A structured approach to calibration intervals, uncertainty documentation, and audit-ready traceability across high-criticality test assets. Covers what traceability really means, how to set intervals from data, and how to organize records so they hold up under audit.
What’s Covered
What traceability really means
Traceability is the unbroken chain that connects a working measurement back to a recognized national or international standard, with documented uncertainty at every step. It is not a logo on a certificate. It is the auditable evidence that the number you reported reflects the physical quantity you claimed to measure, within stated bounds.
For traceability to mean anything, three conditions must hold:
- Every link in the chain documents its calibration uncertainty
- The chain reaches a recognized standard, typically a national metrology institute (NRC in Canada, NIST in the US, PTB in Germany, etc.)
- The chain has no gaps, every step in the chain is itself calibrated by an instrument higher in the chain
Types of calibration and what each delivers
| Type | Delivers | Typical use |
|---|---|---|
| Verification check | Confirms the instrument is within tolerance, no adjustment | Routine confidence check, before-and-after sessions, in-house |
| Calibration with adjustment | Brings the instrument back into tolerance, documents as-found and as-left | Periodic calibration, after over-range events |
| Reference / primary | Calibrates against a primary or secondary standard | Reference instruments, calibration of working standards |
| Field check | Quick functional check using a portable reference | Pre-test confirmation, post-event verification |
Verification checks find problems quickly but do not by themselves restore traceability. A laboratory calibration with as-found / as-left documentation is what supports an audit defense.
Calibration intervals: data, not habit
Annual calibration is the default in most quality systems. It is not always the right answer. Intervals should be set by criticality of the channel and by the drift behavior actually observed.
-
Classify by criticality
Channels tied to safety, regulatory compliance, or product release decisions are critical. Channels used for indicative monitoring or troubleshooting are not. Set the interval baseline by class.
-
Capture as-found history
Track each calibration’s as-found result. Stable assets accumulate evidence to extend intervals; drifting assets accumulate evidence to shorten them.
-
Apply trend rules
If three consecutive calibrations are well within tolerance with no adjustment, the interval can extend. If two consecutive calibrations show drift approaching tolerance, the interval must shorten.
-
Document the change
Interval changes should be approved and documented in the quality system. “We extended because nothing happened” is not a defensible interval rationale during an audit.
-
Trigger calibration on events
Over-range events, mechanical shock, environmental excursions outside spec, or significant configuration changes all warrant calibration regardless of the schedule.
Lifetime cost is dominated by calibration, not capital cost. A premium sensor with 24-month intervals and predictable drift can cost less over five years than a budget unit on a 6-month interval that produces ambiguous as-found / as-left records.
Uncertainty: documenting what you don’t know
Every measurement has uncertainty. Calibration certificates that don’t state expanded uncertainty with coverage factor (typically k=2 for ~95% confidence) are not adequate for audit-grade traceability. Uncertainty must be:
- Quantified for the calibration itself
- Propagated through any subsequent measurement chain that uses the calibrated instrument
- Reported alongside any measurement that depends on it
For test programs that report comparative results, uncertainty is the foundation of statistical conclusions. Without it, “instrument A reads higher than instrument B” cannot be distinguished from “instrument A is correct and B is wrong.”
Records: making calibrations audit-ready
Calibration certificates that survive an audit share several properties:
- Identification of the instrument: model, serial number, asset tag
- Date of calibration and recommended next calibration
- Reference standard(s) used and their own calibration status
- Environmental conditions during calibration
- As-found and as-left results across the calibration range
- Stated expanded uncertainty with coverage factor
- Identification of the calibrating organization and accreditation
- Signature or controlled electronic equivalent of the responsible technician
Store certificates centrally, indexed by serial number, and retain them for the lifetime of the instrument plus the audit retention requirement (often 7+ years).
Durham Instruments calibration services
Durham Instruments operates under ISO 9001:2015 certification and provides traceable calibration services through OEM laboratories and qualified partners. Coverage includes:
- Load cells and force transducers
- Torque transducers
- Accelerometers
- Acoustic sensors and reference microphones
- Reference instruments and calibration standards
Submit RMA and calibration request forms via the support / RMA page. For in-house calibration capability, see Interface 1600 Gold Standard and 1800 Platinum Standard reference load cells in the calibration systems and components catalog.
FAQ
What is the difference between traceable and accredited calibration?
A traceable calibration documents an unbroken chain to recognized standards with stated uncertainty. An accredited calibration is performed by a laboratory whose competence has been verified by an accreditation body (e.g., A2LA, SCC), accreditation provides independent assurance that the calibration process meets defined criteria. Most quality systems accept either, but accredited calibrations carry stronger weight in regulated industries.
How often should I calibrate my measurement instruments?
Annual calibration is the most common default. Intervals should be set by channel criticality and drift behavior observed in the as-found history. Critical channels and drifting assets warrant tighter intervals; stable assets with clean histories can support longer intervals when documented properly.
What is expanded uncertainty?
Expanded uncertainty is the calibration uncertainty multiplied by a coverage factor (typically k=2) to produce a confidence interval, usually about 95%. It defines the range within which the true value of the measurand lies with the stated confidence. Calibration certificates without expanded uncertainty are inadequate for most audit and regulatory purposes.
Do I need to recalibrate after an over-range event?
Yes, for any channel where the over-range event exceeded the safe overload rating. “Safe overload” guarantees the device survives without damage, but it does not guarantee the calibration is unchanged. Document the event, recalibrate, and review affected data.
How do I submit equipment to Durham Instruments for calibration?
Submit an RMA / calibration request through the support page or contact Durham Instruments directly. We will issue an RMA, provide shipping instructions, and return the instrument with a traceable calibration certificate.
Need traceable calibration for your instrumentation?
Durham Instruments provides traceable calibration services across load cells, accelerometers, acoustic sensors, and reference instruments. Submit an RMA to start.