Foam Roof Blister Repair and Removal in Phoenix

Overview

A blister on a foam roof is a pocket of gas or vapor trapped somewhere it should not be, usually between the coating and the foam, between foam passes, or between the foam and the deck. Phoenix makes them worse than almost anywhere else. When a white roof surface hits 150 to 170 degrees on a July afternoon, trapped air and moisture expand hard, and the elastomeric film above them is the weakest link. The blister grows a little each hot day and shrinks each night, and that daily flexing eventually work-hardens the coating until it cracks open.

Understanding why a blister formed determines how it gets fixed. Coating blisters, the shallowest kind, usually come from applying coating over a damp, dusty, or poorly primed surface, or from spraying on a scorching afternoon when solvent or moisture flashes off too fast. Interlaminar blisters form between foam passes when a lift was applied too thick, too cold, or over a contaminated previous pass. Substrate blisters, the most serious, mean the foam has lost adhesion to the deck itself, often because of moisture in the substrate at the time of the original spray.

Correct repair is straightforward but not optional. Every blister gets cut open and inspected, not simply coated over. If the foam beneath is dry and sound and only the coating had released, we clean, prime, and rebuild the coating locally. If the foam is wet, delaminated, or crumbling, that section comes out entirely and gets re-sprayed with fresh closed-cell foam shaped to match the surrounding drainage plane. Coating over an intact blister is the single most common shortcut we find on Phoenix roofs, and it always fails, usually within one summer.

The Physics Of A Blister On A 165 Degree Deck

A blister is a pressure vessel, and Phoenix supplies the pressure for free. Any trapped pocket of air or water vapor expands as it heats, and a white roof surface reaching 160 to 170 degrees on a July afternoon drives a substantial pressure differential against whatever is holding that pocket in. Water is the worse case by an order of magnitude, because a small amount of liquid converting toward vapor generates far more volume than heated air alone. The weakest thing above the pocket is the elastomeric film, so the film is what lifts.

What kills the coating is not one hot day, it is roughly a hundred and eighty of them in a row. The blister inflates every afternoon and deflates every night, and the film at the perimeter of the blister flexes through that cycle thousands of times. Elastomeric coatings tolerate movement, but they have a fatigue limit, and the edge of a blister concentrates strain in a narrow ring. Eventually the film work-hardens along that ring and splits, which converts a sealed pressure pocket into an open funnel pointed directly at the foam.

Three Depths Of Separation And Three Different Repairs

Every blister has to be opened before anyone can say what it is, and what it turns out to be dictates a completely different scope of work. Diagnosing from the surface is guesswork, because a shallow coating blister and a serious substrate delamination can look identical from a standing position. We cut a cross into each one, fold the flaps back, and read what is underneath. The distinction is not academic. It is the difference between an hour of work and cutting several square feet of roof out to the deck.

Coating-Level Separation

The shallowest case is the film releasing from the foam with sound, dry foam underneath. This almost always traces back to application conditions: coating applied over a damp or dusty surface, over chalked film that was never washed or primed, or sprayed on an afternoon hot enough that the surface skinned over before the material below it could release its water. The repair is local. Cut the loose film back to well-bonded coating, clean and prime the exposed foam, then rebuild the coating with base coat, embedded polyester fabric across the transition, and top coat carried past the edges.

Interlaminar Separation Between Foam Lifts

The middle case is separation between two passes of foam. This is an original installation defect, usually a lift applied too thick in one pass, sprayed onto a previous pass that had cooled past its recoat window, or applied over a surface contaminated with dust or overspray. The upper foam is not bonded to the lower foam and will keep releasing outward from the blister as thermal cycling continues. Repair means removing the delaminated upper lift across its full extent, not just the blistered spot, and re-spraying fresh foam in controlled lifts.

Substrate Delamination At The Deck

The most serious case is the foam letting go of the deck itself. This normally means the substrate held moisture at the time of the original spray, or was dusty, oily, or improperly prepared. The consequence is that the roof assembly is no longer mechanically continuous, which matters for wind uplift resistance as much as for water. These areas get cut out to the deck, the deck is dried and prepared correctly, and new foam is sprayed with proper adhesion. If substrate delamination is widespread, the conversation changes to partial or full replacement.

Surveying With Core Cuts And Infrared Before Anything Is Opened

Visible blisters are usually not the whole population. Before opening anything we survey the roof to find the areas that are wet or separated but have not lifted yet, because repairing only the obvious ones guarantees a return visit next summer. An infrared survey run in the evening, once the deck has been heated all day and is releasing that heat, shows wet foam as warmer areas because saturated material holds heat longer than dry material. Conditions have to cooperate, so this is scheduled work rather than something done on any random afternoon.

• Soft or spongy spots that give underfoot without any visible dome in the coating
• Hollow sound when a blunt probe is tapped across the field in a grid pattern
• Warm signatures on an evening infrared scan that do not match a known equipment source
• Coating dome edges showing a fine crack ring, which means the film is already fatiguing
• Fluid movement inside a dome when pressed, indicating liquid water rather than trapped air
• Repeated elevated meter readings in the same location across two or more inspection visits

Infrared shows where to look, and core cuts confirm what is there. A core cut is a small plug taken through coating and foam to the deck, which lets us read the actual layer sequence, measure foam thickness, check adhesion at each interface, and feel whether the material is dry. Cores are taken in both suspect and known-good areas so there is a baseline to compare against, and every core is properly re-foamed and coated the same day. Two or three cores usually settle questions that would otherwise be argued about for weeks.

Repair Geometry: Bevels, Feathering, And The Drainage Plane

A blister repair that leaves a bump or a depression has traded one problem for another, because either result changes how water crosses that part of the roof. Cut edges are beveled outward rather than left square, which gives the replacement foam a mechanical key and lets the finished surface be shaved flush with the surrounding contour. Replacement foam goes in slightly proud and is then trimmed down to match the existing drainage plane, checked by eye along a low sight line and, where the area matters, by running water across it before any coating goes down.

1. Mark every blister found by survey, including soft or hollow areas that have not yet lifted
2. Cut each one open and diagnose the depth of separation before deciding on scope
3. Remove all damp, delaminated, or friable material out to a dry, well-bonded boundary
4. Bevel the cut edges and dry the exposed substrate completely before any new material goes in
5. Shave the patch flush with the drainage plane and confirm water crosses it without pooling
6. Prime as required, then rebuild base coat, embedded fabric, and top coat to full thickness past the edges

Application Conditions That Prevent The Next Generation

Most blisters are avoidable and the avoidance happens during application. Substrate has to be dry, which in practice means checking rather than assuming, since foam sprayed onto a deck that looks dry but holds residual moisture will delaminate later regardless of how good the crew is. Lift thickness has to stay within the range the material is designed for, because thick passes trap heat and produce weak, friable cores. Recoat windows between passes have to be respected, since foam that has fully cooled and skinned needs preparation before the next lift will bond.

Coating application carries its own rules. Surface temperature, ambient temperature, and dew point all have to be inside the acceptable window, and in this climate that usually means starting early and stopping before the deck gets too hot rather than pushing through the middle of a July afternoon. Coating applied to a surface above the material limit skins instantly and traps solvent or water underneath, which is a blister waiting for the next hot day. Crews that stop at ten in the morning in August are not being lazy, they are avoiding exactly this failure.

When Blister Count Argues For Full Restoration

There is a threshold where individual repairs stop being economical. Each blister carries a fixed cost in setup, cutting, diagnosis, foam, and coating detail regardless of how small it is, so twenty small blisters cost far more than one blister twenty times the size. When the count climbs into the dozens, or when core cuts show that separation is a systemic condition across the whole deck rather than a set of isolated events, the right answer is usually to address the entire roof at once under a single mobilization, preparation, and coating cycle.

The deciding question is what the survey says about distribution. Blisters concentrated in one section, near a specific detail, or along a single spray pass generally indicate a bounded original defect that can be cut out and rebuilt. Blisters scattered evenly across every elevation usually mean the entire application shared the same problem, and no number of patches will change that. A survey with infrared and cores costs a small fraction of the wrong decision, which is why we run one before quoting anything larger than a handful of repairs.