Readiness check

Is your single-ply seam and attachment QA ready?

On a single-ply roof the field of the sheet almost never leaks. The seams, the flashings, the penetrations, and the fastener rows do, so seam and attachment QA is the roof QA. Skip the trial weld and you run all day on settings nobody confirmed; skip the probe or the deck pull test and a cold weld or a spinning fastener stays hidden until the rain or the first big wind finds it, and the manufacturer's rep can hold the warranty over it.

1. Do you have the membrane manufacturer's specification on site, with the seam method, weld settings, and weld-width minimum for this system?Yes, the current product spec and warranty are on the roof, and the crew knows the membrane type, weld temp/speed range, and the weld-width minimum before welding startsWe know the membrane type but are working from memory or a generic setting, not the manufacturer's published figures for this systemNo spec on site; the welder is set by habit and the acceptance criteria are assumed

2. Are you pulling a trial weld on scrap at the start of each day and at every shift or condition change, and peeling it to film-tearing bond?Yes, a fresh trial weld is cut and hand-peeled to film-tearing bond before any production seam, and again on new rolls, new operators, or weather changesWe pull one in the morning but do not re-pull when the temperature, wind, or sun shifts through the dayNo trial weld; the welder runs on yesterday's or a set-once setting

3. Is the membrane clean and dry before welding, with aged or soiled laps cleaned with the manufacturer's cleaner and wiped dry?Yes, every lap is confirmed clean and dry; aged, chalky, or soiled membrane is cleaned with the manufacturer's cleaner and wiped dry, and welding stops on dew or damp surfacesWe wipe visible dirt but do not consistently use the manufacturer's cleaner or check for moisture before the weldWe weld over the membrane as it lies, dirty, chalky, or damp

4. Are you probing every lineal foot of seam with a blunt tool after the welds have cooled, and marking every catch?Yes, every foot of seam is probed with a blunt, rounded tool after the weld has fully cooled (commonly at least 20 minutes), and every catch is marked for repairWe probe only a sample of the seams, or probe while seams are still warmWe do not probe, or use a sharp tool that scores the weld

5. Are you checking finished weld width and the squeeze-out bead against the spec, and confirming with destructive peel samples?Yes, we measure finished weld width against the minimum, read the squeeze-out bead (consistent, not absent, not scorched), and cut destructive peel samples for film-tearing bond at the manufacturer's intervalWe eyeball the weld and bead but do not measure width or cut destructive samplesWe do neither; a closed-looking seam is called good

6. Is every T-joint getting the manufacturer's required treatment, confirmed present and bonded?Yes, every seam crossing gets the required T-joint treatment (welded target/skive on TPO-PVC, primed and rolled patch on EPDM), and each one is probed and confirmed bondedWe treat some T-joints but do not confirm each one, or treat them the same as a straight runT-joints are left as running seam with no dedicated treatment

7. On a fastened system, did you pull-test fasteners in the actual deck and set the fastening pattern by zone, tighter at perimeter and corners?Yes, fasteners were pull-tested in the real deck (and re-tested on wet or aged decks), the count backs out with the safety factor, and field, perimeter, and corner zones carry their own pattern per the listed assemblyWe used a catalog pullout or one test reading, or carried close to a uniform pattern with light perimeter and corner enhancementNo deck pull test, and the field pattern runs all the way to the edge

8. Is the seam and attachment log being completed as you go, with the punch list closed before the manufacturer's inspection?Yes, we log by area/zone (weld width, trial-weld and probe results, repairs, destructive samples, deck pull-test values, pattern and uplift rating) and close every flagged spot with welds and patches before the rep arrivesWe keep partial notes but the log is incomplete, or repairs are marked but not reprobed before signoffNo log; we wait for the manufacturer's rep to find the problems