Letters

SPRI responds to NRCA's concerns with steel roof decks

NRCA Associate Executive Director Mark Graham's "Concerns with steel roof decks," January issue, page 12, is the subject of an industry issue SPRI has been working on for several years. The questions raised in the article officially became an industry concern when FM Approvals formally notified participants at the April 2013 Roofing Industry Coalition Meeting that it was revising FM Approvals requirements to include limitations on all steel roof deck assemblies as determined when using Steel Deck Institute (SDI) calculations. The limitations are not restricted to seam-fastened single-ply systems.

Although FM Global had not experienced any steel roof deck failures, it, along with FM Approvals, decided to combine steel roof deck design found in Chapter 22 of the International Building Code (IBC) and roof system design in IBC's Chapter 15 to help its customers better demonstrate compliance with the IBC.

Because of FM Approvals' changes, Miami-Dade County, Fla., was procedurally required to revise all National Oceanic and Atmospheric Administration listings that had been based on approval reports issued by FM Approvals. As a result, these changes have attempted to transfer the responsibility of steel roof deck design from registered professional engineers to roofing material suppliers.

SPRI has been working with the National Research Council of Canada (NRC) to validate, through wind-uplift testing, the SDI design recommendations. What SPRI and NRC have learned so far is:

• SDI steel deck specifications are outdated because of increases in recycled content of coils and need to be updated. Steel decks available in the market for roofing material suppliers that are specified as 33 and 80 kilopounds per square inch (ksi) typically have actual yield strengths of 50+ and 100+ ksi, respectively. Moreover, it is nearly impossible to acquire 33-ksi and 80-ksi steel decks. Accordingly, this reality should be considered in the SDI design and FM Approvals process.
• NRC's measured deck deflections (permanent and transient) and calculated bending stresses are lower than SDI calculated values.
• SDI calculations do not consider the failure mode of the roof assembly.

When completed, the research and work developed through the Special Interest Group for the Dynamic Evaluation of Roofing Systems will be shared with the roofing industry. This will facilitate improved dialogue among all stakeholders, including steel roof deck and roof system designers.

Mike Ennis
SPRI
Waltham, Mass.

Graham responds: Thank you for your letter and SPRI's acknowledgement of the issue.