Opening Preparation and Rot Assessment

Purpose

The new window inherits the opening. Set a sound unit on a soft sill and the fasteners pull, the reveal opens up over the first heating season, and the customer calls it a bad window. Nothing in the unit caused it, and nothing in the unit will fix it.

This procedure guarantees two outcomes. The opening is proven sound by test rather than by eye before the unit is committed to it, and where it is not sound the extent is mapped and the repair is agreed with the customer standing at the opening, not described to them on the phone that night. The second half is what keeps the money conversation honest: a change order signed at the opening with the customer looking into the wall cavity is a different conversation from an invoice line they first meet a week later.

Scope

Covers the window from the moment the old unit is out to the moment the opening is ready for flashing, on insert and full-frame replacement.

Does not cover: removal itself, and in pre-1978 housing every disturbance in this procedure is still governed by the lead-safe removal standard, which does not end when the sash comes out; flashing and pan installation, which begins where this ends; the insert versus full-frame method call, which the jamb and sill condition matrices own; and any repair that changes a header, a king stud or a bearing condition, which stops being window work and becomes a framing scope with its own permit question.

Roles and responsibilities

The handoff that fails is the first row. An installer who finds soft wood, decides it is minor, and proceeds has removed everyone else from a decision that was theirs.

Procedure

1. Photograph the opening before anything is touched. Four corners plus a full frontal, with the old fasteners, sealant and any staining still in place. Acceptance: five images in the job file, timestamped, before the first tool touches the opening. Wrong looks like photographs taken after cleanup, which no longer show the water path. Stop rule: if the light will not give a usable image of a stain, add a work light and re-shoot rather than proceeding, because that stain is the only evidence of where water was entering and it disappears the moment you scrape.

2. Strip the opening back to substrate and leave nothing proud. Remove old fasteners, shims, hardened sealant and any loose material, then dry-sweep or vacuum. Acceptance: no fastener or shim standing above the framing face anywhere the new unit or the pan will bear, and the water-resistive barrier visible where it survives. Wrong looks like a clipped nail left flush, which telegraphs through a sill pan and becomes a puncture. Stop rule: a fastener that will not come out gets driven below the surface and the location noted, not left for the pan to find. Eye protection for the pry and the sweep, and where the sill is masonry or stucco, cutting or grinding it releases respirable crystalline silica, so use a tool with integrated water delivery or a shrouded dust collector under 29 CFR 1926.1153 rather than a dry blade.

3. Probe test at the points that fail first, in that order. Push a sharp awl into the wood at hand pressure at the two sill ends first, then across the sill at 6 in intervals, then the bottom 12 in of each jamb, then the head. Acceptance at every point: the awl stops within about 1/8 in and, when levered, lifts a splinter that breaks with a snap. Wrong looks like the awl sinking 1/4 in or more and the wood lifting as a soft spongy fiber that pulls rather than snaps, which is decay, not weathering. Stop rule at the first failing point: stop probing for a pass and start mapping the extent at step 5. Do not drive the awl toward your other hand, and do not probe blind into a cavity you have not looked into.

4. Take a moisture reading with a pin meter at the same points. Set the meter to the species where it offers a correction, drive the pins across the grain, and record wood moisture content at each probe point. Acceptance: at or under 19 percent, which is the conventional dry-lumber line. Wrong looks like a sustained reading at or over 20 percent, which is where decay fungi become active and stay active. Stop rule: a reading at or over 20 percent in wood the awl says is still sound is not a pass with a note, it is an active leak with the decay not yet visible, so the water source is traced and named before the opening is closed.

5. Map the extent outward until two consecutive sound points confirm it. From the failing point, probe outward at 6 in intervals until two consecutive points both meet the step 3 acceptance, and set the cut line at the second of them. Record the extent as inches from a named datum, such as the inside face of the left jamb. Acceptance: a written extent with a datum and two confirming probe points. Wrong looks like cutting back to where the wood "looks better," which reliably leaves decayed fiber under the new work. Stop rule: if the extent runs past the rough opening into a stud bay, a band joist or a sheathing area, stop cutting and escalate; that is a framing scope. Cutting the sill means a reciprocating saw against embedded fasteners, so cut with a demolition blade rated for nail embedment, keep the shoe against the work to control kickback, and expect the blade to grab.

6. Repair, then re-verify by the same two tests. Replace the removed material with treated or preservative-treated stock, fasten to sound framing beyond the cut line, and re-run the awl and the meter at the repair and at both sides of each joint. Acceptance: every re-tested point passes both step 3 and step 4. Wrong looks like a repair verified visually. Stop rule: a joint that still reads at or over 20 percent has an unaddressed water path behind it and the opening does not close. Cutting or sanding preservative-treated stock releases treatment-bearing dust, so cut it outdoors, wear eye and respiratory protection appropriate to the treatment named on the stock's tag, and do not burn the offcuts.

7. Check the prepared opening for square, level and plane before the unit is offered up. Take both diagonals, level the sill across its width, and check the framing plane with a straightedge at the head and both jambs. Acceptance: diagonals within 1/4 in of each other, sill level within 1/8 in across the opening width, and the framing faces flat enough that a straightedge shows no gap over about 1/8 in. Wrong looks like an opening that measures square and has a jamb bowed into the opening at mid-height, which no amount of shimming corrects. Stop rule: an out-of-plane jamb is planed or shimmed to plane before the pan goes in, because after flashing it is not reachable.

8. Dry-fit the unit, then take it back out. Set the unit into the prepared opening with no fasteners, no sealant and no pan, and check the gap all around. Acceptance: a gap at head, sill and both jambs inside the manufacturer's published shim range, with no side at zero. Wrong looks like a unit that goes in with a shove, which means a side is at zero and there is nowhere for the shims or for movement. Stop rule: a zero-gap side goes back to step 7, not to a bigger hammer. Two people on the dry-fit and the removal both, since the unit is being handled twice with no fasteners in it.

The record this produces

The opening condition record, one per opening, filed to the job with its photographs, plus any change order.

Fields: the five pre-strip photographs, the probe result at every named point, the moisture reading at every named point, the mapped extent with its datum and the two confirming points, the repair method and materials, the re-verification readings, the diagonals, sill level and plane result, the dry-fit gap at all four sides, and the change order number with the signature time if one was raised.

Who reads them later: the warranty desk reads the moisture readings when a customer reports a stain two years on, because a documented 14 percent at install moves the question from "was it wet when you closed it" to "what changed." The estimator reads the mapped extents across a season and learns which elevation and which era of housing in the territory carries hidden sill decay, which is what turns a surprise change order into a line that was already in the bid. The project lead reads the signature time when a customer disputes authorization.

Worked pass: second-floor bedroom, 34 in by 62 in rough opening

Old unit out, five photographs taken before anything moved. Strip found three roofing nails left proud in the sill from a prior repair; two came out, one broke and was driven below the surface with its location noted for the pan.

Step 3 failed at the first point. The awl went 3/8 in into the left end of the sill at hand pressure and lifted a soft fiber that pulled rather than snapped. Probing stopped there and moved to step 5, per the stop rule.

Step 4 at the same point read 31 percent moisture content against a 19 percent acceptance. The right end of the sill read 14 percent and passed both tests, which is what told the crew this was a localized entry rather than a wet wall.

Step 5 mapped it. From the left jamb: 7 in soft, 13 in sound, 19 in sound. Two consecutive sound points, so the cut line went at 19 in from the inside face of the left jamb, replacing 19 in of a 34 in sill, about 56 percent of it. The extent, the datum and both confirming points went in the record.

Step 4's second finding drove the real fix. High moisture at one end with sound wood at the other put the crew looking for a source, and stripping the exterior casing showed no sill pan under the original unit at all, with the sill sloping perceptibly toward the left corner. Every driving rain that got past the exterior sealant bead had landed on bare wood and run to the low end. Replacing the sill without that finding would have put new wood in the same water.

Step 7 came back inside tolerance after the repair: diagonals 70 5/8 and 70 3/4, a difference of 1/8 in against a 1/4 in acceptance, sill level within 1/16 in across the width, and no straightedge gap over 1/16 in at either jamb.

The customer signed the change order at the opening, looking into the cavity, before any cutting started. The sill repair added about 2.5 hours to a 3.0 hour planned install, so that opening ran 5.5 hours, roughly 1.8 times the planned time. That number is worth carrying: on a house of this era the shop now schedules one sill repair per eight openings by default, and an opening that comes in clean gives the crew time back rather than an opening that comes in rotten taking time the day never had.

References

• ASTM E2112, Standard Practice for Installation of Exterior Windows, Doors and Skylights, in the edition your supplier's installation instructions reference, for opening tolerance and preparation.
• 29 CFR 1926.1153, OSHA Respirable Crystalline Silica in Construction, including the Table 1 controls for handheld saws and grinders on masonry and stucco.
• Forest Products Laboratory, Wood Handbook, for the relationship between wood moisture content and fungal decay activity.
• See related: the pre-1978 lead-safe removal SOP, which governs every disturbance in this procedure in older housing; the flashing and water-management detail, which begins where this ends; the jamb and sill condition method matrices.