Uncertainty Budgets


Uncertainty Budgets in ISO/IEC 17025: What They Are and How to Determine Them

ISO/IEC 17025:2017 – Competence of testing and calibration laboratories

What Are Uncertainty Budgets?

Uncertainty budgets are structured, itemized tools (usually tables or spreadsheets) that identify, quantify, combine, and document all significant sources of uncertainty affecting a measurement result. They are the practical way laboratories calculate and report the combined standard uncertainty (uc) and the expanded uncertainty (U) required by ISO/IEC 17025:2017 (clause 7.6).

Measurement uncertainty quantifies the doubt around a reported value — the range within which the true value is reasonably expected to lie (e.g., 100.0 ± 0.5 g at 95% confidence). It accounts for all potential error sources, not just random variation.

“Uncertainty of measurement is a non-negative parameter characterizing the dispersion of the values attributed to a measurand, based on the information used.”
— Guide to the Expression of Uncertainty in Measurement (GUM), JCGM 100:2008

Why Are They Required?

  • Clause 7.6.1: Identify all contributions to uncertainty using appropriate methods.
  • Clause 7.6.2: Calibration labs — always evaluate rigorously.
  • Clause 7.6.3: Testing labs — rigorous where possible; otherwise, estimate based on validation, experience, etc.
  • Clause 7.8: Report uncertainty in calibration certificates (always) and in test reports when needed for interpretation or conformity decisions.

Typical Uncertainty Budget Table Example

SourceDescriptionu_i (standard uncertainty)TypeDistributionContribution to uc
Reference standard From calibration certificate 0.002 g B Normal 0.002 g
Repeatability Std dev from repeats 0.015 g A Normal 0.015 g
Resolution Instrument readability 0.005 g B Rectangular 0.0029 g
Combined standard uncertainty (uc) √(sum of squares) 0.016 g
Expanded uncertainty (U, k=2, ≈95%) k × uc ±0.032 g

How to Determine an Uncertainty Budget – Step-by-Step

  1. Define the measurand and measurement equation (e.g., mass = indication + corrections).
  2. Identify all sources (use fishbone diagram/checklists): reference standard, repeatability, resolution, environment, bias, etc.
  3. Quantify each u_i: Type A (statistical from repeats), Type B (certificates, specs, rectangular → ÷√3).
  4. Apply sensitivity coefficients (usually 1) and distributions.
  5. Calculate combined uc using root sum of squares (RSS): uc = √(Σ(c_i × u_i)²)
  6. Expand: U = k × uc (k=2 for ~95% confidence)
  7. Document in a table, reference sources, review periodically.

Key Takeaways

  • Budgets must be specific to each method or calibration point.
  • Common pitfalls: omitting reference uncertainty, underestimating Type B, ignoring environment.
  • Resources: GUM (JCGM 100), EURACHEM/CITAC Guide, accreditation body templates.
  • Goal: Reliable, comparable, defensible measurements — core to ISO 17025 accreditation.

 

Related Articles