Nailing Pattern and Fastener Audit Before Close

Purpose

This procedure catches fastener defects while the course above is still off, at a stated sample rate, so a bad gun setting costs a square of rework instead of a roof. Nailing is the only installer-controlled variable that decides whether a shingle stays on in wind, and it is completely invisible the moment the next course covers it.

It is also the defect a manufacturer looks for first. A shingle carries a wind rating tested to ASTM D3161 or ASTM D7158, and that rating only describes the shingle as fastened the way the instruction says. A wind claim on a roof with high nails does not get denied on a technicality; it gets denied because the assembly tested was not the assembly installed.

Scope

Covers fastener count, placement, depth and penetration for asphalt shingles, hip and ridge, starter and rake terminations on residential and small commercial steep-slope roofs, during the covering run and at plane close.

Does not cover deck fastening, owned by the deck assessment SOP. Does not cover underlayment cap fasteners or flashing fastening, owned by the underlayment and flashing verification SOP. Does not cover metal panel clips and screws, owned by the metal roof installation SOP. Does not cover the tool's own maintenance schedule.

Roles and responsibilities

The two gates, and the unit each is measured on

There are two, they measure different things, and mixing them is how a crew argues itself past a real problem.

Per shingle. Every nail present, every one inside the manufacturer's nail zone, every head flush and seated, and penetration at least 3/4 in into the deck or fully through decking thinner than 3/4 in. One defect on one shingle is corrected on that shingle, and it is not evidence of anything by itself.

Per plane. The share of audited shingles carrying at least one defect. The default gate is 5 percent of the shingles audited on that plane, and it exists because a rate is what tells you the tool or the operator is wrong rather than the shingle. At or under 5 percent, correct and keep going. Above 5 percent, the gun crew stops. Tune the number to your own crews after a month of data, but commit to one, because "a few high nails" is not a gate.

Procedure

1. Read the fastening spec off the instruction in the bundle wrapper, before the first bundle opens. Write down nails per shingle, the nail zone description, required shank and head, and whether the local wind requirement or the enhanced wind warranty raises the count. Acceptance: nails per shingle, nail zone and minimum penetration written on the ticket, matching the product actually on the truck. Wrong looks like a crew defaulting to 4 nails on a roof the jurisdiction requires 6 on. Stop rule: no instruction on site means the office sends it before the first course. Hazard: none, this is done at the truck.

2. Prove each gun on a scrap before it touches the roof, and record the setting. Set regulator pressure and the depth adjustment, drive into a scrap of the same deck material at the same thickness, and check the head is flush without cutting the mat. Acceptance: heads seated flush, mat not cut, shank fully through into the scrap, and the pressure written next to the gun's identifier. Wrong looks like a setting proved on a warm scrap in the sun that overdrives on cold shingles an hour later. Stop rule: a gun that cannot hold a flush head across five test drives comes off the roof for service. Hazard: the tool is live from here on, so it is carried by the handle with the trigger clear, the muzzle safety stays fitted per 29 CFR 1926.302(b), and a gasoline compressor sits outdoors and well away from any door, window or soffit vent because its exhaust is carbon monoxide.

3. Set the sample and name the auditor before the run starts. Default sample is 4 shingles per square, chosen across the plane rather than from one area, plus the first full course each gun lays after any break. The auditor is not the person who laid it. Acceptance: sample rate, auditor's name and the planes they own written before the first bundle is laid. Wrong looks like the operator auditing their own courses, which reproduces the same misread that caused the defect. Stop rule: nobody free to audit means the crew audits at the end of each square instead of continuously, never not at all. Hazard: none at this step, it is a decision made at the huddle.

4. Audit the starter and the first course on each plane before anything builds on it. Lift the first course, check fastener count, placement relative to the nail zone, and that the starter is fastened at the manufacturer's stated distance from the eave edge. Acceptance: full count present, every fastener in the zone, starter fasteners where the instruction places them. Wrong looks like starter fasteners set so low they show below the first course, or so high they miss the sealant line. Stop rule: a wrong first course is pulled and re-run rather than continued over, because every course above inherits its offset. Hazard: this is eave work, so it is done connected, and any offcut or pulled nail goes in the pouch rather than on the plane where it rolls under a foot.

5. Audit at the sample rate through the run by lifting shingles, not by reading the surface. Lift the tab and look at the fasteners in the course below, checking count, position in the zone, head seating, angle and any nail placed above the zone. Acceptance: every audited shingle either clean or logged with its defect type, running tally kept per plane. Wrong looks like an auditor who walks the plane looking for proud heads, which finds only one of the five defect types and misses shiners entirely. Stop rule: the plane gate at 5 percent of audited shingles stops the gun crew on that plane. Hazard: lifting a sealed shingle in heat tears the mat and the roof surface itself runs far hotter than the air, so audit the sealed area early or in shade with gloves, and never pry a cold shingle, which cracks rather than lifts.

6. Correct on the spot, and re-seal anything you opened. An underdriven head is driven flush by hand. A shiner or a nail outside the zone gets an added correct fastener, the offending nail sealed, and the course above checked. Any shingle lifted to audit is re-sealed with the cement the shingle manufacturer names, in the amount and location the instruction gives. Acceptance: no lifted shingle left unsealed anywhere on the plane, confirmed by the same auditor before the plane closes. Wrong looks like a generous smear of roof cement across a whole tab, which traps water and voids the instruction it was meant to satisfy. Stop rule: a shingle whose mat tore during lifting is replaced, not sealed. Hazard: asphalt roof cement carries solvent, so nitrile gloves and eye protection, no use in an enclosed space, and the SDS for the specific product read before opening.

7. Re-prove the tool after any change, before the next course. Hose swap, compressor swap, tank change, a new operator, or a temperature swing across the day all move the effective driving depth. Acceptance: a fresh scrap test with a flush head after every one of those events, recorded with a time. Wrong looks like a crew that proves the gun once at 7 am and works until 4 pm through a 30 degree rise in surface temperature. Stop rule: no scrap test after a change means the courses laid since the change are audited at 100 percent before the crew moves on. Hazard: this is the step that puts the tool back into service, so the muzzle safety is confirmed fitted and the first drive goes into scrap with the tool pointed away from every person on the plane, not into the roof.

8. Audit terminations, then close the plane. Check hip and ridge cap fastener length against the extra layers at the ridge, check rake and end-of-course fasteners, and confirm penetration where the deck is thin or where cap sits over a vent. Acceptance: cap fasteners long enough to make the same 3/4 in penetration through the stacked material, none exposed below the next cap, and the plane's audited count and defect rate written down. Wrong looks like field-length nails used on ridge cap over a ridge vent, which do not reach the deck at all. Stop rule: any cap fastener not reaching the deck means the ridge run is redone with the correct length. Hazard: the ridge is the highest and most exposed line on the job, worked from a connected position, and a coil of hose dragged over a ridge pulls the operator toward the far side.

The record this produces

One fastener entry per plane: date, plane identifier, product and its wind rating, nails per shingle required and the reason for that number, gun identifiers with proved pressures and every re-prove time, shingles audited, defects by type, the calculated defect rate, whether the plane gate tripped and what happened, corrections made, and the auditor's name.

The lead writes it at plane close. The warranty desk reads it when a manufacturer asks how the roof was fastened, which is the question a wind claim opens with. The office reads the defect rate over a month to see whether one crew or one tool keeps appearing.

Worked pass: 22-square architectural reroof, two guns

The roof is two planes, 12 squares north and 10 south, 22 total. Sample rate 4 shingles per square, so 48 audited north and 40 south, 88 in all. Step 1 recorded 6 nails per shingle for the jurisdiction's wind requirement, the nail zone as printed, and 3/4 in minimum penetration into 7/16 in OSB, which means fully through.

The north plane fails its gate. Of 48 audited shingles, 5 carry a defect. 5 divided by 48 is 10.4 percent, against the 5 percent plane gate, so the stop rule fires and the gun crew comes off. Four of the 5 were underdriven heads standing roughly 1/16 in proud, all on courses laid by gun 2. The fifth was a single nail above the zone on a course laid by gun 1, corrected on the spot under step 6 and not treated as a pattern.

Cause: gun 2 changed hoses at 10:40 am and was not re-proved, which is step 7's trigger. Under step 7's stop rule the courses laid since that change get a 100 percent audit, which covered the 2 squares gun 2 had laid after 10:40. That audit found 14 more underdriven fasteners, all re-driven by hand, and every audited shingle re-sealed under step 6.

Gun 2 was re-proved on scrap at 11:35 with the regulator dropped and the depth adjusted, flush across five test drives, and the time recorded. The south plane then ran clean: 40 audited, 1 defect, which is 2.5 percent and under the gate.

Roof total: 88 audited by sample plus the 100 percent re-audit of 2 squares. Defect rate on the sample was 5 north plus 1 south, or 6 of 88, which is 6.8 percent for the roof. The honest reading is that the roof-wide number is not the useful one. The north plane's 10.4 percent is what fired the stop, and averaging it with a clean plane would have hidden the tool problem entirely, which is exactly why the gate is written per plane.

References

• The shingle manufacturer's installation instructions, which set nails per shingle, the nail zone, fastener specification and penetration, and are the condition of the wind warranty
• ASTM D3161 and ASTM D7158 for asphalt shingle wind resistance classification, which describe the shingle as fastened to instruction
• International Residential Code, Chapter 9, in the edition your jurisdiction has adopted, for asphalt shingle fastener requirements and any local high-wind amendment
• 29 CFR 1926.302(b) for pneumatic power tools in construction work, including the muzzle safety device and hose connection requirements
• See related: the deck assessment SOP, which establishes the deck thickness this procedure's penetration figure depends on