by Ronald J. Ray, RA, CCS, CCCA, CSI, AIA
All roofs eventually leak—it is just a matter of when and where. Nevertheless, the hope is that new roofing systems do not leak right from the start. It is critical to verify the watertightness of roofing, especially if it is to be covered with ballast or a vegetated roof assembly. This verification is a field quality-control measure beyond the scope of a roofing manufacturer’s visual inspections for issuance of a warranty. For existing buildings being considered for a reroofing program, conducting a roof survey to determine the location and extent of wet substrates is essential to making fiscally responsible decisions related to the program’s extent.
Flood-testing to detect leaks and verify the quality of newly installed assemblies has a long history in the roofing industry. ASTM D5957, Standard Guide for Flood Testing Horizontal Waterproofing Installations, is a guide for conducting flood-testing of horizontal surfaces with slopes no greater than 1/4 in. per foot (i.e. 1:48). It is commonly referenced as the standard for flood-testing roof systems. However, ASTM D5957 specifically states the standard “is not intended for use on building roofs.”
ASTM D5957 references a maximum water depth of 100 mm (4 in.). This translates to an added live load during testing of 101.34 kg/m2 (20.76 psf). Water depths and the associated live load at roof drains would be even higher. Further, the National Roofing Contractors Association (NRCA) does not recommend flood-testing of new roof systems. According to Joan P. Crowe, AIA, NRCA’s director of technical services, flood-testing is inappropriate for identifying potential leak sources. Crowe states roof systems are designed to resist the passage of water with a minimal amount of hydrostatic pressure (i.e. flowing water), while waterproofing systems are designed to prevent the passage of water under hydrostatic pressure (i.e. standing water).
For example, roof drains are not designed to be leak-free under hydrostatic pressure, so it is highly likely they would leak at this location’s flashing during a flood-test. Although flood-testing can determine the existence of a leak, it cannot precisely identify the location, even with use of soluble dyes in the water. If flood-testing is still desired by a specifier, it may need to be performed in segments when roof drains are separated by large distances. Critical sheet flashing conditions and terminations may be similarly assessed by localized water testing.
Other technologies, such as electronic leak testing, can detect leaks with far more reliability than flood-testing. Some electronic leak-testing methods include:
- electrical capacitance/impedance testing;
- infrared thermography;
- nuclear hydrogen detection;
- low-voltage electrical conductance testing; and
- high-voltage spark testing.
Electrical capacitance/impedance testing
A nondestructive procedure used to identify entrapped water within a roof system, electrical capacitance/impedance testing may not produce a significant result on new construction. This method works due to the ability of water to store or conduct electrical energy. The procedure involves placing a hand-held moisture meter on the surface of the roof, which sends an electrical current into the roof without penetrating its membrane. Wet insulation provides less resistance to the electrical current than dry insulation. Therefore, if the underlying material is wet, the electrical conductance is great and registers on the meter.
It is important to note these meters only determine whether the area is wet or dry. They do not quantify the percentage of moisture present within the roof insulation, nor do they pinpoint the location of the leak. Additionally, the roof surface must be dry in order for these meters to produce accurate results. ASTM D7954/D7954M, Standard Practice for Moisture Surveying of Roofing and Waterproofing Systems Using Non-destructive Electrical Impedance Scanners, establishes requirements for conducting electrical impedance scans of roof assemblies.