Q: Should I install an on-line DGA monitor, or continue sending oil samples to the lab?
A: Each method has a distinct role. The choice depends on transformer criticality, fault history, and budget.
|
Criteria |
Laboratory DGA |
On-line DGA Monitor |
|
Interval |
Quarterly to annually |
Continuous (every 30 min – 24 hours) |
|
Gas detection |
All 7+ gases (H₂, CH₄, C₂H₆, C₂H₄, C₂H₂, CO, CO₂, O₂, N₂) |
Typically 5–9 gases (varies by model) |
|
Accuracy |
High (lab GC: ±1 ppm) |
Moderate (±5–15% of reading) |
|
Trend sensitivity |
Limited by sampling frequency |
Excellent - captures load-induced changes |
|
Capital cost |
~$50–100 per sample |
$5,000–$30,000 per monitor |
|
Operating cost |
$200–1,000/year (quarterly) |
$500–2,000/year (consumables, calibration) |
|
Best for |
Small fleet (≤ 10 units); non-critical |
> 100 MVA, EHV/UHV; remote; known-fault |
Optimal strategy - hybrid approach:
1. Install on-line monitors on the top 10–20% most critical transformers.
2. Use quarterly lab DGA on all other units.
3. Cross-validate on-line monitors with a lab sample every 6–12 months.
4. If a developing fault is detected, immediately take a lab sample for confirmation + full spectrum.
