Volume 43, Issue 5 - May 2008

Glass vs. Plastic   
Should a Human Impact Resistance 
Standard Cover All Skylights?

by Megan Headley 

The popularity of skylights continues to grow due to increasing demand for daylighting and maximum transparency in buildings (see “Lighten Up” in the September 2007 USGlass, page 96). But does more skylights mean more risk of accidents? Nigel Ellis, Ph.D., president of Ellis Fall Safety Solutions in Wilmington, Del., is chair of the E06.51.25 task group, a part of ASTM committee E06.51 on Performance of Windows, Doors, Skylights and Curtainwalls. The task group is working to produce a “Specification of Human Impact Criteria, with Procedure for Testing and Rating Plastic-Glazed Unit Skylights and Related Products used on Commercial Walkable Roofs for Fall-Through Resistance.”

According to Ellis, the task group was formed “because the toll of occupational deaths is constant from year to year. And, since the design of skylights is controllable by manufacturers, it seems that all skylights should be tested with a uniform test method. That includes glass, plastic and … fiberglass panels.”

The United States Bureau of Labor Statistics (BLS) reports in its Census of Fatal Occupational Injuries for 2006 (the most recent data available) that there were 50 nonfatal falls through skylights and 36 such fatal accidents. The numbers have varied little in the years since this data has been recorded (see chart on page 45). 

Keep in mind, too, that these numbers reflect only falls that occurred on the job, and not those accidents frequently reported in the media when children find a way to play on the roof at school and other such cases. To lower these numbers all around, the task group is developing a test method to test for the impact resistance of all types of skylights. 

“I think that what Occupational Safety and Health Administration (OSHA) would like to see is one test for all,” Ellis says. That means one test for both glass skylights and plastic skylights. And some glass skylight manufacturers fear that just might be the case. 

Don’t Tread on Me
It’s not that skylight manufacturers are resistant to improving safety factors—far from it. “The members of the American Architectural Manufacturers Association’s (AAMA) Skylight Council fully embrace the concept of acknowledging all aspects of health, safety and welfare relating to their products,” says Ken Brenden, technical standards manager for AAMA. 

The Skylight Council is among the industry groups involved in the drafting process. “The members are receptive to standards being developed for impacting ‘unit’ skylights. However, more studies, market analysis and involvement by other associations are needed before human impact loading standards can be effectively developed for all skylights.” Anything that can make a product safer should be embraced. But are glass skylights really a danger?

“The dangers of skylights have been dramatically inflated,” according to John Westerfield, who handles marketing and code compliance for CrystaLite in Everett, Wash. 

CrystaLite is a manufacturer of glass and plastic skylight products. “The fall report data is skewed in that items not defined as skylights by the International Building Code (IBC) have been lumped into these fall statistics, as well as simple holes on the roof during construction. There are tens of millions of skylights in the United States—and a relatively small number of accidents occur annually.”

It’s not simply that there’s a small number of falls through skylights; it’s that there’s purportedly a small number of falls through glass skylights—or at least, no data currently available indicating that glass skylights are a major safety hazard as compared to other types of surfaces.

“I do not believe that there is any evidence at this time which indicates that glass skylights have been involved in any fall-throughs,” says Randy Heather, standard products manager at Naturalite Skylight Systems, a part of Oldcastle Glass in Santa Monica, Calif.

“The further analysis of the fall data will indicate how many falls are through an actual manufactured skylight,” Westerfield says. 

And, as these glass manufacturers note, plastic skylights are as likely to lead to trouble as glass—if not more so. For example, Westerfield says, “Plastic domed skylights are typically installed on flat roofs where workers are more likely to be exposed to skylights where the already established OSHA fall protection requirements may be overlooked. This is typical when a worker must access a flat or low-pitched roof to do a one- or two-day job post-construction. Glass skylights typically are installed on sloped roofs, in which case workers are more likely to follow OSHA requirements and are protected from all falls with a personal fall arrest system of some sort.” 

“You get a lot of plastic, flat skylights, that will be on Walmart, Home Depot, Lowe’s—your commercial buildings with a flat roof,” agrees Tom Kozak, sales director of Acurlite Structural Skylights in Berwick, Pa. As he notes, “A monumental skylight—glass and aluminum—is definitely going to be a lot higher [than a plastic skylight] and at a pitch. It’s not something that can be easily walked on or fallen into. More I would see people tripping over a plastic skylight … than I would a monumental skylight.” In addition, characteristics of these plastic skylights lead glass product manufacturers to believe that’s where the danger lies.

“I would think these plastic or bubble skylights are more susceptible,” says Ron Palombo, president of Acurlite. “I’m not saying they’re not strong, because you can get these things with remarkable test performance capabilities, but unlike the properties and the rigidity of glass—laminated glass, tempered glass—I would think these bubbles or plastic skylights are more susceptible to movement, susceptible to deflection.”

In addition, there is the concept of “rapid aging” mentioned in the scope of the ASTM standard: “Skylights can fail without warning due to rapid aging and there is a need to protect hundreds of thousands of skylights that were installed decades ago,” it reads. While all skylights face aging, this term refers to plastic products. “Ultraviolet (UV) weathering may cause plastic glazing to deteriorate over time, which can greatly reduce its impact resistance,” Westerfield says.

The task group will be looking at all ways in which skylights may be affected by weather and age. “We need to have some sort of idea of how the glass laminates age with UV, too,” Ellis notes. “But the accelerated weathering is a difficult prospect to deal with because it’s done on an individual skylight basis and we have to find new methods to do accelerated UV testing that rely on samples and a considerably shorter period of time to do more years.”

According to information from Acurlite, normal weathering and aging of glass skylights that are 20 to 30 years old usually consists of failed insulating glass seals, breeches in exterior silicone seals and or surface failure, but not structural failures. Structural failures are normally attributed to poor engineering, design and installation, the manufacturer reports. 

Put to the Test
At this point, lacking test data, this test method still is mostly speculation. “There’d be fewer constraints on glass if glass [passes] the test that we decide is relevant,” Ellis says. To ensure that the data collected accurately reflects all types of skylight products, glass manufacturers are encouraged to participate in the process. “The ASTM task group involved with this activity is in the process of collecting data and seeking to establish realistic testing methods for all openings in roofs that could present a hazard,” Brenden says. 

Westerfield adds, “More in-depth analysis of fall data is currently being processed and will be invaluable in determining the ultimate specification and test method for providing fall protection in regard to skylights.” How much could this impact resistance standard change the manufacturing process? “Until such time as a draft test method is prepared and balloted through ASTM, the potential impact to the way in which skylights are manufactured and installed cannot be determined,” Brenden says. 

Westerfield has been representing CrystaLite as an active member on this ASTM specification work group. “Our commitment to be an active member in the development of this standard was made after viewing the initial specification proposal and test methods,” he says. 

“It was clear that little to no research was done as to what the IBC defines and prescribes as to what a skylight is and to what standards it shall be tested to, as well as the limitations to which a testing laboratory could accomplish realistically. Cost effectiveness in testing procedures was not taken into consideration, which, if left unchecked, could have a large financial impact to the industry. In short, we had extreme concerns with the scope and testing procedures as they were.” And, of course, skylight testing does occur already. “A lot of codes ask for a 250-pound concentrated load at any one point on the framing,” Palombo says. 

He notes that many products currently are tested for hurricane resistance—but human impact resistance is another story. “Impact testing for that kind of simulated testing is very extreme and it would more replicate what could happen to a skylight in an extreme condition and beyond someone falling on it. You’re talking about projectiles hitting the skylights at a certain velocity. Not a person walking on it,” Palombo says. But if glass is not found to be a risk, as Ellis noted above, glass skylights may not even be affected by an eventual test method.

“I don’t think that [the standard] will have a big impact on Major’s Guardian 275® or Auburn® glass structural skylights, but it will have a big impact on the polycarbonate and acrylic market,” says Jeff Wirkus, compliance manager of Major Industries Inc. in Wausau, Wis.

The difference between testing all and testing just plastic skylights has huge implications for manufacturers. “If glass is allowed to remain in the scope of this specification; this potentially could have a huge impact on all glass skylights,” Westerfield says. “The current specification, as initially proposed, would require all skylights to pass an impact test under temperature conditions of -40 degrees and 120 degrees … after an outrageous 40 years of Florida sun weathering. This is why we and other manufacturers have taken an active participation in the development of this ASTM standard.”

Danger Ahead
Is testing the answer? Are there other ways to prevent the steady numbers of accidents and injuries resulting from falls through skylights?“

All skylights already are required to have labels on the frames which indicate that the skylight will not support the weight of a human,” Westerfield says. Adding to existing labels likely won’t be keeping people at bay. He notes, “A manufacturer of any product line must be cautious not to over-label any product to prevent the user to be oversaturated, therefore not read any of them.” Wirkus adds, “Who looks for the label and what good is a label during an accidental fall? The best way to protect against an accident is to guard against it by other means than the just the skylight framing and glazing.”

According to information from Acurlite, protective screens can be supplied if called out in the skylight specifications, but are used more frequently with plastic unit skylights than with glass and aluminum. “Many skylight manufacturers, including CrystaLite, have products and methods to protect a worker from an accidental fall through a skylight when such a system is requested,” Westerfield adds, “and have long sought an accredited test standard in which the skylight itself could provide fall protection.” Maybe additional, or improved, labeling isn’t the answer, then. 

“The committee will have to decide … what is an adequate warning and what it should say,” Ellis says. “Certainly there is much greater skill levels that need to deal with the prescriptions of warning label standards, which are quite advanced right now in this country and part of the Z535 family of standards. So we’ll have to discuss the relevance of warning labels. How long do they last? They also have an aging problem.”

More to the point—is it really the responsibility of the skylight manufacturer to prevent accidents that occur because people are improperly using their products or disregarding safety regulations? The Glass Association of North America (GANA) notes in its glass informational bulletin “Skylights and Sloped Glazing Are Not Walking Surfaces” that “while architectural flat glass can be designed and engineered for use as a walking surface, glass typically used in skylights and other sloped glazing applications is rarely designed to support the weight of live loads such as a human standing or walking on the glass …” Should it have to be?

“Major Industries does make clear in its installation manuals that walking on our skylights is prohibited,” Wirkus says. “Currently, the fiberglass used in our Guardian 275® translucent panels are tested using a form of UL-972, but this test is really for rating the ability to stop an intruder and not as a primary means of protection against an impact from a fall or someone running onto or over a skylight.”

As the statistics indicate (see chart on page 45), it’s construction and maintenance professionals who most run into trouble with skylights. Skylight manufacturers note that safety standards exist to protect these individuals and should be strictly enforced. “We have a close network of dealers/contractors who go to great lengths to protect themselves by requiring all of their workers to follow safety guidelines already established by OSHA,” Westerfield says. 

“The case reports presented at this point indicate that the roofer/contractors installing the skylights are not those who are at risk of falling through skylights. Rather, it is those that access the roofs post-construction for various short-term projects that typically overlook the already required safety procedures, therefore putting themselves at risk for all falls from the roof surface.”

“You’ve got to follow all of OSHA’s guidelines,” Palombo adds. “If you come within six feet of an opening or a hole, you have to be completely tied off. And, when working with skylights or replacing glass on a skylight, you must follow all of OSHA’s guidelines because, in theory, you are opening the skylight and creating that hole again.”

 If a worker is putting himself in danger by not using proper fall protection—how much can a skylight standard really do to protect them? Aren’t they in as much danger by virtue of being on a rooftop?

The Number One Priority
Safety is the most important part of any job, any (responsible) construction professional will agree, whether it’s a glazier or roofer or anyone else. A professional on a roof—or scaffold, or other dangerous location for that matter—should be wearing fall protection equipment and no amount of product testing will protect a worker who isn’t protecting himself. “I’ve seen people in harnesses get yanked off the side of the building. Thank God they had a harness; they’re hanging a unit and the wind picks up the unit and them and takes them right off the side of the building,” Palombo says. 

“Safety has got to be number one or you’re going to get hurt. You’re going to fall.” Keeping contractors and others away from skylights is intended to lessen the danger. What about the danger, Wirkus asks, that once skylights are tested to prevent falls, then the misuse of the products actually grows? “Giving a skylight an impact resistance rating would only give individuals a false sense of security,” Wirkus says.

“Skylights are designed to allow light to enter the building, and are not to be used as a walking surface. By giving ratings to a new generation of skylights, we’re giving people the green light to walk on all skylights, putting the general public in greater danger. How is one to know if the skylight is new or old and non-rated?” 

At this point, the goal is simply to determine if any action can be taken to increase the safety of anyone who sets foot on a roof. “We look forward optimistically to having the manufacturers provide the confidence that the public needs and the trades people need to do work around skylights whether it’s with additional protection or at least they’ll have a test that assures the public that that particular skylight works for their safety,” Ellis says. 

OSHA Cites Winter’s Architectural Roofing for Workplace Safety Violations
On April 2, the U.S. Department of Labor’s Occupational Safety and Health Administration (OSHA) issued Winter’s Architectural Roofing Co. in Carbon Cliff, Ill., eight citations for alleged safety and health violations following its investigation after the death of one employee. 

Proposed penalties total $224,000.On October 10, 2007, an employee was killed when he fell 16 feet through a skylight. OSHA issued eight willful citations to the company for its failure to provide fall protection in hoisting areas and on low-sloped roofs, failing to cover skylight openings to prevent falls and not training employees about fall hazards. Seven of the citations allege per-instance willful violations of three OSHA requirements. 

A willful violation is defined as one committed with an intentional disregard of, or plain indifference to, the requirements of the Occupational Safety and Health Act and regulations.

Prior to this investigation, OSHA inspected Winter’s Architectural Roofing in 2002, resulting in a serious citation for failing to provide fall protection during roofing operations. The company has been in business since 1937.Under the Occupational Safety and Health Act of 1970, employers are responsible for providing a safe and healthy workplace for their employees. 

OSHA’s role is to promote the safety and health of America’s working men and women by setting and enforcing standards; providing training, outreach, and education; establishing partnerships; and encouraging continual process improvement in workplace safety and health. 

Megan Headley is the editor of USGlass magazine.

© Copyright 2008 Key Communications Inc. All rights reserved.
No reproduction of any type without expressed written permission.