Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC Guide 115 — Uncertainty of Measurement
Evaluation and Expression of Measurement Uncertainty in Electrical Testing
1. Introduction to Measurement Uncertainty in IEC
IEC Guide 115 provides essential guidance on the evaluation and expression of measurement uncertainty for electrical and electronic testing. It harmonizes with the internationally recognized ISO/IEC Guide 98-3 (GUM — Guide to the Expression of Uncertainty in Measurement) while adding specific considerations relevant to IEC product standards and conformity assessment.
Every measurement result is incomplete without a stated uncertainty. Guide 115 provides the framework to ensure your uncertainty statements are both rigorous and practically applicable in a testing laboratory environment.
The guide addresses the full measurement chain: from defining the measurand and identifying influence quantities, through Type A and Type B evaluation methods, to calculating combined and expanded uncertainty. It also covers the specific challenges of correlated input quantities and non-linear measurement functions.
2. Type A and Type B Evaluation Methods
Guide 115 distinguishes between Type A evaluation (based on statistical analysis of repeated observations) and Type B evaluation (based on scientific judgment using all available information about the measurement process). Both approaches produce standard deviations that are treated identically when calculating combined uncertainty.
| Evaluation Type | Basis | Typical Examples | Degrees of Freedom |
|---|---|---|---|
| Type A | Statistical analysis of repeated measurements | Repeatability, reproducibility | n – 1 (directly calculated) |
| Type B — Calibration | Calibration certificate data | Reference standard uncertainty | Effective df from certificate |
| Type B — Resolution | Instrument digital resolution | DMM last digit, scale division | Infinite (rectangular dist.) |
| Type B — Environmental | Temperature, humidity specifications | Drift over temperature range | Estimated from experience |
| Type B — Method | Test method limitations | Imperfect alignment, loading effects | Engineering judgment |
A common mistake is underestimating Type B contributions. Engineers often focus on instrument calibration uncertainty while neglecting larger contributions from test setup variation, operator technique, and environmental factors. A thorough uncertainty budget should include all significant sources.
3. Practical Uncertainty Budget Construction
Building a measurement uncertainty budget requires careful identification of all influence quantities and their contribution to the total variability. Guide 115 recommends a structured step-by-step process: define the measurand, identify uncertainty sources, quantify standard uncertainties, compute combined uncertainty using the law of propagation of uncertainty, determine effective degrees of freedom (Welch-Satterthwaite formula), and calculate expanded uncertainty using the appropriate coverage factor k (typically k=2 for 95% confidence).
From an engineering design perspective, understanding measurement uncertainty is critical when defining test limits and guard bands. If a product must comply with a specified limit, the test decision must account for the measurement uncertainty — a product whose measured value falls within the guard band around the limit may still pass or fail depending on the confidence level required.
Designing products with a 15-20% margin below the regulatory limit (rather than at the limit) eliminates most conformance testing risks arising from measurement uncertainty. This “design guard band” approach costs little in early design stages but saves significant rework expense.
The guide also addresses the reporting of uncertainty in calibration certificates, test reports, and declarations of conformity. Uncertainty statements should include: the expanded uncertainty U, the coverage factor k, and the approximate confidence level.
4. Frequently Asked Questions
Q: Do I need to calculate uncertainty for every test I perform?
A: Accredited laboratories are required to maintain uncertainty budgets for all test methods. For routine testing, the pre-established uncertainty budget can be referenced rather than recalculated.
A: Accredited laboratories are required to maintain uncertainty budgets for all test methods. For routine testing, the pre-established uncertainty budget can be referenced rather than recalculated.
Q: What is the difference between accuracy and uncertainty?
A: Accuracy describes how close a measurement is to the true value, while uncertainty quantifies the dispersion of values that could reasonably be attributed to the measurand.
A: Accuracy describes how close a measurement is to the true value, while uncertainty quantifies the dispersion of values that could reasonably be attributed to the measurand.
Q: How do I handle correlated input quantities?
A: Correlation occurs when two uncertainty sources share a common cause. Guide 115 recommends using covariance terms in the combined uncertainty formula or, practically, avoiding correlation through experimental design.
A: Correlation occurs when two uncertainty sources share a common cause. Guide 115 recommends using covariance terms in the combined uncertainty formula or, practically, avoiding correlation through experimental design.
Q: Is a coverage factor of k=2 always appropriate?
A: k=2 (≈95% confidence) is conventional for most industrial testing. For safety-critical applications, k=3 (≈99.7%) may be required, while k=1 (≈68%) suffices for internal screening.
A: k=2 (≈95% confidence) is conventional for most industrial testing. For safety-critical applications, k=3 (≈99.7%) may be required, while k=1 (≈68%) suffices for internal screening.
