IEC 60970 Insulating liquids - Methods for counting and sizing particles
9 Method A - Automatic particle size analyzer
9.1 Summary of method
The sample is agitated to suspend the particles, then passed at an optimum flow rate through the sensor unit of the particle counter. After the required fluid volume has been passed through the sensor the count is terminated and the results recorded.

9.2 Apparatus and auxiliary materials
- An automatic particle counter fitted with a sensor operating on the light interruption principle and suitable for counting within the range of 2,0 µm to greater than 200 µm. The automatic particle counter shall be capable of:
• Sorting particles into at least > 4μm, > 6 μm, > 14 μm size ranges.
• Providing a specific particle count and distribution for a measured quantity of reference material to a precision of +/- 10 % of the total count.
• Not saturating (i.e. providing a lower than expected count as a result of particle coincidence within the sensing zone) when analysing a suspension containing less than 2000 particles per millilitre of fluid.

NOTE Alternative detectors may be used to extend the size range of the particles measured.

- Compressed gas - a pressurized source of air or nitrogen free from oil or water contamination filtered through a 0,45 µm membrane filter. The capacity, pressure range and constant pressure controls should meet the requirements of the particular equipment in use.
- Solvent dispenser fitted with a 0,45 µm membrane filter at the outlet.
- Solvent - petroleum ether with a boiling-range of 40 °C to 70 °C, or suitable alternative filtered through a 0,45 µm membrane filter.
- Detergent - liquid, water soluble, commercial grade.
- Calibration standard ISO MTD.
- Ultrasonic bath.

9.3 Calibration procedures
The calibration of the instrument shall be carried out using one of the procedures given in Annex B.

Calibration should be advisably carried out annually or when any repair/change is made to the sensor or whenever results are suspect. It is recommended to check instrument's performance at least each 6 months, by analysing a standard sample of standard reference material (SRM). The results must be consistent with the reproducibility in 9.11.

NOTE Laboratory reference materials (LRM) can be used for instrument's checks if their stability and uncertainty are demonstrated to be equivalent to standard reference materials.

9.4 Preparation of the apparatus for counting
Check that the instrument is set to the calibration numbers for the required size ranges in accordance with the procedure given in the instrument manual.

When the instrument has been switched off, the check procedure shall be carried out prior to use or daily, whichever is less frequent.

If the instrument is permanently switched on, the check procedure shall be carried out at least monthly.

9.5 Preparation of sample before counting
- Remove any visible contamination from the exterior of the bottle.
- Samples which are found on visual examination to contain water or suspended solids (e.g. sludge) likely to affect the performance of the sensor shall be rejected.
- Sample dilution should be avoided as far as possible. However, if in order to comply with the instrument operating parameters dilution is necessary then a liquid of the same type as the sample, or a compatible solvent, may be used. The solvent must be filtered through a 0,45 µm membrane before mixing. Then record the dilution ratio.

NOTE In order to meet the instrument operating parameters, the viscosity of the liquid may be modified by heating. In this case, due regard to the instrument manufacturer's recommendations regarding sensor life should be noted. It should also be established that the precision obtained is acceptable. If sample heating is applied, note the analysis temperature on the Report of results.

9.6 Preparation of sample for counting
- Detach particles coalesced on the vessel by immersing the bottle in an ultrasonic bath for 5 min.
- Shake vigorously by hand for 30 s.
NOTE Suitable mechanical apparatus can be used for shaking several samples at the same time.

- After shaking the sample must be degassed by immersing the bottle in an ultrasonic bath for 5 min.
NOTE Vacuum degassing can either be used, if a sufficient repeatability can be achieved (see 9.10). It is advisable, when using automatic bottle sampling apparatus, that the sensor uptake tube be filled with the test fluid before applying the vacuum.

- Count the sample immediately after the degassing stage. If unable to count within 2 min then the particle suspension shall be maintained by continuous rolling of the sample container.

9.7 Counting procedures
- The counting procedure shall be carried out in accordance with the instrument manufacturer's operating instructions, paying particular attention to the required flow rate and selecting the cumulative mode of counting.
- The sample shall be counted in at least three equal volumes, greater than or equal to 10 millilitres. These three counts should agree within 10 % in the smallest size range (> 4 μm). If this requirement is not attained the sample shall be discarded or re-agitated and re-counted.
NOTE The above minimum count of 20 particles ensures that the counting reproducibility is in accordance with 9.11 of this standard.

- After each sample has been counted remove the sample bottle and flush the system with solvent, filtered through a 0,45 µm membrane filter. It is not recommended that the sensor be dried out after flushing.
- Calculate the particle concentration in each size range from the mean of the counts obtained for each sample aliquot, taking into account the volume of the aliquot and the dilution ratio.

9.8 Report
Report the cumulative number of particles per millilitre of the original sample in at least the following size ranges (see notes):
- ISO code, expressed according to ISO 4406:
• (AA)/BB/CC
• AA is the scale number representing the number of particles equal to or larger then 4 μm(c) per millilitre of fluid
• BB is the scale number representing the number of particles equal to or larger then 6 μm (c) per millilitre of fluid
• CC is the scale number representing the number of particles equal to or larger then 14 μm (c) per millilitre of fluid
NOTE 1 The μm(c) notation of the size ranges means that the measurement is carried out using an automatic particle counter which has been calibrated in accordance with ISO 11171 [4] (ISO MTD calibration).

NOTE 2 When the raw data in one of the size ranges results in an actual particle count of fewer then 20 particles, the scale number for that size range should be labelled as: ≥ scale number (e.g.: a code of 14/12/≥7 signifies that the counting for size range of 14 μm(c) was more than 0,64 and up to and including 1,3 particles, but less than 20 particles were counted).

NOTE 3 When the raw data in one of the size ranges is 'too numerous to count' report a «*» sign (e.g.: a code of */22/7).

NOTE 4 When no counts are detected in one of the size ranges report a «-» sign (e.g.: a code of 1 2/9/-).

NOTE 5 According to ISO 4406, size ranges of 6 µm (c) and 14 µm (c) are equivalent to the old 5 µm and 15 µm particle sizes obtained using the now def unct ISO 4402:1991 [5] method of calibrating automatic particle counters.

- Total number of particles (where p is the diameter in micrometers)
• p > 4 μm(c)
• p > 6 μm(c)
• p > 1 4 μm(c)

- The report shall also include the following:
• method of calibration;
• date of sampling;
• date of analysis.

9.9 Precision
Precision data for this method has been provided by a round robin test performed on three samples of mineral insulating oil taken from transformers with different levels of particle contamination.

NOTE Repeatability and reproducibility are referred to a 95 % confidence level.

9.10 Repeatability
The repeatability has been estimated independently by different laboratories by preparing sample batches of mineral oil and making 7 to 10 replications of the analysis.

When the same laboratory analyses twice the same sample, the difference in each one of the three scale numbers of the ISO code should not exceed 1.

9.11 Reproducibility
The reproducibility has been estimated on the results provided by different laboratories on the same samples.

When different laboratories analyse the same sample, the difference in each one of the three scale numbers of the ISO code should not exceed 2.