How to use
Propagate flow measurement uncertainty into delivered quantity
Start from the field measurement equation and calibration basis. Confirm that flow is stable for the entered duration, that rate components use the same actual or standard reference condition, and that density is appropriate before extending volume into mass flow.
- Enter the indicated rate, unit basis, meter range, and fluid condition exactly as observed.
- Apply only the signed bias correction supported by calibration and keep correction uncertainty as a separate component when required.
- Interpret distributions and coverage before entering resolution, calibration, repeatability, and fluid-condition effects as standard components.
- Set the coverage factor, observation time, and matching density while recording excluded time-base and density uncertainty.
- Review component ownership and the flow interval first, then propagate it into volume and use a fuller covariance-aware budget for governed quantity or mass-flow decisions.
Flow-precision fundamentals
Indication correction
Moves the reading toward the calibrated estimate; it does not shrink uncertainty.
Flow repeatability
Spread from repeated readings at the same stable flow condition.
Fluid-condition effect
Allowance for temperature, density, viscosity, or reference-condition influence.
Volume propagation
For fixed time, flow uncertainty multiplies directly by elapsed seconds.
Mass-rate extension
Density converts corrected volume rate to mass rate but brings its own uncertainty, not modeled here.
Result interpretation
Carry one corrected flow interval into volume without hiding assumptions
Corrected flow is the measurement estimate after the signed bias adjustment. Expanded uncertainty defines the displayed lower and upper rate limits. Multiplying both estimate and half-width by the same elapsed time produces a delivered-volume interval under the fixed-rate propagation assumption.
Longer time increases absolute delivered volume and absolute half-width proportionally; it does not automatically reduce relative systematic uncertainty. Near zero, report absolute rather than relative uncertainty. The mass-rate extension uses fixed density and should not be presented as a complete mass uncertainty statement when density or fluid condition varies.
Correct the flow rate before expanding and propagating uncertainty
The indicated flow is converted to m³/s and corrected. Independent standard components are combined by root-sum-square and expanded with the selected coverage factor.
Zero-flow behavior
Relative uncertainty becomes unstable near zero; use absolute flow uncertainty and meter zero specifications.
Time correlation
Multiplying one uncertainty by total time assumes a stable systematic interval, not independent random errors at every sample.
Density coupling
Mass flow requires uncertainty in density when temperature, composition, or pressure varies materially.
Meter operating range
Relative accuracy often worsens near the lower range
Resolution, zero stability, pulse quantization, leakage, and threshold behavior can dominate at low flow. Dividing a fixed absolute component by a small reading makes relative uncertainty grow rapidly even when the instrument has not changed.
Confirm the meter's calibrated range, specified turndown, zero check, and minimum pulse or velocity before interpreting percentages. Report absolute limits near zero and consider a lower-range instrument or longer governed collection method when the expected rate approaches the usable minimum.
Installation effects
Calibration does not remove field-profile error
Elbows, valves, swirl, insufficient straight run, orientation, vibration, deposits, and partially filled pipe can change meter response after laboratory calibration. Fluid properties and installation geometry may also differ from certificate conditions.
Compare the field arrangement with manufacturer requirements and the calibration setup. Add a justified installation component or correction when evidence supports it; if the effect is unknown and potentially material, investigate or change the measurement method rather than assuming laboratory uncertainty covers it.
Mass-flow extension
Density introduces another uncertainty budget
The displayed mass rate multiplies corrected volume rate by one fixed entered density. A professional mass-flow or total-mass statement must also propagate density uncertainty from temperature, pressure, composition, phase, and the property method used.
Volume rate and density can be correlated when both depend on the same pressure or temperature sensors, so independent root-sum-square may be inappropriate. For compressible gas or changing mixtures, use a condition-aware mass-flow equation rather than treating density as a constant label.
Visual reading guide
Compare component ownership with time propagation
The primary view centers the coverage interval on corrected flow and compares absolute standard components on one rate basis. The supporting view carries both rate estimate and expanded half-width through the entered duration to show delivered-volume propagation.
Changing coverage factor widens the interval without changing variance shares; changing time scales volume and its half-width together; changing density affects mass rate but not volume uncertainty. Neither view includes covariance, profile bias, time-base uncertainty, or changing flow unless explicitly represented in inputs.
Detailed calculation process
Qc = Qi(1 + c); uc = √Σuᵢ²; UQ = kuc; V = Qct; UV = UQt; ṁ = ρQc
Qi and Qc are indicated and corrected volumetric rates, c is bias correction, UQ is expanded flow uncertainty, V is volume, and ṁ is mass rate.
| Qc | corrected volume rate | m³/s |
| uc | combined standard uncertainty | m³/s |
| UV | propagated volume uncertainty | m³ |
| ṁ | density-derived mass rate | kg/s |
Propagation check:
Flow metrology evidence
Retain calibration, installation and fluid-state records
Keep raw indication, signed correction and uncertainty, actual or standard unit basis, meter range and turndown, resolution or pulse value, calibration coverage, repeat runs, zero checks, drift history, installation geometry, fluid temperature and pressure, density source, and synchronized timing.
Use components from the same measurand, condition, and period; avoid duplicating calibration or fluid effects under several labels. Reconcile corrected rate to the certificate model, volume to a totalizer or gravimetric reference when available, and all timestamps to the entered observation duration.
Limits and exclusions
What this independent-component model omits
The model excludes covariance, nonlinear meter curves, detailed pulse statistics, time-base uncertainty, density uncertainty, compressibility, multiphase behavior, installation-profile correction, transient integration, leakage, and correction uncertainty unless explicitly included. It assumes stable flow and symmetric components.
Accordingly, do not use the interval alone for custody transfer, regulated emissions, dosing compliance, or other governed quantities when omitted effects are material. Apply the approved measurement equation, sampling method, traceability chain, and decision rule for those uses.
Flow metrology glossary
Terms used in the propagated interval
Turndown ratioUsable maximum-to-minimum flow range.
Zero stabilityMeter behavior at no flow.
Installation effectBias caused by field geometry or profile.
Pulse resolutionVolume represented by one pulse.
Coverage intervalEstimate plus and minus expanded uncertainty.
PropagationTransfer of input uncertainty into a derived result.
Time-base uncertaintyUncertainty in elapsed-duration measurement.
Density couplingDependence of mass rate on fluid state.
Worked cases
Two flow uncertainty decisions across operating range
One-hour batch transfer
Inputs: 250 L/min with signed correction, four standard components, k = 2 and 60 minutes.
Calculation: correct the rate, combine independent components and multiply the rate interval by 3,600 seconds.
Decision: report both corrected volume and interval; add density uncertainty before reporting mass quantity.
Low-flow dosing check
Inputs: 5.00 L/min indication, −1% correction, 0.10 L/min resolution, 0.5% calibration, 0.4% repeatability, 0.3% fluid effect, k = 2 and 30 minutes.
Calculation: corrected flow is 4.95 L/min. Combined standard uncertainty is about 0.0454 L/min, expanded uncertainty about 0.0908 L/min, and the interval about 4.859 to 5.041 L/min. Delivered volume is 148.5 ± 2.72 L.
Decision: the interval straddles a 5.0 L/min target and resolution is material. Verify zero and installation behavior or use a lower-range governed dosing method before conformity decisions.
Important note
For custody transfer, regulated emissions, or dosing, use the applicable metrology standard and include time, density, reference-condition, and installation uncertainties.
Frequently asked questions
Flow precision questions
Why correct before combining uncertainty?
The reported estimate should include a supported signed bias correction, while uncertainty describes remaining doubt after correction. Applying percentage components to the uncorrected rate can place them on the wrong magnitude basis.
Can manufacturer accuracy be used directly?
Only after identifying whether it is a limit, standard uncertainty, or expanded uncertainty and interpreting its distribution, coverage, range, and operating conditions. Convert it to the same absolute standard-rate basis as other components.
Why is density absent from volume uncertainty?
Density is unnecessary for volumetric quantity because volume equals rate multiplied by time. It becomes essential for mass rate or total mass, where density value, uncertainty, condition, and possible covariance must be included.
Does longer time reduce volume uncertainty?
Not in this fixed-rate propagation model. Estimate and systematic half-width both scale with duration, leaving relative uncertainty unchanged. Averaging random samples can behave differently but requires a time-series and correlation model.
What happens near zero flow?
Relative uncertainty becomes unstable and resolution, pulse quantization, leakage, and zero behavior can dominate. Report absolute limits, confirm turndown and zero checks, and consider a more suitable range or method.
Should installation effects be added?
Yes when field geometry, orientation, profile, fluid condition, or straight-run arrangement differs materially from calibration. Use evidence or manufacturer guidance for the component; investigate rather than inventing a convenient allowance.
Can components be correlated?
Yes. Temperature, pressure, correction, density, and repeat data can share sensors or causes. When covariance is material, use sensitivity coefficients and covariance terms instead of the independent root-sum-square calculation.
Does the interval prove delivered quantity?
Only under the entered stable-rate, duration, unit-basis, and fixed-density assumptions and with a valid measurement model. It does not prove traceability, field installation adequacy, or absence of omitted bias.
When is a regulated method required?
Use the approved regulated method for custody transfer, emissions, dosing, tax, safety, or other governed measurements. Such methods may prescribe sampling, calibration, uncertainty, rounding, records, and decision rules beyond this calculator.