How to Document a Condition You Found Before You Touch It

Why this matters

A year after a job, nobody remembers anything. The technician has moved on, the customer remembers a version of events in which they were told nothing, and the only thing still standing is whatever got written down at the time. A record made before the condition was altered is your evidence that the condition pre-existed your work. What such a record is worth in a dispute depends on the forum and on how it was made and kept, which a checklist does not settle. A record made after the repair proves only that you repaired something, which is precisely the fact nobody is disputing.

This is not about paperwork discipline in the abstract. It is about the single most common way a shop ends up paying for damage it did not cause: it fixed the evidence.

What outranks the record

If the condition you found is a hazard, the hazard action comes first and the camera waits.

• Gas odor or a suspected fuel leak: everyone leaves the building immediately, nobody touches a switch, nobody turns a light on or off, no phone is used inside, and the call to the gas utility and to the customer is made from outside and upwind. Any documenting happens later, from outside, after the utility has released the building. A camera is an ignition source like any other.
• Exposed or damaged energized conductors: keep everyone clear at a distance, barrier or guard the area so nobody walks into it, and do not open, probe or photograph inside the enclosure while it is energized. If the condition has to be documented inside the enclosure, de-energize at the source, apply your lock and tag, and prove the circuit dead with a meter rated for it, checked on a known live source before and after the test, which is the live-dead-live sequence in NFPA 70E-2021, 120.5, a consensus standard that binds through your employer's electrical safety program or where your jurisdiction or contract has adopted that edition. The lockout and tagging requirement for work on electrical circuits sits at 29 CFR 1910.333(b)(2) for general industry, with the construction counterpart at 29 CFR 1926.417.
• Water at or near energized equipment: de-energize before anyone approaches the standing water, using the same prove-dead sequence, because water paths are invisible and a wet floor is a contact path.
• Structural or overhead instability: clear the area beneath and around it, keep the customer out of it as a matter of ordinary care, and do not load, lever or lean on it to get a better photograph.

Note the split in duties: OSHA obligations run to your own employees. Your obligation toward the homeowner standing behind you is the ordinary duty of care, and it is not discharged by citing a standard at them. Get them out of the way and tell them why.

Once the hazard is controlled, the record gets made, and it gets made before anything is repaired.

The one job the record has

A stranger, a year later, has to be able to look at it and conclude four things without asking you a question:

1. What it is. Identified specifically enough that it cannot be confused with a similar component elsewhere in the building.
2. Where it is. Located by something that will still be there next year.
3. How bad it is. Scaled, so the reader can judge size and extent instead of trusting an adjective.
4. When it was seen, relative to your work. Established as before you touched it.

Every field below exists to serve one of those four. Anything that serves none of them is administrative and can be dropped.

Observation, conclusion and recommendation are three separate sentences

The most damaging habit in condition records is collapsing all three into one line. "Bad valve, needs replacing" is a conclusion, a recommendation and no observation at all, and a year later there is nothing behind it.

Write them separately, in this order, and label them:

• Observed: what your eyes and instruments registered. No inference. "Green crystalline deposit at the joint, roughly 40 mm along the fitting, dry to the touch, no active drip during a 10-minute watch."
• Concluded: what you think it means, stated as your professional read. "Consistent with a slow past seep at the joint, likely long-standing."
• Recommended: what you propose. "Replace the fitting and the 300 mm of line either side, quoted separately."

The reason for the separation is that the observation survives being wrong about the conclusion. If your conclusion turns out to be mistaken, a clean observation still stands and still shows the condition pre-existed. A merged sentence falls entirely when any part of it is challenged.

The photograph set

Three frames per condition, always in this order, because the order is what makes the set intelligible later.

Wide, for location. Far enough back to include a permanent landmark: a doorway, a service panel, a window, a room corner. The single most common defect in trade photography is a close-up of a component with no way to prove which component, in which building, it is.

Medium, for context. The component with its immediate surroundings and its connections visible, so the reader sees what it is attached to and what runs alongside it.

Close, with a scale. A rule or a known object in frame, in the same plane as the condition. Use a non-conductive rule near any electrical equipment, and never insert a metal rule into an enclosure, energized or not, since proving dead does not make the practice acceptable and does not protect the next person who reads it as normal. Photograph a surface temperature reading with a non-contact instrument rather than touching a hot surface to check it.

Do not clean the area first. A wiped-down joint is a modified joint, and the wiping is exactly the kind of small helpfulness that removes deposits which would have dated the seepage.

The record, filled in

Here is a complete condition note for a real shape of finding: heat damage discovered at a connection point inside an enclosure while doing unrelated work.

Job: service call, unrelated fault on a separate circuit. Condition found: discoloration and deformation at a terminal connection. Location: main enclosure, interior right, second position from the top, garage north wall. Wide photograph shows the enclosure and the garage door for orientation. State when found: enclosure cover already off for the original scope of work. Circuit de-energized at the source, locked and tagged, proven dead with a rated meter checked on a known live source before and after, before any close inspection or photography inside the enclosure. Observed: brown discoloration on the insulation for approximately 25 mm back from the terminal. Terminal body discolored. Insulation stiff and cracked at the discolored area. No soot deposit on the enclosure interior. Adjacent terminals show no discoloration. Measured: none taken. The circuit was de-energized before inspection, so no load reading was available at the time of the finding. Concluded: consistent with sustained heating at this connection over time rather than a single event, on the basis that the discoloration is graded rather than sharply bounded and the neighbouring terminals are clean. Cause not established. A loose connection, an undersized conductor for the load and a failing device all remain possible. Recommended: do not re-energize this circuit until the connection is examined and the cause established. Quoted separately. Photographs: 3 frames, wide, medium and close with a non-conductive rule. Customer notified: same day, in person, and by written summary the same evening. Customer acknowledged receipt. Time spent: 0.4 hours, recorded on the ticket. Not done: nothing was tightened, cleaned, replaced or re-terminated.

Two fields in that record do work that shops routinely skip.

"Measured: none taken" with the reason. A blank field reads as an oversight. A field that states what was not measured and why reads as a deliberate limit, and it protects you from the later question of why there is no load reading. Recording an absence is a real part of the artifact.

"Not done" is the field that decides the dispute. It is the affirmative statement that you left the condition exactly as found, and it is the difference between a record and a repair log. Without it, a reader a year later has to infer from silence that nothing was touched, and inference is what the other side gets to do too.

The recommendation is also worth reading twice. It says do not re-energize until the cause is established, and it does not name the loose connection as the fault. A discolored terminal has at least three candidate causes and treating the visible one as the answer is how a shop replaces a component and gets the same finding back in six months.

Total cost of that record: 0.4 hours. The work it prices out is a quoted repair, and what it buys you, a year later, is a dated position in the argument over whether your technician caused the damage in that enclosure, instead of your word against theirs.

Where the record gets destroyed

Three moments, all of them well intentioned.

The technician who tidies. Tightening the terminal takes fifteen seconds and feels like the responsible thing. It also converts an observation into an unprovable claim and, if the connection later fails, puts your hands on it last.

The second visit that photographs the repair. Plenty of shops document thoroughly and do it all after the fix. Those photographs prove workmanship, which is worth having, and prove nothing about what pre-existed.

The verbal-only notification. The condition was found, the customer was told, nobody wrote it down. A year later that conversation did not happen. Same day, in writing, with an acknowledgement, or it is not a notification.

What changes the approach

Customer or facility restricts photography. Some commercial sites and some tenants prohibit cameras. Then the record is written, in more detail, with measured dimensions and positions substituting for the frames, and it notes in the record that photography was declined and by whom. Do not let a restriction become an excuse for no record.

Occupied residential spaces. Frame tightly enough to capture the condition and its landmark without capturing more of somebody's home than the record needs. Tell the customer what you are photographing and why before you start.

Conditions in confined or difficult spaces. Do not enter a tank, pit, vault or similar space to document a condition on the basis that you are only going to look. Space entry is governed by the permit-required confined spaces requirements at 29 CFR 1910.146 for general industry, with the construction counterpart in 29 CFR 1926 Subpart AA, and evaluating the atmosphere before entry is part of that. Photograph from outside the space, or arrange the entry properly as its own task.

References

• 29 CFR 1910.333(b)(2), lockout and tagging for work on electrical circuits and equipment, general industry, with the construction counterpart at 29 CFR 1926.417
• NFPA 70E-2021, 120.5, establishing an electrically safe work condition, including the live-dead-live instrument check; a consensus standard binding through an employer's electrical safety program or through adoption by a jurisdiction or contract
• 29 CFR 1910.146, permit-required confined spaces, general industry, with the construction counterpart at 29 CFR 1926 Subpart AA
• See related: Documenting the Unsafe Condition You Found; Documenting a Mid-Job Discovery to Protect the Re-Quote; What a Differing Site Condition Actually Is
• What a contemporaneous record is worth in a dispute depends on the forum, on your state's rules, and on how it was made and kept; this piece is about making a record that answers the four questions, not about how it will be treated legally