11.5 As-Found / As-Left Documentation

Definitions

As-found values are the test results captured before any maintenance, adjustment, or repair is performed. As-left values are the results captured after the work is complete and the equipment is returned to service condition. The distinction is not bookkeeping: as-found documents the relay's condition before the technician made any adjustment, which is the only evidence of how the equipment had actually been performing in service; as-left confirms the final calibrated state that the system will now rely on. Both are required on formal NETA acceptance and maintenance reports.

Why Both Matter

As-found proves what was wrong. If a relay was picking up at 4.8 A when the setting called for 5.0 A, the as-found record is the evidence — it supports a misoperation analysis, justifies the corrective action, and protects the technician and the owner from later disputes about whether the equipment was actually out of spec before the work began. As-left proves what was corrected; the same relay now picks up at 5.0 A within tolerance. Without as-left, you cannot demonstrate that the equipment left the shop in spec.

Together they form the trend baseline. Year over year, the as-found/as-left pair lets the owner see drift, predict failure, and plan maintenance instead of reacting to it. If the as-found pickup drifts from 4.99 A to 4.95 A to 4.90 A over three maintenance cycles, the trend tells the owner the relay is heading out of tolerance and should be replaced before it misoperates — even though each individual as-found was still within ±5 percent.

Required Content of a NETA Test Record

A defensible as-found/as-left entry includes:

• Equipment identification — asset tag, serial number, location, one-line reference.
• Test performed — the name of the test (e.g., "instantaneous pickup", "insulation resistance 1 minute") and the standard section it maps to.
• Instrument ID and calibration date — every measured value is traceable to a calibrated instrument; an instrument out of calibration invalidates the test.
• Test setup — connections, applied voltage or current, range, whether primary or secondary injection.
• Environmental conditions — temperature and humidity where they affect the result. Insulation resistance is temperature-corrected to 40 °C; humidity affects IR on large rotating machinery.
• Measured values — the actual numbers, all three phases where applicable.
• Acceptance criteria — the tolerance band or pass/fail threshold (e.g., "pick up within ±5 percent of setting").
• Pass/fail verdict — explicit, not implied.
• Technician signature and date — accountability.

NETA Test Report Format Essentials

NETA acceptance and maintenance test reports follow a consistent structure:

1. Project and owner identification.
2. Equipment under test, with serial numbers.
3. A table of test points with columns for as-found, as-left, and acceptance criteria.
4. Instrument IDs and calibration dates for every instrument used.
5. Environmental conditions (temperature, humidity) where they affect results.
6. A deficiencies and recommendations section.
7. Technician and reviewer signatures.

The one-line or three-line diagram is referenced, not redrawn. Deviations from the standard test plan are noted and justified. The report is the legal record of the equipment's condition on that date — write it for the reader five years from now who has only the report, not for yourself.

Traceability

Every value traces to a calibrated instrument; every instrument traces to a calibration standard; every setting traces to the coordination study or the manufacturer's instruction book; every deviation traces to an authorized engineering decision. A test report that says "IR = 1200 megohms" without an instrument ID is not traceable and is not a NETA-compliant record. As-found and as-left values must each carry their own instrument ID — using the same instrument for both is fine, but the record must show which instrument produced each value on each date. If an instrument is found out of calibration after a test, every value it produced during the out-of-cal period is suspect and must be flagged or reverified.

Worked Example — Trending a Drifting Overcurrent Relay

A feeder 51 relay is calibrated annually. The setting is 5.0 A pickup, ±5 percent acceptance band (4.75 to 5.25 A). Over four years the as-found pickup reads: Year 1 — 4.98 A; Year 2 — 4.95 A; Year 3 — 4.90 A; Year 4 — 4.82 A. Every as-found is still within the ±5 percent band, so a report that records only pass/fail would show "pass" four times and the drift would be invisible. The as-found/as-left pair, recorded to the digit, makes the trend visible: the relay is drifting toward the low end of the band at about 0.05 A per year and will cross 4.75 A sometime in Year 5 or 6. The owner can plan a relay replacement at the next scheduled outage instead of discovering a misoperation in service. This is the trending value that justifies recording the actual numbers, not just a checkmark.

Why As-Found/As-Left Is a NETA Hard Requirement

NETA acceptance and maintenance specifications require both as-found and as-left values on the test report because each answers a different question. As-found answers "what was the equipment doing before I touched it?" — the question that matters for misoperation analysis, warranty disputes, and owner liability. As-left answers "what is the equipment doing now that I have finished?" — the question that matters for return-to-service, future trending, and the next technician's baseline. A report with only as-left cannot prove the technician actually corrected anything; a report with only as-found cannot prove the equipment was returned to spec. Both are required, both must be traceable, and both must be recorded with the same rigor.

Common Documentation Errors

• Recording a pass/fail verdict without the actual measured value — destroys the trend.
• Recording as-left values without the matching as-found — cannot prove a correction was made.
• Omitting the instrument ID on one of the two values — breaks traceability for that measurement.
• Failing to temperature-correct IR before recording — produces a value that is not comparable year over year.
• Backdating or batch-filling the record at the end of the day instead of in real time — invites missed steps and is not a defensible legal record.