10.2 Evaluating Test Results Against NETA, Manufacturer, and Baseline Data

Evaluating Test Results Against NETA, Manufacturer, and Baseline Data

Quick Answer: Level III domain 3.3 Troubleshooting and Analysis (10-15 % of the Level III exam) contains task 3.3.2, "Analyze and interpret tests performed" — explicitly including reviewing "test results recorded by Level II and III technicians and compare to manufacturers' [data]." The hierarchy is fixed and NETA states it plainly: manufacturer's published data first; the NETA table only in its absence.

---

1. The hierarchy of acceptance criteria

NETA's tables carry the qualifying language in the standard itself: "In the absence of consensus standards dealing with insulation-resistance tests, the Standards Review Council suggests the above representative values." They are a fallback, deliberately so, because a generic table cannot know the design of the specific apparatus in front of you.

The practical consequence: a technician who reports "fails NETA" on a reading that satisfies the manufacturer's own published limit has misapplied the hierarchy. And the reverse is more dangerous — a reading that satisfies a generic NETA minimum but falls below the manufacturer's specific limit is a failure, and reporting it as a pass conceals a real defect.

2. The three comparisons that give a number meaning

A single absolute reading is the weakest form of evidence. Three comparisons make it strong.

Phase-to-phase within the unit. The most sensitive comparison available, because ambient temperature, humidity, instrument calibration, lead configuration, and technique are common to all three phases and therefore cancel. Three windings that read 2,100, 2,050, and 1,150 MΩ tell you something even if all three exceed the table minimum. Typical field expectations are that phases agree within a modest percentage; the outlier is the finding.

Against the unit's own baseline. A baseline taken at commissioning is the single most valuable data point in a maintenance program, because it captures that specific unit's normal condition. A winding that has fallen from 5,000 MΩ at commissioning to 800 MΩ today is deteriorating rapidly, even though 800 MΩ comfortably exceeds any table minimum. The rate of change, not the absolute value, is the diagnosis.

Against sister units. Where several identical units of the same age and duty exist, they form a natural population. A unit that sits well outside the group is worth investigating regardless of where the group sits relative to a table.

3. Corrections that must precede comparison

Comparing raw numbers taken under different conditions produces false trends. Before any comparison:

Record the conditions or the correction is impossible. Ambient temperature, winding or apparatus temperature, relative humidity, load current, and the instrument used and its serial number are not clerical extras — they are what make the data comparable next year. A test record without temperature cannot be trended, which means the test was, in trending terms, wasted.

The direction rule that catches people: insulation resistance goes down as temperature goes up, so a hot reading corrected to 20 °C moves up. Conductor resistance goes up as temperature goes up. Applying either correction in the wrong direction produces an error of exactly twice the correction.

4. What a below-minimum reading actually means

NETA's language for a low insulation resistance is that values below the table or the manufacturer's recommendation "shall be investigated." It does not say "shall be rejected."

That distinction is deliberate and heavily examined. The correct response chain:

1. Verify the measurement. Repeat it. Check lead condition, guard connections, instrument range, and whether the specimen was properly isolated and discharged.
2. Check the obvious external causes: surface contamination on bushings and insulators, moisture and dew, a connected item that should have been disconnected (a surge arrester, a PT, a capacitor), or a still-connected cable adding its own surface leakage.
3. Apply the corrections in Section 3 and re-evaluate.
4. Compare across phases and against baseline.
5. Run complementary tests. A low IR with an acceptable power factor and normal capacitance suggests a surface problem; a low IR with elevated power factor suggests bulk deterioration. DAR and PI distinguish moisture and contamination from bulk aging.
6. Then form a conclusion and recommend an action.

Condemning equipment on a single uncorrected low reading is the failure mode this sequence exists to prevent — and it is expensive, since the recommendation may be to take a critical asset out of service.

5. ATS versus MTS — different questions, same numbers

The commonly held belief that ATS acceptance limits are numerically stricter than MTS limits is not generally true, and the shared Table 100.1 is the clearest counterexample: one column of values, applicable in both standards. What genuinely differs is which tests are required, what stress level is applied, and what the result is compared against.

The reduced field test voltage on aged insulation is the real ATS/MTS difference in stress: a factory-level overpotential applied to service-aged insulation can cause the failure it was looking for, so maintenance testing uses lower levels.

6. Writing the evaluation

An evaluation states four things:

1. What was measured, with units and conditions.
2. What it was compared against, named specifically — "manufacturer's published minimum of X" or "NETA Table 100.1 in the absence of manufacturer data" or "commissioning baseline of Y."
3. The conclusion — satisfactory, satisfactory with monitoring, deficient, or unsatisfactory.
4. The recommended action, with urgency.

"IR low" is not an evaluation. "C-phase winding-to-ground insulation resistance 1,150 MΩ corrected to 20 °C, against 2,100 and 2,050 MΩ on A and B phases at the same test and a commissioning baseline of 3,400 MΩ; power factor on the same winding is also elevated at 1.4 % against 0.6 % on the other phases; conclusion is bulk moisture ingress in the C-phase winding; recommend removing from service for a dry-out and retest before re-energization" is.

Exam trap: A question gives an insulation resistance reading below the NETA Table 100.1 value and offers "remove the equipment from service and replace it" as an option. NETA's instruction is that such a value shall be investigated — the correct next steps are verification, correction, and complementary testing, not condemnation on a single number.