Reroofing Over an Existing Deck: When Is the Assembly Still Compliant?

The attachment method is the point where I’d insist on written manufacturer confirmation, not just rely on the new membrane’s UL or FM listing. For a future project, I’d want the roofing manufacturer to confirm that the exact layer sequence, including the retained adhered membrane and insulation, cover board type, fastener type, spacing, and perimeter enhancements, is covered by a tested or approved assembly.

That letter shouldn’t be treated as a substitute for the design review. The design professional still needs to verify the existing steel deck’s condition and thickness, fastener pullout capacity, perimeter and corner wind zones, trapped moisture, and the code-required total R-value. If any of those differ from the tested assembly, the manufacturer should identify what engineering analysis or field verification is acceptable.

If the manufacturer can only approve the new membrane by itself, I’d consider that a documentation gap rather than proof the reroof is compliant. That’s usually where the cost comparison changes: a little more investigation up front may be cheaper than discovering later that the retained layers can’t be credited for fire, wind, or energy compliance.

Don’t count retained insulation toward the required R-value unless its type, thickness, continuity, and moisture condition can be documented well enough for the code official to verify it.

An old assembly report or original drawings may establish what was specified, but they don’t necessarily show what remains after years of leaks, compression, wetting, or undocumented repairs. If the insulation is inaccessible, the design may need to base compliance only on new insulation, or use an approved investigation method that can verify the existing layers. Otherwise the project could have a manufacturer-approved wind assembly while still lacking support for its claimed energy performance.

The uncertainty around wet or compressed insulation is also a budget issue, not just a code issue. Before comparing tear-off and reuse bids, I’d price a limited set of exploratory cuts or another agreed investigation method. A roof can look like a good candidate for reuse on paper, then lose much of that advantage when concealed insulation has to be removed, dried, replaced, or excluded from the R-value calculation.

Those cuts should document the actual insulation type and thickness, moisture condition, cover board, membrane layers, and deck condition in representative areas, including near drains, penetrations, seams, and suspected leak paths. It’s worth having the roofing designer and manufacturer agree in advance on how the findings will affect the design. Otherwise a low initial reuse price may simply postpone the tear-off decision until construction, when the change is more expensive and disruptive.

The exploratory cuts should probably check more than moisture and insulation thickness. How well the retained membrane and insulation are bonded or mechanically restrained matters too. Delamination, voids, loose boards, or areas that have separated from the deck could affect the existing assembly’s wind behavior and make the proposed fastener pattern unsuitable, even if the materials are dry and the specified R-value is documented.

Could the investigation report map those conditions by roof zone, especially along perimeters, corners, seams, and around previous repairs? A manufacturer may be able to accept limited variation, but widespread debonding could mean the retained layers can’t reasonably be treated as a stable substrate for the new system.

Mapping debonding by roof zone is useful, but exploratory cuts show conditions only where they are made. Unless the investigation method supports interpolation, areas between cuts remain uncertain, especially near transitions, repairs, perimeters, and corners.

The design documents should set a conservative acceptance threshold in advance: what amount and distribution of separation is acceptable, how unverified areas are treated, and when the retained layers must be removed or the reroof redesigned. A manufacturer letter can address the documented condition, but it shouldn’t turn unexamined areas into assumed continuous restraint.