Phil Childs checks the moisture content in roof insulation for Building 2518.

As homes and commercial buildings age, their roofs often begin to leak. When a roof is 15 to 20 years old, a typical building owner will have it replaced. But, is a costly replacement of a roof needed so soon? Not always, say researchers at ORNL's Buildings Technology Center.

The researchers are finding ways to dry a wet roof and extend its life by five years at one-third the cost of replacing it. They also are identifying methods of building new low-sloped roofs that dry easily and can be re-covered to make them last longer. These techniques, they say, will save energy and money.

BTC researchers David Kyle and Andre Desjarlais have written an Assessment of Technologies for Constructing Self-Drying Low-Slope Roofs, which evaluates currently available techniques for removing water from flat roofs. According to Tom Smith, director of the National Roofing Contractors' Association, this report has been cited by roofing contractors as "one of the most important reports . . . to come out in the last five years."

Traditionally, after 15 years of use, an old roof assembly is torn from the metal deck and replaced at a cost of about $9.00 per square foot. Adding to the expense is the cost of landfill disposal of the large volume of old roof parts, including asphalt, foam, and asbestos, which are no longer classified as a single construction waste.

To reduce the cost of roof repair to one-third that of the traditional tear-off approach, more and more building owners are asking roofers to dry out their roof insulation and cover it with a new top layer. BTC researchers say this approach offers benefits to consumers and the country.

"An increase in roof service life of 5 years in the United States should reduce the cost of roofing by 21%, saving the country $2.5 billion a year and cutting by 25% the landfill waste from roofing, which currently represents almost 4% of the total volume of solid wastes in the United States," says Jeff Christian, manager of ORNL's Buildings Envelope Research Center. "This savings is possible if roofers could dry out and re-cover old roofs and build new self-drying roofs that can be safely re-covered later."

Starting from the bottom up, a typical roof comprises a corrugated metal deck, two layers of foam insulation, and a membrane made of sheet rubber or plies of felts mopped together with asphalt. Water eventually enters gaps in the roof assembly, leading to its deterioration. As the moisture accumulates, metal fasteners may corrode, the ability of roof insulation to keep out heat may decrease by as much as 40%, and dripping water may enter building interiors, causing damage and motivating owners to consider roof repair.

How can an old roof be dried most rapidly before covering it with a new membrane? ORNL researchers and their industry partners at the Buildings Technology Center have identified effective ways to reduce and prevent moisture accumulation in roofs.

"We are studying moisture and heat flow in low-slope roofs to learn how to make them last longer," Christian says. "To determine the best alternative solutions for roof repair, we have combined field diagnostics, laboratory experiments, computer modeling, and assessments of available techniques for drying out roofs."

Researchers have found some special materials, called vapor retarders, that block the flow of water into a roof assembly. Other materials absorb water as liquid and release it downward as vapor. When this vapor reaches the building interior, it can be removed by air conditioning.

Schematic of an engineered self-drying roofing system envisioned for the future.

BTC researchers Christian, Desjarlais, and Phil Childs in collaboration with Dow Corning researchers are demonstrating "downward drying" in Building 2518, which is home to ORNL's Plant and Equipment Division. Using infrared cameras and neutron sources, they found that 40% of the insulation in the 27-year-old roof was saturated with water.

They repaired the leaks and increased the roof's insulating ability from R-2 to R-13 by spraying the roof with polyurethane foam. Before reroofing, holes were drilled in the metal deck and vapor retarder to let vapor from the insulation pass through the roof deck below. In this way, the wet insulation dried out.

In the summer the large difference in the pressure of vapor in the roof and in the air-conditioned space inside forces the vapor down into the building, where it is removed by air conditioning. The roof's outer membrane was covered with white granules (donated by 3M Corporation) that reflect sunlight more effectively than typical black roofs, reducing the air-conditioning load on the building.

The researchers removed wet roof samples from Building 2518 and thermally tested them in the Large-Scale Climate Simulator while weighing them on "load cells." The Large-Scale Climate Simulator was programmed to reproduce summer outdoor weather conditions. They measured loss of moisture mass over time, confirming downward drying.

They also conducted neutron gauge surveys of the moisture content in the Building 2518 roof to measure vapor losses. They sent sample materials to the National Institute of Standards and Technology to measure their vapor permeability and capillarity. Based on periodic data obtained, they modeled the roof system on the computer so they could predict how fast the roof will dry out.

By not replacing the roof at Building 2518 by the typical procurement process and by having Dow Corning re-cover the roof as part of its contribution under a cooperative research and development agreement, the Buildings Technology Center effort saved ORNL $250,000 in roof replacement and disposal costs. Based on computer modeling of energy use before and after the roof work, the researchers found that the installed insulation and reflective granule reduced heating costs by 42.5% and cooling costs by 18.5%, saving the Laboratory $2300 a year in energy costs.

"From our experiences with Building 2518, we learned that roofs that have vapor retarders will not dry out easily," Christian says. "For the engineered self-drying roofing system of the future, we recommend installing a vapor retarder only if it is water-permeable and a perforated deck so that water vapor will pass into the interior for removal by air conditioning. We also recommend an easily removable membrane as well as permeable insulation and a wicking layer that absorbs liquid water and passes it into the building as a vapor.

"Lessons learned from this work are being incorporated into a software program that will help roofing contractors decide whether re-covering individual old roofs is a viable alternative. The software may also guide the design of future roofing systems."

Christian says that the Building Envelope Research Center is negotiating users' agreements with roofing companies on designing engineered self-drying roofs for new buildings. "We think our knowledge can be applied to new construction, as well as roof retrofits," he says. "Our research results and experience suggest that re-covering self-drying roofs and old wet roofs that are dryable should extend roof life, reduce landfill waste, improve building energy efficiency, and save consumers money."

New Type of Steel Stud Tested at ORNL

Because steel is becoming cheaper than wood, is recyclable, and is resistant to mold and termites, steel studs are increasingly being selected by builders over wood studs for home construction.

Recently, a Buildings Technology Center user, aided by an ORNL researcher, tested a new type of steel wall stud that will enable future new homes to hold in--or keep out--more heat. The tests showed that the new studs would make walls 10 to 15% more energy efficient than wall systems containing conventional steel studs. The results suggest that widespread use of new steel studs in new home construction, when compared with conventional studs, will reduce the consumption of fossil fuel to heat or cool new homes and lower carbon dioxide emissions.

LeRoy Landers, a professor of architecture at the University of Pennsylvania and a candidate for a doctorate in architecture design, conducted a series of tests at ORNL using BTC equipment. He conducted his research with the help of Andre Desjarlais of ORNL's Energy Division.

Thermal performance is improved because of the metal stud's reduced surface contact with the inside and outside wall sheathing. The experimental steel stud was tested in a wall with fiberglass batt insulation filling the cavity which is sandwiched between plywood and gypsum board sheathing. The wall was exposed to a temperature difference across the wall of 50°F in BTC's Rotatable Guarded Hot Box.

One use for measurements on new wall systems such as Landers' is to help develop a way to evaluate the thermal performance of a total wall system, not just its insulation. Jeff Christian, manager of BTC's Buildings Envelope Resesarch Center, says, "We must consider the whole wall system, including structural parts such as studs and materials used to hold walls to other walls, doors, windows, roofs, and floors. Our goal is to offer the building industry a method of developing a new consumer label that represents the whole wall and not just the misleading `center-of-cavity' R-value rating on insulation."


Where to?

[ Next article | Search | Mail | Review Home Page | ORNL Home Page ]