EURAMET cg-18 Calibration of Weighing Instruments Explained
EURAMET cg-18 is the European guideline for calibrating non-automatic weighing instruments. A cg-18 calibration tests repeatability, errors of indication across the range and eccentricity, and reports each error with its measurement uncertainty, traceable to the SI unit of mass. It gives no legal pass or fail: that is verification under OIML R 76 and the NAWI Directive.
EURAMET cg-18 is the European guideline for calibrating scales and reporting their errors with a measurement uncertainty. This article explains the tests, the uncertainty, the difference with legal verification and what to consider on a vessel. Calibration, adjustment and verification in general are covered in calibration, verification and re-verification of marine scales. This is general information, not legal or accreditation advice.
What EURAMET cg-18 is
EURAMET cg-18 is Calibration Guide No. 18 of EURAMET, the European Association of National Metrology Institutes: "Guidelines on the Calibration of Non-Automatic Weighing Instruments", Version 4.0 (11/2015), prepared by the Technical Committee for Mass and Related Quantities (TC-M) and listed on the EURAMET calibration guidelines page.
It covers the static calibration of self-indicating non-automatic scales: the measurements, the calculation of results, the uncertainty and the content of the certificate. It is guidance, not a mandatory procedure, written so that results from different laboratories are equivalent across Europe. It sets no limit for uncertainty; laboratory and client agree what suits the use. Automatic instruments such as graders and checkweighers fall outside it: they come under Annex VIII (MI-006) of the Measuring Instruments Directive 2014/32/EU where Member States prescribe them, and EURAMET publishes Calibration Guide No. 26 for automatic catchweighing instruments (Version 2.0, 12/2024).
What a cg-18 calibration tests
A cg-18 calibration normally consists of three tests: repeatability, errors of indication across the range and eccentricity. It covers zero to maximum capacity unless a smaller range or individual loads are agreed.
| Test | What it shows | How it is done (section 5) | On the certificate |
|---|---|---|---|
| Repeatability | Scatter when the same load is weighed again | One load, often between half and full capacity, placed at least five times (three times for loads of 100 kg and more) | Standard deviation of a single indication |
| Errors of indication | Difference between indication and the reference value of the load | At least five loads spread over the range, for example zero, a quarter, half, three quarters and full capacity | Error per load, or an error curve, with expanded uncertainty |
| Eccentricity | Effect of load position | About one third of capacity or more, at the centre and four positions towards the corners | Result per position |
| Auxiliary | Buoyancy and convection effects | Air temperature, pressure or altitude, weight temperature | Conditions during calibration |
The procedure should resemble normal use, such as single loads or tare with containers. The scale must be clean, undamaged, warmed up, level and loaded once before testing. Display resolution limits how small the uncertainty can be; see capacity versus readability.
Measurement uncertainty in plain words
Measurement uncertainty is the margin around a measured error within which the true error is expected to lie. If a certificate reports an error of +4 g with an expanded uncertainty of 3 g at 20 kg, the true error is expected between +1 g and +7 g.
It combines the rounding of the display, repeatability, eccentricity and the test weights themselves, including their calibration, drift, air buoyancy and temperature. The certificate states an expanded uncertainty: the combined standard uncertainty multiplied by a coverage factor. In line with the Guide to the Expression of Uncertainty in Measurement (JCGM 100:2008), cg-18 asks for a coverage probability of 95.45 %, which gives a coverage factor of 2 for normally distributed, reliable estimates. Certificates therefore usually state a factor of 2 and about 95 % coverage, and should explain any other factor.
Uncertainty at calibration is not the full uncertainty in daily use. Section 7.4 adds later effects such as temperature changes, drift since calibration, tare operations, creep and off-centre loads, and Appendix G uses this to define a minimum weight: the smallest load that still meets a required relative accuracy, such as 0.1 %.
Calibration versus legal verification
A cg-18 calibration reports errors and uncertainty; it does not decide whether a scale is legally acceptable. Legal verification does that, against the maximum permissible errors of OIML R 76-1 and, in the EU, the NAWI Directive 2014/31/EU.
| Aspect | Calibration (EURAMET cg-18) | Verification (NAWI 2014/31/EU, OIML R 76) |
|---|---|---|
| Question | How large is each error, and how certain? | Are errors within the permissible limits for the class? |
| Basis | Voluntary guidance, laboratories working to ISO/IEC 17025 | EU directive, OIML recommendation, national law |
| Result | Certificate with errors and uncertainty; no legal pass or fail | Pass or fail, with mark, seal or certificate |
| Needed for | Quality systems, audits, customers, research | Regulated uses such as commercial transactions |
| Interval | Chosen by the user | Set nationally |
A calibrated scale is not automatically verified, and a verified scale has no uncertainty statement. An optional statement of conformity to a specification agreed with the client is contractual, not a verification mark. Which regime applies is covered in NAWI and MID directives explained and OIML R51 vs R76.
Traceability: from your scale to the kilogram
Traceability means a result is linked to the SI unit of mass through an unbroken chain of calibrations, each with a stated uncertainty: from a national metrology institute, through a laboratory's reference weights, to the weights on your platform.
- Weights: cg-18 asks for traceable standard weights with suitable density, surface and magnetic properties; weights meeting OIML R 111-1 satisfy this, and their accuracy must suit the scale interval.
- Large platforms: where standard weights run out, for example at 3 000 kg, substitution loads are allowed, but each step adds noticeably to the uncertainty.
- Accreditation: ISO/IEC 17025 accreditation shows a laboratory's competence within a defined scope; accredited certificates name the accreditation body and number.
When you need a cg-18 calibration
You need a traceable calibration when a quality system, auditor, customer or research protocol asks for evidence that weighing results can be trusted.
- ISO 9001: clause 7.1.5.2 of ISO 9001:2015 requires calibration or verification at specified intervals against traceable standards where measurement traceability is required.
- GLP, GMP and ISO/IEC 17025 laboratories: calibration records for balances and scales that affect results.
- Food safety audits: where portion or pack weights are part of a HACCP plan or specification.
- Internal quality: yield and giveaway figures are only as reliable as the scale.
- Research at sea: sampling programmes that need documented weights; keep calibration changes in the audit trail, as described in weighing data logging.
For regulated uses, calibration comes in addition to legal verification, never instead of it.
Preparing for a calibration and reading the certificate
Agree the scope in advance and check the certificate against how the scale is actually used.
- Scope: full or partial range, test loads, and whether uncertainty in use or a minimum weight is wanted.
- Use: single loads, tare with containers or subtractive weighing from a hopper.
- Adjustment: scales normally adjusted before use are adjusted before calibration unless agreed otherwise; ask for results before and after.
- Scale and weights: clean, level and warmed up, zero tracking as in normal use, and safe handling for the weights.
The certificate should show the laboratory and any accreditation, the instrument with capacity, scale interval and place of installation, the date, site and conditions, any adjustment, the errors with expanded uncertainty and coverage factor, repeatability, eccentricity and a traceability statement. At your typical load, the error plus its uncertainty should be small compared with the deviation you can accept.
Calibrating a marine scale: at sea or in port?
A calibration describes the scale under the conditions during calibration, so calibrate a marine scale where and how it is used, or document those conditions.
cg-18 states that calibration is normally performed at the place of use, and that moving the instrument afterwards can invalidate it through differences in gravity, environment and transport stresses. For a vessel this means:
- Calibrate as installed: a scale calibrated ashore and then reinstalled has been moved after calibration.
- Record conditions: port or position, sea state, temperature, and whether alongside or under way. During on-site calibration, vibration and platform inclination become part of the results and the uncertainty, so a calibration under way usually shows more scatter.
- Consider gravity: gravity differs by about 0.5 % between equator and poles; see gravity, latitude and calibration.
A calibration alongside shows the basic performance of the installed scale, not its behaviour in a seaway. Routine checks with test weights at sea, logged with the sea state, fill that gap, as described in the calibration and verification guide.
How WPL approaches this
Every WPL scale, in the marine weighing range and the industrial scales, can be supplied with a EURAMET cg-18 calibration certificate as an option. For on-site calibration and service, see service; when you request a quote, state the range, test loads and place of use. Terms are defined in the glossary, and related guides are on the legal metrology hub.
Frequently asked questions
How often should a scale be calibrated to EURAMET cg-18?
The guide does not set an interval. Base it on how intensively the scale is used, the conditions it works in, how much its errors changed between previous calibrations and what your quality system or customers require. A yearly calibration is a common starting point; shorten the interval after impacts, repairs or large changes between calibrations, and lengthen it only when the history supports that.
Does a calibration also adjust the scale?
No. Calibration only determines the errors and their uncertainty. Adjustment changes the scale so that it reads correctly and is a separate step. cg-18 advises that scales normally adjusted before use are adjusted before calibration unless agreed otherwise, and the certificate should state what adjustment was made. On a verified scale, adjustment can mean breaking seals.
Can a cg-18 certificate replace legal verification?
No. A calibration certificate reports errors and uncertainty but does not assess them against the maximum permissible errors of OIML R 76 and the NAWI Directive, and it carries no verification mark. For regulated uses such as trade or official catch weighing, the scale needs the conformity assessment and verification required by law; a calibration can be kept alongside for quality purposes.
Can a marine scale be calibrated while installed on board?
Yes, and the guide's preference for calibrating at the place of use supports it. The laboratory needs safe access for the test weights, conditions stable enough for the tests and a record of position, sea state and temperature. The certificate then describes the scale as installed, which is more meaningful than a workshop calibration of a scale that is reinstalled afterwards.
Why is the uncertainty on my certificate larger than the scale's readability?
Readability is only the smallest step the display shows. The expanded uncertainty also includes repeatability, eccentricity, the uncertainty of the test weights, and buoyancy and temperature effects, and it is multiplied by a coverage factor, usually 2. It is therefore normally larger than one display step, especially for heavy loads or scales used in less stable conditions.
Sources
- EURAMET Calibration Guide No. 18, Guidelines on the Calibration of Non-Automatic Weighing Instruments, Version 4.0 (11/2015), TC-M
- EURAMET Calibration Guide No. 26, Guidelines on the Calibration of Automatic Catchweighing Instruments, Version 2.0 (12/2024), TC-M
- JCGM 100:2008 Evaluation of measurement data - Guide to the expression of uncertainty in measurement (GUM)
- OIML R 76-1:2006 Non-automatic weighing instruments, Part 1: Metrological and technical requirements - Tests
- OIML R 111-1:2004 Weights of classes E1, E2, F1, F2, M1, M1-2, M2, M2-3 and M3
- Directive 2014/31/EU on non-automatic weighing instruments (NAWI Directive)
- Directive 2014/32/EU on measuring instruments (MID), Annex VIII MI-006 automatic weighing instruments
Written and reviewed by WPL Industries weighing engineers. Technical and regulatory content is checked against the cited sources. Editorial policy