What a calibration certificate does and does not say

A calibration certificate is a research record of a measurement that already happened. The application is to decide whether that record still supports the volume you are about to publish or to put into a standard curve. The limit of the claim is the page itself: identity, date, reference, volumes, and errors. Anything the pipette suffered after that date is outside the sentence the certificate can speak. How you produce a volume worth certifying is in accurate micropipetting technique.

What the page is for

Read the identity first. The serial or the laboratory mark on the pipette should be the mark on the page. A certificate in a folder, unmatched to a pipette, is a story about a different object. Read the date. The date is when the errors were true under the conditions of the test. It is not a prediction.

Read the reference. A useful certificate says what it compared the pipette with: a stated method, a manufacturer specification, or the maximum permissible errors specified for that class. For pipettes, those class limits are the subject of ISO 8655-2:2022. This application will not copy them onto the bench. If the certificate does not name the specification, you cannot know what "pass" meant. ISO 8655-1:2022 is the terminology and user-recommendation part of the same series. It defines how to talk about the apparatus. It is not itself a filled-in certificate, and it is not evidence that a particular pipette is accurate today.

Read the volumes. A test at the nominal volume does not automatically describe a setting near the bottom of the range, where relative error is often wider. If your work lives at that low setting, a certificate that never visited it does not cover the decision. Read the errors. Systematic error is the shift of the mean. Random error is the scatter. You want both. A single word, pass, without the errors, cannot tell you how close the mean sat or how wide the repeats were.

Read the conditions if they are printed: temperature, tip, mode. A forward-pipetting water test is the usual air-displacement case. A filter tip or a low-retention tip used every day, if it was not the selected tip on the certificate, is a different apparatus. ISO 8655-2 treats the pipette as complete with its selected tip. The certificate can say that only if the tip is identified.

NIST's Office of Weights and Measures and BIPM are why the milligram and the microlitre on the page belong to a shared system. They do not sign your certificate.

What the page cannot do

It cannot see a drop. If the pipette fell after the date, the shaft and the piston may have moved. Check it before it returns to service. The old pass is not a cushion.

It cannot see solvent. Liquid that entered the nose can swell a seal that was fine on the calibration bench. The certificate does not absorb that chemistry.

It cannot see heavy use. A pipette used hard, every working day, for a long stretch such as six months, can leave the errors that were printed even though nobody dropped it and nobody scheduled a moral failure. Six months is an illustration of elapsed heavy use, not a law that every laboratory must adopt as the interval. The interval is set locally from usage, the liquids, and the manufacturer recommendation. The certificate simply stops being evidence once that local rule, or an event, says the conditions have changed.

It cannot convert itself into a glycerol method, an ethanol method, or a robot method. Those are different measurements. An interim check, in checking a pipette between calibrations, compares the pipette you have now with the specification you claim. Water density in that check is about 0.998 grams per millilitre near room temperature and must come from a table at the measured temperature.

A way to use the document, and the branch when life has moved on

Before a critical series, confirm the identity, the date, the volumes, and that no drop or solvent event is unrecorded. If those are intact and the date is inside your local interval, the certificate supports the water-like, same-tip use it describes. Perform the work in the mode it assumed. Forward pipetting for aqueous liquids. Do not reverse-pipette and then cite a forward certificate.

Branch if any event happened. Tag the pipette, check or service it, and do not cite the old page in a notebook as if it were current. Branch if the volume you need was not on the certificate. Either check that volume yourself, with a balance that can see it, or do not claim the certificate covers it. Branch if the tip family changed. Check the pair. A passing old page plus a new leaky box is a leak.

When a new certificate arrives, read the errors before you file it. A fail that nobody noticed stays in service.

Failure modes when people use the page as a charm

Citing a certificate after a known drop is the serious failure. Using a nominal-volume pass to bless a tiny volume is the quiet one. Photocopying one pipette's page into another pipette's drawer is a records failure. The measurement was never that transitive. The page stays true about the past.

Safety and research limits

Certificates do not assess biological risk. Follow the institutional assessment. This application is about reading a volume record. It is not a diagnostic validation.

If the laboratory is hot, or far above the altitude of the calibration site, the certificate remains a true report of where it was produced. Check locally before you treat it as a description of this bench. Write the water temperature on that check.

What to ask for when a document is part of the enquiry

Ask for identity, date, volumes you actually use, systematic and random error, the specification used as the reference, the test temperature, and the tip family. Say whether solvent exposure is part of the life the document must be relevant to, so service recommendations are not quoted from a water-only story. Pipettes are in the scientific instruments catalogue. Tips that must match the certificate are in the laboratory consumables catalogue. Send the list with the quotation request.