The difference lies in the origin of the value. A primary standard has its value determined directly, from the definition of the unit or from well-established physical principles, by means that do not require calibration against another standard of the same quantity. A secondary standard has its value assigned by calibration against a primary one, so its uncertainty includes that of the primary standard and that of the comparison.
What it looks like in practice
Primary standards are usually found at national metrology institutes and in a few highly specialized laboratories, because their construction and operation are complex. Examples vary widely by quantity: systems based on quantum effects for voltage and electrical resistance, high-quality pressure balances with characterized dimensions and masses for pressure, or fixed-point cells for temperature. Secondary standards are the ones disseminated most widely.
Why they are distinguished
- Traceability: the primary standard is the root of the chain; the secondary depends on it.
- Uncertainty: that of the primary is lower, and each step down adds to it.
- Availability: primary standards are scarce, so they are used to calibrate secondary standards, which in turn calibrate working standards.
- Maintenance: the primary standard requires checks and international comparisons to support its value.
A clarification on the term
The classification does not depend on what the device looks like, but on how its value is obtained. The same type of device can be primary in one laboratory, if it is characterized independently, and secondary in another, if it is calibrated against a reference. Nor should primary be confused with accurate in absolute terms: what defines it is the method by which it is established and its place in the hierarchy.
When it matters
When reviewing a certificate, knowing the level of the standard used makes it possible to verify that its uncertainty was appropriate for the instrument calibrated and that the traceability chain is complete.