Transformer Oil Sampling Guide: IEC 60475 Methods, Container Selection, Light-Exclusion & Chain of Custody

Abstract

Oil sampling is the most common — and most commonly botched — diagnostic procedure in transformer maintenance. A contaminated or poorly handled sample can produce DGA results that trigger unnecessary outages or, worse, mask an incipient fault. IEC 60475:2011 establishes the globally accepted methodology for sampling insulating liquids from transformers, reactors, bushings, LTCs, cables, and switchgear. This guide translates the standard into actionable field procedures: which container to use and when, how to flush the sampling port, why glass syringes outperform aluminum bottles for DGA, the physics of light-induced degradation, and the complete chain-of-custody documentation required for defensible results. Every step is described at the level of detail a production engineer or commissioning technician needs to execute correctly on the first attempt.

1. Why Sampling Technique Determines Diagnostic Accuracy

Consider a DGA sample that is supposed to detect 2 ppm of acetylene (C₂H₂) — a critical indicator of arcing — but loses 1.5 ppm of gas to an unsealed container during the 3-day transit to the lab. The transformer has an active high-energy discharge, but the lab report shows "C₂H₂ = 0.5 ppm — within specification." The transformer fails catastrophically three months later.

This scenario is not hypothetical. Studies by CIGRE Working Group D1.32 have demonstrated that:

30-50% of dissolved hydrogen can be lost from a sample through an unsealed plastic syringe within 24 hours at ambient temperature
Acetylene loss of 20-40% occurs in samples exposed to direct sunlight for 4+ hours (photo-degradation)
Moisture contamination of 5-15 ppm is introduced when sampling in high-humidity conditions without proper port flushing

The chain from transformer to lab report is only as strong as its weakest link — and the weakest link is almost always the sample extraction.

2. Container Selection: Glass Syringe vs. Aluminum Bottle vs. Glass Bottle

IEC 60475 permits several container types, each with specific use cases.

2.1 Glass Syringe (Recommended for DGA)

Construction: Borosilicate glass barrel with ground-glass or PTFE plunger, Luer-lock tip, capacity 50-250 mL.

Advantages:

Gas-tight when fitted with a stopcock or Luer-lock cap with septum
Transparent — allows visual inspection for air bubbles, free water, or particulate contamination
No headspace required: the plunger moves to accommodate thermal expansion, maintaining a zero-headspace condition
Compatible with automated headspace samplers used in modern DGA labs
Reusable after proper cleaning and drying

Disadvantages:

Fragile — requires padded transport case
Plunger can stick if oil contains sludge or the syringe was improperly cleaned
Temperature-sensitive — plunger seal may leak if oil cools and contracts, drawing air past the seal
Limited volume (generally ≤250 mL)

Field procedure for glass syringes:

Flush the syringe with sample oil at least 3 times before taking the final sample
Fill to 80-90% of barrel volume (leave room for thermal expansion during transport)
Immediately cap with a gas-tight Luer-lock cap with PTFE septum
Wrap in aluminum foil (light protection)
Place in a padded, rigid transport case
Transport within 4 days; analyze within 14 days of sampling

2.2 Aluminum Bottle (Alternative for DGA, Physical/Chemical Tests)

Construction: Seamless aluminum bottle with self-sealing PTFE-lined cap, capacity 500-1000 mL.

Advantages:

Robust — survives rough shipping without breakage
Light-tight (opaque aluminum blocks UV)
Large volume suitable for combined DGA + physical/chemical testing from one sample
Self-sealing cap prevents gas loss during transport

Disadvantages:

Cannot visually inspect for bubbles or sediment
Residual air in the bottle headspace can absorb dissolved gases from the oil (partitioning effect), reducing gas concentrations — though this is mitigated by the self-sealing cap and the Ostwald coefficient of each gas
Internal corrosion possible if bottle is stored wet after cleaning
More expensive per sample than glass syringes

2.3 Glass Bottle (Physical/Chemical Tests Only — NOT for DGA)

Construction: Amber borosilicate glass bottle with PTFE-lined cap, capacity 500-1000 mL.

Usage: Suitable for BDV, moisture (Karl Fischer), acidity, IFT, tan δ, viscosity, and PCB testing. Not suitable for DGA because the large headspace and non-zero-headspace design allow dissolved gases to escape into the bottle air space, rendering gas-in-oil concentrations unreliable.

2.4 Container Selection Matrix

Test	Preferred	Acceptable	NOT Acceptable
DGA (dissolved gas analysis)	Glass syringe (gas-tight)	Aluminum bottle (self-sealing cap)	Glass bottle, plastic bottle, open container
Moisture (Karl Fischer)	Glass syringe	Aluminum bottle, amber glass bottle with septum cap	Open container, unsealed plastic
BDV (breakdown voltage)	Amber glass bottle	Aluminum bottle	Open plastic container
Acidity, IFT, tan δ, viscosity	Amber glass bottle	Aluminum bottle	Any container contaminated with detergent or solvent residue
Furan (HPLC)	Amber glass bottle or glass syringe	Aluminum bottle	Plastic container (phthalate contamination)
PCB analysis	Amber glass bottle	Aluminum bottle	Plastic container (PCB absorption into plastic)

3. The Sampling Procedure: Step-by-Step per IEC 60475

3.1 Pre-Sampling Preparation

Check transformer condition: The transformer should be at normal operating temperature (60-80 °C top oil) — warm oil flows more easily and dissolved gases are at equilibrium. Cold oil (<10 °C) may not be representative; if sampling from a cold transformer is unavoidable, note this on the sample record.
Assemble equipment:
Sampling container (glass syringe with Luer-lock cap, or aluminum bottle)
PTFE or nitrile sampling hose (1-2 m length, 6-8 mm ID), rated for 100 °C
Sampling adapter (matching the transformer sampling valve thread)
Waste oil container (5-10 L)
Nitrile gloves (do not handle sample containers with bare hands — skin oils contaminate)
Aluminum foil for light protection
Permanent marker for labeling
Chain-of-custody form
Padded transport case or insulated cooler box
Label the container before filling: Transformer ID, sampling point (main tank, OLTC, bushing), date, time, and sampler's name.

3.2 Flushing the Sampling Port

This is the single most important step for obtaining a representative sample. The sampling port and any dead volume in the valve body contain stagnant oil that has been in contact with the atmosphere and is not representative of the bulk oil. IEC 60475 requires:

Open the sampling valve and drain at least 2 liters of oil (for a standard ½-inch valve) into the waste container. For large transformers with long sampling pipes, drain 5-10 liters.
Observe the oil flow — it should be continuous, free of air bubbles, and representative in color. If the oil appears milky (free water), dark (sludge), or contains visible particles, continue flushing.
Attach the clean sampling hose to the valve adapter. Flush the hose with at least 500 mL of the flowing oil before connecting to the sample container.
Reduce the oil flow rate to a gentle stream — turbulent flow introduces air bubbles and can strip dissolved gases from the oil.

3.3 Filling the Container

Glass Syringe Method (DGA):

Connect the sampling hose to the syringe Luer-lock inlet.
Slowly draw oil into the syringe, filling to approximately 20-25% of barrel volume.
Disconnect, invert the syringe, and expel the oil into the waste container. This flushing step removes residual air and contamination.
Repeat the flush at least 2 more times (total 3 flushes).
On the final fill, draw oil to 80-90% of barrel volume.
Immediately disconnect the hose and cap the syringe with the gas-tight Luer-lock cap (or close the stopcock).
Do not leave air bubbles in the syringe. If a small bubble is present, hold the syringe tip-up and gently push the plunger to expel it through the Luer port before capping.
Wrap in aluminum foil immediately.

Aluminum Bottle Method:

Insert a clean PTFE tube through the bottle cap opening to the bottom of the bottle.
Allow oil to overflow from the bottle for at least 30 seconds (or until at least 2 bottle volumes have overflowed). This displacement filling minimizes air contact.
While oil is still overflowing, slowly withdraw the filling tube and immediately tighten the self-sealing cap.
Invert the bottle and check for leaks. If any oil seeps through the cap, discard the sample and re-sample with a new bottle/cap.
Wipe the exterior clean and wrap in a sealed plastic bag.

3.4 Post-Sampling

Close the transformer sampling valve.
Clean any oil spills from the transformer and ground.
Place the sample container in a padded, opaque transport case.
Complete the chain-of-custody form (Section 5).
Transport to the laboratory as quickly as possible — ideally within 24 hours for DGA, 4 days maximum.
If immediate transport is not possible, store the sample in a cool, dark place (15-25 °C) and ship as soon as practical.

4. Light Protection: Why It Matters

Transformer oil is photochemically reactive. When exposed to ultraviolet (UV) radiation (sunlight, fluorescent lighting), two degradation pathways are activated:

Photo-oxidation: UV photons cleave hydrocarbon chains, generating free radicals that react with dissolved oxygen to form peroxides, aldehydes, ketones, and ultimately organic acids. This affects acidity, IFT, tan δ, and color.

Dissolved gas photodegradation: Certain dissolved fault gases are photolabile:
Acetylene (C₂H₂) is particularly sensitive — UV exposure converts C₂H₂ to polymerized products, reducing the measured concentration
Hydrogen (H₂) can photochemically react with unsaturated hydrocarbons

Field rule: All oil samples must be protected from light from the moment of extraction. Wrap glass syringes in aluminum foil immediately after capping. Aluminum bottles are inherently light-tight but should still be stored in opaque transport cases. Laboratory storage prior to analysis should also be light-protected.

5. Chain of Custody and Documentation

A complete chain-of-custody record establishes the legal and technical defensibility of the sample results. IEC 60475 specifies the minimum information required:

5.1 Sample Label (on container)

Unique sample ID number
Transformer asset ID / serial number
Sampling point: "Main Tank — Bottom Valve" or "OLTC Compartment"
Date and time of sampling
Sampler's name and signature

5.2 Chain-of-Custody Form (accompanies sample to lab)

Field	Description
Sample ID	Matches label on container
Transformer ID	Serial number, substation name, bay designation
Manufacturer & rating	e.g., "ABB, 50 MVA, 132/33 kV, 2018"
Oil type	Mineral (inhibited/uninhibited), natural ester, synthetic ester, silicone
Sampling point	Main tank, OLTC, bushing (specify which), cable box
Transformer status	In service / out of service / recently energized
Top-oil temperature at sampling	°C
Ambient temperature & humidity	°C, %RH
Oil appearance	Clear / cloudy / dark / particles visible
Sampling method	Container type, flushing volume, any deviations from IEC 60475
Tests requested	DGA, moisture, BDV, acidity, IFT, furan, PCB, etc.
Date/time sampled	DD/MM/YYYY HH:MM
Sampler	Name, signature
Lab receipt	Date/time received, received by, sample condition on arrival

6. Special Sampling Situations

6.1 OLTC (On-Load Tap Changer) Oil

OLTC oil is sampled separately from the main tank — the OLTC compartment is sealed from the main tank by the diverter switch housing. OLTC oil typically has higher dissolved gas content and carbon particles from arcing during tap changes, making cross-contamination a significant concern. Use a separate, dedicated sampling hose for OLTC sampling. Never use the same hose for main tank and OLTC without thorough solvent flushing and drying — acetylene concentrations as low as 1-2 ppm carried over from OLTC to main tank will trigger a false arcing alarm on the main transformer.

6.2 Bushing Oil

Capacitance-graded bushings contain a small oil volume (typically 5-50 L). Because the volume is small, the flushing volume must be correspondingly reduced to avoid lowering the bushing oil level below the minimum. Flush only 1-2 liters, then fill a small (50 mL) glass syringe. Always check the bushing oil level gauge before and after sampling.

6.3 Cold Climates

When sampling at ambient temperatures below -10 °C, the oil viscosity is high and dissolved gas equilibrium may be disturbed. If possible, energize the transformer and allow it to warm up (top oil >30 °C) before sampling. If sampling cold is unavoidable, insulate the sampling hose, use a shorter hose to minimize cooling, and note the temperature on the chain-of-custody form.

FAQ

Q: Why do we use glass syringes instead of plastic syringes for DGA?

Plastic syringes (polypropylene, polyethylene) are permeable to gases. Hydrogen — the smallest and most mobile gas molecule — diffuses through plastic walls at a significant rate: studies show 5-10% H₂ loss through a polypropylene syringe wall within 2 hours. Plastic syringes also typically have rubber plunger seals and tip caps that absorb hydrocarbons (especially acetylene and ethylene) and can outgas plasticizers, contaminating the sample. Glass syringes with PTFE plungers and Luer-lock caps are gas-impermeable and chemically inert. The only case where a plastic syringe is acceptable is for immediate on-site BDV testing, where the sample is analyzed within minutes of extraction.

Q: How much oil should I flush before taking a sample?

IEC 60475 specifies flushing at least 2 liters of oil from the sampling port before sample collection for a standard ½-inch valve. However, the actual required volume depends on the length and diameter of the sampling pipe between the transformer tank wall and the valve. The goal is to flush 2-3 times the dead volume of the sampling pipe. For a transformer with a 2-meter sampling pipe of 12 mm ID, the dead volume is approximately 0.23 L; 2 L represents roughly 9 dead volumes. For transformers with 3-5 meter sampling pipes (common on large units), increase flushing to 5-10 liters. Always observe the oil appearance — continue flushing until the oil is clear, bubble-free, and representative of the bulk oil.

Q: What temperature should the transformer be at for oil sampling?

Ideally, 60-80 °C top-oil temperature, which represents normal operating conditions. At this temperature: (1) the oil viscosity is low, ensuring good flow; (2) dissolved gases are at thermal equilibrium with the oil — this is essential because the partition coefficients (Ostwald coefficients) used to interpret DGA results assume the oil was at operating temperature when sampled; (3) moisture is dissolved in the oil rather than separated as free water. If the transformer is cold (<20 °C), dissolved gases may not be representative of the equilibrium condition, and moisture may exist as free water rather than dissolved. Always record the oil temperature at the time of sampling.

Q: How long can an oil sample be stored before analysis?

For DGA samples in gas-tight glass syringes: 4 days maximum from extraction to lab receipt; analyze within 14 days. For physical/chemical tests (BDV, moisture, acidity): 7 days maximum for sealed containers. The critical concern for DGA is gas loss — hydrogen is the most volatile and shows measurable loss after 4 days even in gas-tight containers. After 14 days, acetylene concentrations can also degrade measurably. If analysis cannot be performed within these timeframes, note the delay on the report and interpret results with caution.

Q: What happens if an air bubble gets into the DGA syringe?

An air bubble in a DGA syringe is problematic because dissolved gases in the oil will partition into the air bubble to reach equilibrium (Henry's Law). This reduces the measured concentrations in the oil phase, with lighter gases (H₂, O₂, N₂) being most affected. A bubble occupying 5% of the syringe volume can reduce the measured hydrogen concentration by 20-40%. If a small bubble is noticed immediately after sampling, hold the syringe vertically (tip up), gently tap to coalesce bubbles, and expel them through the Luer port before capping. If a large bubble is present, discard the sample and re-sample.

Q: Can I sample OLTC oil and main tank oil through the same hose?

Only if you thoroughly flush the hose with clean, dry oil (or a suitable solvent followed by complete drying) between sampling points. Acetylene concentrations in OLTC oil routinely reach 10-100 ppm, while a healthy main tank should show <0.5 ppm C₂H₂. Even 0.1 mL of OLTC oil residue in the hose can contaminate a main tank sample with enough acetylene to trigger a DGA alarm. Best practice: maintain separate, clearly labeled sampling hoses for main tank and OLTC sampling, dedicated to each transformer.

References / Standards

Reference	Title
IEC 60475:2011	Method of sampling insulating liquids
IEC 60567:2011	Oil-filled electrical equipment — Sampling of gases and analysis of free and dissolved gases — Guidance
IEC 60599:2022	Mineral oil-filled electrical equipment in service — Guidance on the interpretation of dissolved and free gases analysis
IEC 60422:2013	Mineral insulating oils in electrical equipment — Supervision and maintenance guidance
IEC 60296:2020	Fluids for electrotechnical applications — Mineral insulating oils for electrical equipment
ASTM D923-15	Standard Practices for Sampling Electrical Insulating Liquids
ASTM D3612-02(2017)	Standard Test Method for Analysis of Gases Dissolved in Electrical Insulating Oil by Gas Chromatography
CIGRE TB 296	Recent Developments in DGA Interpretation

*Authored by Du Fu, Production Engineer at ZY POWER. Sampling procedures described herein are based on IEC 60475:2011 and the author's field experience. Always comply with site-specific safety protocols, including arc-flash PPE requirements when sampling from energized transformers.*