ISO 8655Piston-operated volumetric apparatus

ISO 8655 is the international standard family that governs piston-operated volumetric apparatus (POVA) — the mechanical and electronic pipettes, multi-channel pipettes, dispensers, burettes, diluters and titrators on which essentially every QC, microbiology, stability, R&D and clinical lab depends. The 2022 revision (Parts 1 through 9) replaced the 2002 family and aligned terminology, the gravimetric method and the user-side performance-verification programme with modern metrology practice.

The standard is not a regulation, but it is the de-facto reference in pharmacopoeial method validation (USP, Ph. Eur., JP), ISO 17025 accredited labs, FDA Form 483 expectations and ISO 15189 medical-lab accreditation. A QC lab that cannot show ISO 8655-compliant pipette verification will get a finding under any of them.

ISO 8655 separates two performance metrics: systematic error (the bias between the mean delivered volume and the selected volume) and random error (the dispersion of repeated deliveries). The standard publishes maximum permissible errors (MPE) by nominal volume and dispensing volume — for a 1000 µL air-displacement pipette set to 1000 µL the MPE on systematic error is typically ±0.8 % and on random error ≤0.2 %.

Critically, the MPE is a function of dispensed volume, not nominal volume. A 1000 µL pipette set to 100 µL has a much wider relative MPE than the same pipette at 1000 µL. The user must verify performance at the volumes actually used in routine practice, not just at nominal.

Part 6 specifies the gravimetric procedure: dispense a number of replicates (10 is the standard) onto a balance, record each mass, apply the Z-factor to convert mass to volume at the test conditions, and calculate systematic error (mean) and random error (standard deviation).

The Z-factor is the conversion from balance reading (in grams) to volume (in microlitres) accounting for water density at the test temperature, air density (buoyancy), pipette material thermal expansion, and atmospheric pressure. Part 6 publishes Z-factor tables and equations; modern pipette-verification balances (Mettler XPR, Sartorius Cubis) compute Z internally from measured air temperature, water temperature, humidity and pressure.

• Balance must be selected so that the dispensed volume is well above the balance's minimum weight (USP <41>) — a 10 µL pipette test on a balance with 0.1 mg readability is acceptable; on a 1 mg readability balance it is not.
• Water must be Type 2 or Type 3 (ASTM D1193) and equilibrated to room temperature; thermometer at ±0.2 °C resolution.
• Environmental conditions controlled — temperature 20 °C ± 5 °C with stability ±0.5 °C during the test, humidity > 50 % to suppress evaporation.
• Evaporation trap or sequential dispensing into a covered vessel for volumes < 50 µL.
• Pre-rinse the tip three times with the same liquid before the first recorded delivery — critical for hydrophobic plastic tips at low volumes.

ISO 8655-1 distinguishes three programmes the user must operate: (1) calibration — a formal as-found / as-left gravimetric calibration against the manufacturer's MPE, performed annually or per the lab's risk-based schedule; (2) routine performance check — typically quarterly, an abbreviated gravimetric verification at the volumes used in routine work; (3) daily / pre-use check — visual inspection, tip-seal check, and a one-replicate volume verification at the first volume to be used.

The lab's pipette programme must document: which pipettes, what frequency, what method (full Part 6 versus abbreviated Part 7), what acceptance criteria, what happens when a pipette fails. Failed pipettes go through the same as-found / as-left flow as any other instrument — the data they generated since the last successful check must be evaluated for impact.

ICH Q2(R2) and ISO/IEC 17025 require analytical methods to declare a measurement uncertainty budget. The pipette contribution is calculated from the published manufacturer accuracy and precision combined with the verification data and reasoned environmental uncertainties. For a method that pipettes a 100 µL aliquot of a 0.1 mg/mL standard, the pipette uncertainty propagates directly into the assay uncertainty — a poorly-characterised pipette can be the dominant uncertainty contributor.

Air-displacement vs positive-displacement matters here. Air-displacement pipettes have a column of air between the piston and the sample and are sensitive to density, viscosity, vapour pressure and temperature differences between the sample and the air column. Positive-displacement pipettes (where the piston contacts the sample directly) are required for volatile, viscous, dense or foaming samples and have a tighter ISO 8655 envelope.

1. Calibration certificate present but only at nominal volume — routine use is at 10 % of nominal where the MPE is much wider and the actual error unknown.
2. Gravimetric test performed without Z-factor (mass treated as volume) — systematic bias of 0.1–0.3 % unrecognised.
3. Pipettes used past calibration due date with no impact assessment on results generated.
4. Daily check missing — no pre-use verification programme.
5. Multi-channel pipette calibrated only on one channel — inter-channel variation never quantified.
6. Failed-pipette events not flowed through as-found / as-left, no investigation of results generated since last good check.
7. Calibration provider (in-house or contract) not assessed for ISO 17025 scope covering 8655.
8. Air-displacement pipette used for organic solvents or formulation buffers with significant vapour pressure — chronic systematic under-delivery.

• Pipettes are first-class assets with model, serial, channel count, volume range and intended-use volumes.
• Calibration schedule is enforced — overdue pipettes are blocked at the kiosk until the as-left certificate is uploaded.
• Daily pre-use check is a one-tap action at the kiosk; failures route to maintenance and quarantine the pipette automatically.
• Verification results are entered with the test volumes — V5 evaluates systematic and random error against MPE without manual calculation.
• When a pipette fails as-found, V5 builds the impact-assessment ticket: every result generated against that pipette since the last good calibration is listed for evaluation.
• Measurement-uncertainty templates per method reference the assigned pipettes and propagate their published accuracy / precision into the method uncertainty budget.