Fence Repair vs Replace Field Assessment Standard

Purpose

This SOP guarantees that the walk a technician runs before recommending repair or replacement produces a counted, defensible finding, not an impression formed from the section the customer pointed at. It is what makes the thresholds in Fence Repair vs Replace Decision honest: a percent-intact figure built from eyeballing the fence at the gate is not the number a probe at every post produces, and that gap is where a shop patches a run that needed replacing, or quotes a full replacement over one bad hinge. Skipped or rushed, this walk produces the two failure modes the decision reference already names, the upsell that costs trust and the undersell that costs a callback, because both start from a count that was never taken.

Scope

Covers the on-site assessment of an existing fence run being evaluated for repair versus replacement, from the initial full-property walk through a counted report ready for the recommendation conversation. It does not cover the criteria that turn a count into a recommendation; Fence Repair vs Replace Decision owns the age and percent-intact bands and the customer conversation. It does not cover the first-response walk after a storm, where life-safety hazards come first; Storm Damage Fence Repair Triage Standard owns that sequence, and this procedure picks up once any live hazard from it is cleared. It does not cover the utility locate itself (Fence Line Utility Locate and Hand Dig Standard, gating step 5) or repairing any single finding (the procedures routed in step 8).

Roles and responsibilities

Role Owns Hands off
Technician Walks the full run, probes and levels every post, checks hardware, photographs and counts every finding A complete report with a counted numerator and denominator before any recommendation is voiced
Crew lead Confirms whether a current locate record is on file for this visit, resolves any section the technician can't safely assess A go or no-go on same-visit hidden probing, with a return date where it's a no-go
Shop Runs the recommendation conversation against Fence Repair vs Replace Decision, files the report against the property A written record the customer can be shown if the recommendation is later disputed

Procedure

  1. Walk the entire fence line, not just the section the customer described, and confirm the post and section count against the property's fence record or a physical count on site. Acceptance: every post and section tallied in writing before probing starts. Wrong looks like assessing only the reported section and treating that as the assessment, when a run commonly carries a second problem on a section the customer rarely looks at. Stop rule: a tally that doesn't match the known post count halts the walk until resolved. Hazard: a full-property walk can cross an unclear property line or run through vegetation hiding debris or uneven grade; stay on the confirmed side and route around anything blocking a clear sightline.

  2. Probe every post at and just below grade with an awl on two faces, not only the posts already showing visible lean or damage. Acceptance: the point sinks less than about a quarter inch on both faces, the soundness figure Leaning Post Reset Standard Procedure and the shop's method matrices use, logged pass or fail per post. Wrong looks like probing the three or four obviously bad posts and extrapolating the rest as fine, when punky wood under paint or stain only shows on the probe. Stop rule: any post not individually probed doesn't get logged as sound; a percent-intact figure waits until the rest of the run is probed. Hazard: kneeling at grade along an older run puts hands near broken wire ends, old staples, and exposed nail heads; gloves on, and check the area by hand before kneeling, not after.

  3. Check plumb on every post individually with a level against two adjacent faces, never by eye against the neighboring posts. Acceptance: bubble centered within its lines on both faces for a pass, bubble off its lines on either face logged as a lean finding. Wrong looks like calling a post fine because it matches the lean of the post beside it, which measures the neighbor, not the post. Stop rule: a post checked only against its neighbor doesn't get logged as plumb; go back with the level before the count is finalized. Hazard: reading the upper face of a tall privacy post may call for a ladder; foot it on level ground and never brace it against the post being assessed, since that post's soundness hasn't been confirmed yet.

  4. Check every gate, hinge, latch, and panel splice point for movement, corrosion, and elongated fastener holes by hand, not only the hardware at the reported section. Acceptance: each hardware point tested for play by hand, pass or fail logged, corrosion or elongation noted. Wrong looks like checking hardware only at the gate the customer complained about and skipping a corroding hinge two sections over, which becomes the next callback. Stop rule: any hardware finding gets logged even where it wasn't the reported issue, distinguished from what was originally called in. Hazard: a gate or panel with play can swing once released during the check; keep clear of its arc rather than standing in the path it would take if it moved.

  5. Where the run's age is at or past the shop's expected-life band (PT pine 15 years, cedar 20, vinyl 30, the bands Fence Repair vs Replace Decision uses) or the property carries three or more repairs in the trailing two years, probe below surface depth at a post that passed step 2; a post can read sound at the surface and still fail six to eight inches down, the gap Leaning Post Reset Standard Procedure found mid-repair. Acceptance: a current locate on file for this visit, probe taken, hole backfilled and tamped before the visit ends. Wrong looks like skipping the probe because every visible post passed step 2, when the trigger exists because a surface pass on an aged run isn't reliable alone. Stop rule: no current locate on file halts any probe past surface depth; log it deferred to a return visit under the Fence Line Utility Locate and Hand Dig Standard rather than digging without one. Hazard: same sharp-metal exposure as step 2, plus a buried-line strike risk; gloves on, area cleared by hand first, no probe without a current locate.

  6. Photograph every finding against a fixed reference, a post number, a tape measure, or a chalk mark, and log a numerator of failing units against the denominator counted in step 1. Acceptance: a photo tied to an identifier for every finding, numerator and denominator recorded as counts, not an estimated percentage. Wrong looks like writing "about a third is failing" from a walking impression, which the decision reference's bands can't apply to honestly without a counted fraction. Stop rule: a report without a counted denominator doesn't get a percent-intact figure attached; raw counts go on the report until the full run is counted. Hazard: none new, documentation only.

  7. Record the apparent cause for each finding individually, rot, heave, impact, wear, insect, or hardware fatigue, rather than one dominant cause across the whole run. Acceptance: a cause recorded per finding, checked against the pattern of any single known event on the property. Wrong looks like attributing every failure to the storm event the customer called about, missing a second independent cause on an unrelated section. Stop rule: any finding whose cause doesn't fit the known dominant event gets flagged for separate review rather than folded under it. Hazard: on posts predating roughly 2003, cause assessment may raise chromated-copper-arsenate treated lumber; do not sand, scrape, or grind a suspect post to inspect its core, since CCA dust is an inhalation and skin hazard. Note the suspected treatment era and route any confirmation cut to a controlled method instead.

  8. Compile the counted report, denominator, numerator, per-finding cause, hardware findings, and the hidden-probe result or its deferred status, and hand it to whoever runs the recommendation conversation before any repair-or-replace call is voiced. Acceptance: a complete report with every prior step's findings attached exists before the conversation starts. Wrong looks like telling the customer "this should just need a patch" mid-walk, before the hardware check or hidden probe is done, then walking it back once the rest of the count comes in worse. Stop rule: no recommendation is voiced until this report is complete; a partial walk isn't treated as sufficient because the visible damage looked minor. Hazard: none new, documentation and handoff.

The record this produces

The assessment report ties the property and date to a per-post probe and plumb result, per-hardware-point findings, the hidden-probe result or its deferred status with a return date, per-finding cause, and a counted numerator over denominator. Fence Repair vs Replace Decision's percent-intact bands apply against that denominator rather than an estimate, giving the shop a written basis for a disputed recommendation, or to revisit if a deferred probe changes the count.

Worked pass, with a step that failed

A wood privacy run behind a house, 24 posts total per the property's fence record, customer called about "a rough-looking section" near the patio.

Step 1: full walk confirms 24 posts, matching the property record. The reported section is one of six on the shaded side of the yard.

Step 2: every post probed at grade on two faces. Three posts fail the quarter-inch standard, two in the reported section, one on the far side the customer hadn't mentioned. Twenty-one posts pass.

Step 3: every post leveled individually. Two further posts read off plumb, neither among the three that failed the probe; one sits against a large shrub whose root growth is pushing the footing sideways.

Step 4: hardware checked at the gate and both splice points. The gate's lower hinge shows play and surface corrosion; the splice points pass.

Step 5: the run is 16 years old, PT pine, past the 15-year trigger, so a hidden probe is required at a representative post that passed step 2. Only the original install-era locate is on file, not a current one for this visit. Fails the acceptance condition, which requires a current record before any probe past surface depth. Stop rule taken: the probe is not run; it's logged as deferred and routed to a return visit under the Fence Line Utility Locate and Hand Dig Standard.

Step 6: photographs taken against post numbers for all five flagged posts and the hardware point. Denominator 24, numerator 5 (3 rot, 2 lean), the hardware finding reported separately. Nineteen of 24 posts pass clean, about 79 percent post-level intact.

Step 7: causes recorded individually. The three rot failures share one cause, ground-contact deterioration consistent with the run's age. The two lean findings don't share that cause; one is root pressure from the shrub, the other shows no adjacent root growth and is logged undetermined pending a closer look. The hinge corrosion is logged separately as fastener fatigue.

Step 8: report compiled with the 79 percent figure, three distinct causes, the hardware finding, and the deferred hidden probe flagged with its return date, handed to the shop before any recommendation is given. The report notes the figure as a surface-level floor, not a final count, until the deferred probe runs.

References

  • See related: Fence Repair vs Replace Decision, for the percent-intact and age-band thresholds this procedure's counted report feeds.
  • See related: Storm Damage Fence Repair Triage Standard, for the first-response sequence that precedes this procedure after a storm or wind event.
  • See related: Fence Line Utility Locate and Hand Dig Standard, for the locate sequence gating any probe past surface depth in step 5.
  • See related: Leaning Post Reset Standard Procedure, for the buried-section soundness finding this procedure's hidden-probe step is built on.
  • AWPA U1, Use Category System: User Specification for Treated Wood, for ground-contact treatment expectations used when probing buried wood.