A calibrator can only judge an instrument if it is clearly better than the instrument itself. The ratio between the tolerance you are checking and the doubt in your own measurement tells you how much you can trust every pass or fail.
Every pass or fail decision in a calibration lab carries some risk of being wrong. This guide explains how TUR measures that risk, where the 4:1 rule came from and how guard bands keep false accepts low.
What Is the Test Uncertainty Ratio?
The test uncertainty ratio, or TUR, compares the tolerance of the unit under test with the expanded uncertainty of the calibration process used to check it. A higher ratio means the measurement doubt is small compared with the tolerance, so pass and fail decisions are more reliable, a topic tied closely to measurement uncertainty in calibration.
Morehouse defines it as the span of the tolerance divided by twice the 95% expanded uncertainty of the measurement process. For a symmetric tolerance of plus or minus T, that simplifies to T divided by U.
The unit under test tolerance usually comes from its datasheet, such as the classes in pressure gauge accuracy classes. The uncertainty comes from the lab budget, not only from the reference accuracy.
TUR Versus TAR
| Aspect | TAR | TUR |
|---|---|---|
| Full name | Test accuracy ratio | Test uncertainty ratio |
| Denominator | Accuracy specification of the standard | Expanded uncertainty of the whole process |
| Includes | Reference only | Reference, resolution, repeatability, environment |
| Typical result | Looks better on paper | More realistic, usually lower |
| Used in | Older procedures | ANSI/NCSL Z540.3 and ISO 17025 work |
A TAR only compares accuracy figures, while TUR includes every uncertainty contribution in the process. Terms like accuracy and uncertainty are explained in instrument calibration common terms.
Where the 4:1 Rule Came From
Transcat notes that ANSI/NCSL Z540 1 of 1994 required the collective uncertainty of standards not to exceed 25% of the acceptable tolerance. That limit is the same as a 4:1 ratio.
Morehouse explains that a 4:1 ratio gives less than 1% false accept risk and slightly over 1.5% false reject risk, assuming k equal to 2 and 95% end of period reliability. Those assumptions do not hold for every instrument.
The later ANSI/NCSLI Z540.3 standard moved the focus to risk. Its section 5.3 states that the probability of false accept shall not exceed 2%.
5 Smart Rules for Using TUR
Good reference choice raises the ratio, for example a dead weight tester for high accuracy pressure work. Temperature labs choose among the types of temperature calibrators in the same way.
TUR Formula and Example
Guarded acceptance limit = T minus U
Example, transmitter tolerance ± 0.10 bar:
Expanded uncertainty U = 0.02 bar, k = 2
TUR = 0.20 ÷ (2 × 0.02) = 5.00:1
Acceptance limit = 0.10 minus 0.02 = ± 0.080 bar
Meets the 4:1 target
This calculator uses the Morehouse definition, tolerance span divided by twice the expanded uncertainty. Enter the half tolerance T and the expanded uncertainty U in the same units.
Test Uncertainty Ratio Calculator
ILAC G8 Decision Rules and Guard Banding
Pass if the reading lies inside the tolerance.
Acceptance limit shrinks by a guard band w, often equal to U.
Reject only when the reading is beyond tolerance plus w.
ILAC G8 requires the lab to agree the decision rule with the customer and state it in the report, in line with ISO 17025 requirements. With a guard band equal to U, a reading just inside the new limit carries about 2.3% false accept risk at most for a normal distribution.
Guard banding lowers false accepts but raises false rejects, so good instruments are sometimes adjusted without need. A higher TUR reduces both risks at once.
- Reliable pass and fail decisions.
- Small or no guard band needed.
- Fewer false rejects and rework.
- Easier audit justification.
- Ignores where the reading falls.
- Assumes normal distributions.
- Hard to reach for precise devices.
- Says nothing about end of period reliability.
Transmitter specifications often include several error terms, as explained in pressure transmitter accuracy specifications. Combine them carefully before you calculate the ratio.
Field checks with a multifunction process calibrator often give a lower ratio than lab work. Adjust the calibration interval and correction factors with that in mind.
Understanding TUR White Paper
TUR Explained Video
Test Uncertainty Ratio FAQ
It is the ratio of the unit under test tolerance to the expanded uncertainty of the calibration process. A higher ratio gives more reliable pass and fail decisions in the laboratory.
Morehouse defines it as tolerance span divided by twice the expanded uncertainty. For a symmetric tolerance of plus or minus T it simply equals T divided by U.
TAR uses only the accuracy specification of the reference standard in the denominator of the ratio. TUR uses the full expanded uncertainty of the complete measurement process instead.
TUR therefore includes resolution, repeatability, environment and the reference itself. It is usually lower than TAR but gives a much more realistic picture of the actual risk in each decision.
ANSI/NCSL Z540 1 limited the uncertainty of standards to 25% of the instrument tolerance. That limit is exactly equal to a ratio of 4 to 1 between tolerance and uncertainty.
Morehouse notes it gives under 1% false accept risk under typical assumptions. It became a simple rule that laboratories and auditors could apply easily without detailed statistical calculations.
It requires the probability of false accept to stay at or below 2% for calibrations. This is a risk based requirement rather than a fixed ratio between numbers.
Laboratories can meet it with a high TUR or by applying guard bands. Many labs still use the ratio as a quick screening check before detailed risk analysis.
It is a margin taken off the tolerance to form a tighter acceptance limit for decisions. A common choice under ILAC G8 is a guard band equal to the expanded uncertainty.
This lowers the chance of passing an instrument that is actually out of tolerance. It also raises the chance of rejecting an instrument that is actually good.
Use a better reference standard, reduce uncertainty sources or apply a suitable guard band. You can also calculate the actual false accept risk for that measurement using the methods in ANSI/NCSLI Z540.3 or ILAC G8.
Record the decision rule used on the calibration certificate. Discuss the result with the customer before the instrument is released for use in the plant or laboratory.
No, a reading right at the tolerance limit still carries a high false accept risk. Morehouse notes that this risk can reach 50% for a reading exactly at the limit.
Where the reading falls matters as much as the ratio itself. Guard bands are the practical way to handle these borderline results safely and consistently.
Related Articles
- Measurement Uncertainty in Calibration
- ISO 17025 Calibration Laboratory Requirements
- Instrument Calibration Common Terms
- Pressure Transmitter Accuracy Specifications
- Dead Weight Pressure Tester
External References
- Understanding Test Uncertainty Ratio, Transcat
- Why a 4:1 TUR Is Not Enough, Morehouse
- Measurement Uncertainty, Wikipedia
What We Learn Today
- The test uncertainty ratio compares unit under test tolerance with expanded calibration uncertainty, and a higher value means more reliable pass and fail decisions.
- The 4:1 rule came from ANSI/NCSL Z540 1, while Z540.3 now limits the probability of false accept to 2% instead.
- When the ratio is low or readings sit near limits, ILAC G8 guard bands shrink the acceptance zone to control false accept risk.