Calibration intervals

The equipment owner sets the calibration interval for the intended use. A due date sets the schedule. Evidence of current suitability also depends on the instrument’s condition. Fluke calibration interval guidance

NIST sets no general interval. Where contracts or regulations leave it open, use accuracy requirements, observed stability and environment, then refine the interval using later calibration results. NIST recommended calibration interval FAQ

In the drift example, error rises from +2 µm after calibration to +5 µm as found at six months. Linear extrapolation gives +8 µm at twelve months against a ±10 µm allowance. One comparison at one point leaves uncertainty, variation and other functions untested, so this projection alone cannot justify extension.

Define the measurement requirement first

Identify the equipment, functions, ranges, and uncertainty needed for the work. A device acceptable for a broad process-control limit may be unsuitable for a tighter product-acceptance decision. Ask what changes in error would matter and what the consequences would be if those changes went unnoticed. NIST’s GMP 11–2019 treats calibration history, uncertainty data, and tolerance limits as prerequisites for interval management. It is a laboratory template that must be adapted to its scope, not a blanket schedule for factory gauges. NIST GMP 11, purpose and prerequisites

Calibration and verification are different activities. The VIM defines verification as supplying objective evidence that specified requirements are met, with measurement uncertainty considered where applicable. A pass/fail verification therefore needs a stated requirement; “checked” without a check method, limit, and outcome is incomplete information. A verification does not automatically replace calibration or provide the same information. JCGM VIM, verification

Choose an initial interval, then review it

ILAC-G24:2022 / OIML D 10:2022 is joint guidance on determining and reviewing intervals. Its initial risk assessment considers required uncertainty, equipment type, drift/wear, use severity, environment, transport, manufacturer advice, checks, and legal requirements. It also calls for review as history accumulates. The PDF cover identifies Edition 2022 (E); this is guidance, not a universal equipment-specific interval table. ILAC/OIML guidance, sections 5–6

Manufacturer specifications can inform the starting choice, but their time basis and conditions need to match actual use. Fluke explains that mechanical wear and electronic drift can be influenced by temperature, humidity, vibration, and shock. An instrument used in conditions harsher than its design assumptions may require a shorter interval. The owner knows its use and history and is responsible for selecting the interval, subject to applicable external requirements. Do not extend a schedule merely because recalibration is inconvenient. Fluke interval ownership and operating conditions

Review methods and their boundaries

These are the five approaches outlined in ILAC/OIML section 6. No method fits every situation. An in-service check device also needs its own stability control; a check cannot detect every failure simply because it is frequent. ILAC-G24 / OIML D 10, review methods

As a bounded research example, NIST Technical Note 2146, published April 2, 2021, applies interval-analysis methods to data from two radiation-survey-meter models and an electronic dosimeter. It also discusses control charts for individual instruments. Those examples demonstrate analysis of actual populations and histories; their conclusions should not be copied into a dimensional-gauge or torque-tool schedule. NIST TN 2146 publication and scope

Preserve the condition found on return

Request results before adjustment or repair where needed to evaluate the elapsed interval, as well as results afterward. An instrument returned in good condition after adjustment does not establish how it performed before that adjustment. NIST recommends retaining both as-submitted and post-calibration accuracy/precision information when validating or refining an interval. Compare the same relevant functions and points over time rather than treating a certificate’s presence as the entire history. NIST interval FAQ, calibration history

If equipment returns significantly out of tolerance, assess earlier measurements that relied on it. NIST’s dimensional-metrology FAQ notes that this can require rechecking parts or other instruments. The affected period and work depend on the failure, previous checks, usage records, and measurement margins; neither automatic acceptance nor automatic rejection of every earlier part is a sound general rule. Shortening the next interval alone does not resolve earlier product decisions. NIST Engineering Metrology Toolbox FAQ

Illustrative drift review

Assume a hypothetical comparator has an internal permitted indication error of ±10 micrometres at one specified point. Its last post-calibration error was +2 micrometres; after six months, the as-found error is +5 micrometres. The observed change is +3 micrometres, or 0.5 micrometre per month over that interval.

A simple linear extrapolation gives +8 micrometres at twelve months. That arithmetic is not sufficient justification for a twelve-month interval: it assumes unchanged linear drift, ignores variation and measurement uncertainty, represents only one point, and does not establish the behavior of other functions. A repair, changed environment, or intermittent fault could invalidate it. A defensible review would examine repeated comparable results, checks, the actual decision rule, and consequences. These numbers are invented to explain the reasoning, not recommendations for any comparator. Basis for using history and uncertainty in interval review

Keep the schedule usable

Record equipment identity, intended use, calibration scope, interval basis, last calibration, due date or usage trigger, intermediate-check arrangements, and who can approve changes. Separate a planned due date from evidence that equipment remains suitable after damage, a failed check, or a change in use. NIST GMP 11 requires tracking and scheduling and rejects inadequate funding or poor due-date tracking as technical reasons for extending intervals. Its detailed legal-metrology provisions apply within that context. NIST GMP 11, tracking and interval adjustment

Related reference

Inspection and quality documents