SPARKYQC
← Back to Sparky QC
Last updated 3 August 2026
A specification tells you which standard applies. The standard tells you what to record. This is the step in between — for each record on an electrical job, what asks for it, what has to be written on it, and what a reviewer lines it up against.
The project specification governs, not this page. Where the spec is silent, these are the documents it usually points at. Section numbers move between editions, so check the edition your spec actually names — a citation that was right in 2017 may not be right in the book on your desk.
This summarizes what a standard asks you to record. It does not reproduce the text of any standard, and it is not a substitute for the code, a permit, or a qualified engineer.
Almost every industrial and construction electrical spec is built from some combination of these. They do different jobs, and only two of them are about testing at all.
| Document | What it governs | What it leaves you holding |
|---|---|---|
| NFPA 70 National Electrical Code |
How the installation is built — conductors, raceway, terminations, grounding and bonding, identification, working space. | Torque records, ground resistance results, panel and circuit directories, conductor identification, and the manufacturer’s instructions the installation was judged against. |
| NFPA 70E Electrical Safety in the Workplace |
How people work on or near energized equipment. Nothing to do with whether the installation is correct. | Energized electrical work permits, shock and arc flash risk assessments, equipment labels, job briefings, qualified-person training records, and the lockout/tagout procedure with its audits. |
| ANSI/NETA ATS Acceptance Testing Specifications |
Testing new equipment before it is energized and handed over. | A visual and mechanical inspection plus an electrical test report for each piece of equipment, with results compared against published test values. |
| ANSI/NETA MTS Maintenance Testing Specifications |
Testing equipment already in service. | The same shape of report as ATS, but read against the last set of results as much as against a table. The trend is the finding. |
NETA in turn points at IEEE for method. IEEE 43 for insulation resistance on rotating machinery, IEEE 81 for measuring ground resistance, IEEE 400 and 400.2 for field testing shielded power cable, IEEE C57.12.90 for transformer tests. If your spec says “test in accordance with NETA ATS,” it has already pulled those in behind it.
70E is not law, and that catches people out. The enforceable rules are OSHA — 29 CFR 1910.147 for lockout/tagout, 1910.269 for power generation and distribution work, 1926 Subpart K for construction. NFPA 70E is the consensus standard your program gets measured against when an inspector or an expert witness asks what “safe work practice” meant on your job. Contracts routinely make it binding regardless.
A test that happened and a test you can prove happened are different things, and the difference is what is written in the header of the sheet. This is the minimum that makes a record defensible.
| Record | Asked for by | What has to be on it |
|---|---|---|
| Cable schedule | Project spec; NEC for identification and sizing | Cable number, from and to equipment tags, size, number of cores, insulation and voltage class, routing, and the drawing revision it was taken from. Everything else on the job references this number. |
| Cable pull | Project spec; manufacturer’s limits | Cable number, date, reel and lot, length pulled, and the pulling tension and bend radius limits if the spec set them. |
| Terminations and torque | NEC 110.14(D); NETA; manufacturer |
Cable number, termination point, lug type, the torque value applied, where that value came from, and the tool ID with its calibration date. |
| Insulation resistance | NETA ATS/MTS; IEEE 43 for machines | Cable or equipment ID, test voltage, duration, every phase combination tested, the reading for each, temperature, and the acceptance criterion it was judged against. |
| Continuity | Project spec; NEC Article 250 for bonding paths | Circuit or cable ID, core identification at both ends, method, and who witnessed it. This is the record that proves the core landed on the terminal the schedule says it did. |
| Ground resistance | NEC Article 250; IEEE 81 for method | Electrode or grid identified, method used, probe spacing, measured value in ohms, soil condition and recent weather, and the limit set by the spec. |
| Loop check | Project spec; commissioning procedure | Loop number, every device in the loop by tag, the injection point for each check, values injected and read at the display, and the date — which must fall after the termination it depends on. |
| Relay and protection testing | NETA ATS/MTS; IEEE | Relay tag, model and firmware, the settings sheet revision tested against, pickup and timing results recorded as-found and as-left, and the engineer who issued the settings. |
| Functional and operational test | NETA ATS; commissioning procedure | What was operated, the interlocks and permissives proven, expected versus observed behavior, and everyone present. |
| Panel and circuit schedule | NEC 408.4 |
A directory at the panel identifying each circuit by its specific purpose — not “lights” or “spare” where it is actually in use — kept current as circuits are added. |
| Instrument calibration | NETA; the test standard being worked to | Instrument make, model, serial, calibration date and due date, and traceability to a national standard. Every test report has to name the serial number of the instrument that produced it. |
| Qualifications | NFPA 70E; project spec | Who is qualified for what, on what evidence, and when it expires. 70E puts an outside limit on how long training stands before it is refreshed. |
| Energized work permit | NFPA 70E 130.2(B) |
The circuit and location, why de-energizing was not feasible, the shock and arc flash risk assessments, the PPE required, how access is restricted, evidence of the job briefing, and the approval signatures. |
Insulation resistance changes sharply with temperature. A reading taken on a cable at 38 °C and one taken at 15 °C are not comparable, and the acceptance criterion in your spec is written at a base temperature. Without the temperature on the sheet the reading cannot be corrected, which means it cannot be compared to anything — including the reading you took on the same cable last month.
The other half of it is coverage. A three-conductor cable has six readings, not one: each phase to ground, and each pair of phases. A single number in a box labeled “megger” does not say which of the six it is, and a reviewer who has been burned will assume the worst.
On machine windings, IEEE 43 is what a spec means by polarization index — the ten-minute reading over the one-minute — and by dielectric absorption ratio, the sixty-second over the thirty. Both are ratios of the same instrument on the same winding, so both need the same run recorded, not two visits.
Recent NEC editions require that where a manufacturer has provided a torque value for a termination, it is tightened with a calibrated tool to that value. That is why “torqued to spec” is not a record. The sheet has to say which value and whose: the manufacturer’s instruction for that lug, or the published table your spec fell back to where the manufacturer gave none.
NETA publishes bolt torque values for electrical connections in its test values section — Table 100.12 in current editions — and that table is the usual fallback. It is a fallback. Where the manufacturer has spoken, the manufacturer wins, because NEC 110.3(B) makes the listing instructions part of the installation requirement.
The NEC figure most often quoted back as a target is about a single made electrode — a rod, pipe or plate. Where one of those measures above roughly 25 ohms, the Code’s answer is to add a second electrode, not to fail the job. It was never a performance limit for a grounding system.
Substation and plant specs routinely set something far tighter, sometimes low single-digit ohms, and that number comes from the spec and the engineer’s study, not from Article 250. Record the measured value, the method, and the limit you were actually working to. Writing “<25 Ω — PASS” against a spec that demanded 5 is how a whole grid gets retested at your cost.
Method matters enough to be on the sheet. IEEE 81 fall-of-potential needs the probes far enough out to reach flat ground, and a clamp-on reading on a grid with multiple parallel paths is measuring something other than what you think. Record which one you used and the spacing.
None of these records is worth much alone. Their value is that they agree, and what makes them agree is a single identifier carried through all of them.
One cable number should appear on the schedule, on the pull record, on both termination records, on the insulation resistance result, and on the loop check that depends on those terminations. When it does, the chain is checkable by anyone in a few minutes. When the schedule says C-2104 and the megger sheet says C2104, they are two different cables to every system that will ever read them, including ours.
Agree the format at kickoff and hold it. Cable numbers, equipment tags, loop numbers, instrument IDs. This is the cheapest thing on this page to get right and the most expensive to fix at energization, because by then the sheets are signed.
A turnover reviewer is rarely re-deriving your test results. They are checking whether your own records contradict each other. The ones that get found:
Sparky QC’s Audit tab finds the subset of these that are objectively true from the records — expired dates, open items, missing attachments. It does not find the others. How Sparky QC works is explicit about which is which.
An 811 locate is a state statute, not an NEC requirement, and the rules differ by state — how long before you dig, how long the ticket lives, what counts as a positive response from each operator, and what you owe when a mark is missing. Do not carry one state’s numbers onto a job in another.
What travels is the record: the ticket number, the date called, the expiry, which operators responded, and a photograph of the marks before the first bucket. Duct bank and ground grid work is where this bites electrical crews hardest, because both start with excavation on a schedule set by someone else.
It is a map of which document asks for which record. It is not the standard, and it will not settle an argument with an engineer.
Corrections are welcome and get made. If something here is wrong or has moved in a newer edition, write to info@sparkyqc.com and say which edition you are reading.
See also the blank form pack — the sheets these records go on, free and ungated.
Spec quick reference — sparkyqc.com/spec-reference