, the committee held hearings for several years. In the beginning hearings were held 3 days a week, 4 or 5 hours per day. Producers, builders, designers, regulators, and other key players provided testimony. As a result of his experience, Mitrani does not think that osb performs adequately as structural roof sheathing. He states, “During that experiment we call Andrew, osb did not behave as an equal to plywood. I saw shattered osb all over the place and I did not see a lot of plywood.”

Mitrani severed as vice chair of a Task Force subcommittee investigating the performance of osb. The subcommittee reviewed mountains of reports submitted by consulting engineers and industry members. APA the Engineered Wood Association, a trade association representing the osb and plywood manufacturers provided test data demonstrating how osb performs when it’s wet and when it’s dry. Mitrani says, “The data was convincing. But when we plotted the data our way, the behavior of osb was simply not as good as plywood.” Mitrani is convinced that dry osb behaves as well as or even better than plywood, but wet osb does not. He thinks that osb swells when it is wet and causes nails to pull through the panels easily. APA rejects Mitrani’s condemnation.

Ed Keith, senior engineer with APA was there from the beginning too. He was part of Florida’s official damage assessment team and claims, “The problems with roof sheathing were related to the fact that builders were not building correctly, and the inspectors were not inspecting adequately.” Keith insists that APA’s quality assurance testing of osb and plywood is reliable and as result, builders get durable roof sheathing. Panel manufacturers must submit to random testing to earn an APA grade stamp. APA-stamped osb and plywood are tested for bending strength, stiffness, shear strength, and concentrated load performance. APA tests glue bonds with an accelerated weathering wet/dry cycle tests to assure durability and measures fastener holding capability. The APA tests outlined in PS2/PRP 108 are product performance standards and they predict that osb and plywood perform as equals for their intended end use. This begs the question: Why did Mitrani and others see so much failed osb during inspections? Proponents of osb say it’s simple. Most homes were built with osb and the installation was poor.

“When I did inspections after Andrew, I literally saw panels with only 4 nails in the corners of the panels and roof framing members with no holes in them indicating that the sheathing had never been nailed off or that sheathing nails had completely missed the roof framing targets.” says Keith. He claims the overall consensus among experts who have looked at roof-sheathing failures report that failures resulting from hurricane Andrew were related to “workmanship and improper fastening” not the use of osb.

John Pistorino, a principle of Miami-based Pistorino and Alam Consulting Engineers considers osb to be a viable product. Pistorino has been a special consultant for the South Florida Building Code since 1974. His company was a key participant in the production of the 1993 FEMA report Building Performance: Hurricane Andrew In Florida. “Every picture in that report was mine. I saw more of the hurricane damage than most people because of the position I was in.” says Pistorino. The tests he conducted convinced him that osb provides the structural capacity required for roof sheathing. According to his investigations, osb is not much more susceptible to water than plywood.

Pistorino states, “Immersing osb in water for a long period of time certainly shows that osb swells more than plywood, but I am not sure this is what you see as a normal condition in the field.” He believes that osb roof sheathing is more prone to rot and weakening when a roof is allowed to fall into disrepair, but considers this a maintenance issue. Pistorino found that improper fastening was the cause of most roof-sheathing failures during Andrew. This is the opinion expressed in most reports. He advocated the banning of staples that is now part of the current Southern Florida Building Code. Tighter nailing schedules and ring-shank nails are now recommended. Florida currently has a statewide building code that allows the use of osb and plywood as roof sheathing – except in Dade and Broward counties.

Code View
Recently, I served on a National Research Council panel where we determined that there were literally thousands of code interpretations enforced in this country. Even states with a statewide code have communities with different interpretations and requirements for a given application. So it’s hard to say anything absolute about “the code.” However, there has been a move to create a uniform code through the development of the International Code Congress (ICC), which has been adopted by 18 states. The ICC permits the use of osb roof sheathing interchangeably with plywood in its 2003 International Residential Code (IRC).

Section R803.2 of the IRC regulates “Wood structural panel sheathing” based on standards used to qualify panel performance, not the material used to make the panel. Of all the building codes that I am familiar with, it appears that Dade and Broward counties stand alone in its ban of osb as an acceptable roof sheathing material. The reason: As it turns out, osb fails the large missile impact test, where a 9-pound 2×4 is fired at a panel from a cannon at a speed of 34 mph. The bar is set at the level of resistance provided by 19/32-inch thick plywood, which is prescribed by Dade and Broward counties. Osb must be 30% thicker to pass this test, making it economically non-competitive even if osb were allowed by these codes. Some experts think the established test limits are arbitrary. They think it might make more sense to fire a 25-pound roof tile at 50 mph and see what happens to a panel.

Performance Characteristics
Scientists and engineers who study the performance of wood-based materials generally think that the structural performance of osb and plywood are equivalent. However, these are different materials with different characteristics worth considering.

Structural-use plywood and osb share the same Exposure 1 durability classification (95% of all structural panels are Exposure 1), suitable for temporary exposure to the weather like those experienced during construction. Osb and plywood also share the same set of performance standards and span ratings. Both materials are installed on roofs following the same set of recommendations. The prescription for blocking, fastening, and use of H-clips are identical. Independent research conducted by professor Poo Chow, a researcher at the University of Illinois, and others; demonstrate that nail withdrawal and pull-through for osb is as good or better than plywood. One of the big differences in overall sheathing performance is related to their response to moisture.

Plywood gets saturated much faster than osb. Plywood can get saturated when exposed to a couple of days of rain. It will take more than a week to saturate osb. However, plywood dries quickly and osb does not. And some consider this the Achilles heel of osb.

The biggest knock leveled at osb is its tendency to swell around the panel edges. It swells faster around the perimeter of the panel than the center. This swelling is irreversible. Once osb swells, it stays swollen and expanded panel edges can telegraph through thin roof finishes like asphalt roof shingles. Manufacturers have reduced the likelihood of Ghost Lines, the telegraphing of panel edges, by coating osb panel edges to reduce water absorption, but the phenomenon is still common. Plywood does not have this problem. Plywood swells uniformly, less dramatically, and reversibly. Compounding this weakness is the fact that osb is often made from aspen and poplar, which are not rot-resistant woods. Since osb holds water, prolonged exposure to roof leaks or excessive humidity and condensation in unvented attics can cause osb roof panels to degrade faster than plywood panels. Osb, in its current state of development, is more sensitive to moist conditions. Plywood is more forgiving. However, proper construction and appropriate maintenance will generally result in good service from either material.

Wood is a widely variable material, but osb is a fairly consistent product. Osb is more homogenous than plywood. It is perhaps 50 strands thick as opposed to the 4-veneer thickness of plywood, so the range of variability is narrower in osb. All sheets are consistently stiff and strong. While plywood may be slightly stronger on average, osb is more uniform so you will sense less difference from sheet to sheet. You never have a “soft spot” in an osb panel because 2 knot holes overlap. And you don’t have to worry about knotholes at the edge of a panel where you are nailing the panel in place.

Lessons Learned

Hurricane Andrew redefined the way we think about risk and damage. The dramatic increase in population along our coastlines has upped the anti when it comes to structural damage control. It’s been estimated that more than 80% of the losses caused by Andrew were related to roof failures and associated water damage. So roof construction deserves special attention. Improperly fastened roof sheathing is easily ripped from framing supports. And significant damage results when windborne projectiles crash into walls, windows and roofs. As wind enters the structure, it acts against the interior surfaces of the home, which are not designed to withstand this force. Roof punctures may not immediately level a home, but do allow rain to penetrate. Interiors are wet and structures weakened as rain-soaked ceilings collapse, removing the reinforcement provided by ceiling joists.

There are several things that can help minimize roof-sheathing failure in hurricane zones. Build homes with hip- or steep-pitched roofs when possible. They withstand high winds best. Fasten roof sheathing beyond what code requires: glue the sheathing in place; increase nail-size and frequency at the corners and along the edge of the roof; and inspect the roof from the attic side to verify that all nails hit the rafters. Don’t use staples or clipped-head power-driven nails. They don’t provide adequate resistance in high-wind exposures. Roof shingles will blow off. Apply strips of peal-and-stick membrane or flashing tape over all seams in the sheathing to prevent water leakage. And then keep your fingers crossed.

Most reports blame roof-sheathing failures on shoddy construction and improper fastening. Proper fastening as outlined by APA and the IRC work equally well for osb and plywood roof sheathing. Osb and plywood roof sheathings work fine when installed and maintained correctly. But osb can’t escape the fact that it simply is not as resistant to impact as plywood. Is osb’s impact resistance “strong enough?” Many experts think it is. But when pressed, few will give up the advantage plywood provides in High Velocity Hurricane Zones (HVHZ) to save a few dollars per sheet.

Last updated: April 1, 2009