FAT Reports: What to Ask For and How to Read Them
The salesperson tells you: "Don't worry, the transformer passed all the tests." They send you a 3-page PDF — a table of green checkmarks. You sign the acceptance. The equipment arrives on site. Six months later, it fails.
When you trace it back, you discover the "test report" was an internal factory checklist — no measured values, no acceptance criteria, no reference to IEEE C57.12.90.
A misread FAT is more dangerous than no FAT at all: it gives you false confidence.
We've reviewed hundreds of FAT reports. The good ones run multiple pages, with numerical values, standard references, and a real inspector's signature. The bad ones are a checkbox table that wouldn't survive 30 seconds of scrutiny. This guide tells you how to tell the difference.
What a FAT Is — and Why It's Your Last Line of Defense
The Factory Acceptance Test is the complete suite of tests performed on a finished transformer before it leaves the factory and before it hits the ocean. It's your last chance to catch a problem before the equipment is on a container ship.
Once it leaves, the cost to correct a defect multiplies by a factor of 5 to 20.
The Tests You Must Demand (Per IEEE C57.12.90)
Routine Tests — Every Single Unit
| Test | IEEE C57.12.90 Reference | What It Verifies |
|---|---|---|
| Resistance measurements | §6 | Continuity, detect loose connections |
| Polarity and phase-relation | §5 | Correct terminal markings and vector group |
| Ratio tests | §5 | Turns ratio within tolerance at every tap |
| No-load loss and excitation current | §8 | Core quality — high losses = low-grade steel or poor assembly |
| Load loss and impedance voltage | §9 | Copper quality and winding integrity |
| Applied-potential dielectric test (60 Hz, 1 min) | §10.2 | Winding-to-winding and winding-to-ground insulation |
| Induced-potential dielectric test | §10.3 | Turn-to-turn insulation |
For pad-mounted transformers specifically: IEEE C57.12.34 and ANSI C57.12.22 add requirements for the enclosed compartment design — including the integrity of the tamper-resistant enclosure, the bushing well interfaces for separable insulated connectors (loadbreak and deadbreak), and the parking stands.
Design Tests — When the Design Is Not Yet Certified
| Test | IEEE C57.12.90 Reference | Key Parameter |
|---|---|---|
| Temperature rise | §11 | Top-oil rise, winding rise, hottest-spot — ≤ 65°C winding rise is standard |
| Lightning impulse (BIL) | §10.4 | Full-wave and chopped-wave — the BIL rating on your nameplate isn't theoretical |
| Short-circuit withstand | IEEE C57.12.00 §5 and C57.12.90 §12 | Mechanical and thermal withstand — this is the test that separates real capability from catalog claims |
How to Read a Real FAT Report
A professional FAT report contains these elements. If any one is missing, you didn't receive a test report — you received an opinion.
1. Nameplate Data (Must Match Your Purchase Order Exactly)
Serial number. Rated kVA/MVA. Primary/secondary voltage (kV). Connection diagram. Impedance (%). Frequency (Hz). If any of these don't match your PO, the entire report is invalid.
2. Test Conditions
Ambient temperature during testing. Oil temperature (if applicable). Test power frequency. Which measurement instruments were used — including make, model, and calibration date.
3. Measured Values — Not "PASS / OK"
Every test must show: measured value (a number), acceptance criterion (a limit or tolerance), result (pass/fail), and the referenced standard clause.
4. Inspector's Signature
Name. Title. Date. Not a blurry "QC Department" stamp.
The Five Most Serious FAT Report Errors
Error 1: Checkmarks only, no values. Reject it. Demand measured numbers.
Error 2: Test frequency or conversion method is not stated. Some measurements are frequency-dependent, and an unexplained result cannot be audited. Ask which tests were performed at the rated frequency, whether any approved conversion method was used, and where the governing standard permits it.
Error 3: No temperature correction. Resistance and losses vary with temperature. The report must state the reference temperature (load loss typically at 75°C, insulation resistance typically corrected to 20°C).
Error 4: No acceptance criteria. Every measured value needs a criterion next to it: "≤ 65 K," "≥ 1,000 MΩ," "≤ 0.5% deviation." Without criteria, you don't know what passed and what didn't.
Error 5: No inspector's signature. A signature represents accountability. When something goes wrong, you need to know who to call.
Witnessed FAT vs. Remote FAT
| Witnessed FAT (On-Site) | Remote FAT (Live Video) | |
|---|---|---|
| Watch tests in real time | ✅ | ✅ |
| Request a repeat of any test | ✅ | ✅ |
| Verify equipment identity | ✅ | ⚠️ Partial |
| Travel cost | $2,000–$5,000 | $0 |
| Time required | 3–5 days | 4–6 hours |
For high-value orders, a witnessed FAT pays for itself — the travel expense is a fraction of the cost of receiving one defective unit.
What Happens When the FAT Fails
- Document the non-conformance in writing — test number, measured value, expected value, acceptance criteria
- Demand root cause analysis — manufacturing error, design error, or measurement error?
- Require correction and re-test — at the manufacturer's expense
- Do not accept "we'll handle it on site" — factory corrections are ten times cheaper and ten times faster
Pre-Acceptance Checklist: Before You Sign a FAT Report
- Nameplate data (kVA, kV, Z%, connection) matches my purchase order — line for line
- All IEEE C57.12.90 routine tests completed (resistance, ratio, polarity, no-load loss, load loss, applied potential, induced potential)
- Every test shows measured values — not "OK"
- Every test has explicit acceptance criteria
- Test frequency (50/60 Hz) is correct for my grid
- Report states the reference temperature for loss and resistance measurements
- Inspector's name, title, and signature are present
- Photos of the transformer during testing are attached
- For pad-mounted units: compartment integrity, bushing well interface, and parking stand verification included
Conclusion
The FAT is not a formality. It's your insurance that the transformer you paid for is the transformer you receive. A professional report runs multiple pages, contains numerical values, cites IEEE C57.12.90 clauses, references calibrated instruments, and is signed by a real human being. If what you received is a table of green checks with no numbers, you didn't get a test report — you got an opinion. And opinions don't survive a short circuit.
From TransformerGrid
TransformerGrid does not operate the factory test laboratory. We coordinate the FAT and review the report package between the buyer and the selected manufacturing partner: approved nameplate data, applicable standard and clauses, measured values, acceptance criteria, instrument calibration status, deviations, signatures, and supporting photos or live witness arrangements.
Our current verified manufacturing partner can provide FAT evidence for 50 Hz and 60 Hz projects, supported by KEMA documentation covering 250 MVA, 345 kV, and 60 Hz capability. The test method and test frequency must be confirmed for each required test rather than assumed from the destination frequency alone.
Do not accept a report made only of green checkmarks. Ask TransformerGrid to coordinate a relevant sample FAT report and a project-specific FAT review.
Related articles: