Meter Proving Fundamentals

1. Purpose of Meter Proving

Every flow meter has inherent inaccuracies from manufacturing tolerances, installation effects, and operating condition variations. Meter proving quantifies these inaccuracies by comparing the meter's output to a reference standard of known accuracy. The resulting meter factor is then applied to all subsequent meter readings to correct for the systematic error.

Proving is required at custody transfer points because the financial value of the measured fluid justifies the cost of maintaining traceable measurement accuracy. A 0.1% error on a station flowing 100 MMSCF/d at $3.50/MMBTU represents approximately $130,000 per year of undetected measurement bias.

Without Proving

Unquantified Error

Meter reads within manufacturer spec (typically 0.5-2.0%) but actual bias is unknown.

With Proving

Corrected Measurement

Meter factor corrects bias to within proving uncertainty (typically 0.02-0.05%).

Financial Impact

$130k+ per 0.1%

On a 100 MMSCF/d station at $3.50/MMBTU, every 0.1% matters.

Traceability chain: The prover's reference volume is calibrated against a primary standard (water draw, gravimetric), which is traceable to NIST. This unbroken chain of calibration provides the legal basis for custody transfer measurement accuracy.

2. Meter Factor

Meter Factor Definition: MF = V_prover / V_meter Where: MF = Meter Factor (dimensionless) V_prover = Volume indicated by prover (reference) V_meter = Volume indicated by meter under test Applied as correction: V_corrected = V_meter_indicated × MF Example: Prover indicates: 100.12 gallons (calibrated reference) Meter indicates: 100.00 gallons (under test) MF = 100.12 / 100.00 = 1.0012 All subsequent meter volumes are multiplied by 1.0012 to correct for the meter's slight under-reading. For K-factor (pulse) meters: K = Pulses / V_prover MF = K_base / K_proving (ratio to baseline K-factor)

Meter Factor Interpretation

MF Value	Meaning	Typical Range
MF = 1.0000	Meter reads exactly correct	Ideal (rare)
MF > 1.0000	Meter under-reads (correction adds volume)	Common for turbine meters
MF < 1.0000	Meter over-reads (correction reduces volume)	Common for new meters
MF > 1.02 or < 0.98	Significant deviation, investigate	May indicate problem

3. Prover Types

Prover Type	Volume	Min Runs	Typical Repeatability	Application
Pipe Prover	10-300 bbl	3	0.01-0.02%	Large liquid stations, permanent installation
Compact Prover	0.01-1 bbl	5	0.02-0.04%	Smaller stations, portable proving
Master Meter	N/A	5	0.02-0.05%	Gas metering, transfer standard
Tank Prover	100-5000 gal	3	0.01-0.02%	Primary calibration, water draw

Compact vs. pipe provers: Compact provers use a small piston and many passes to accumulate sufficient volume. Because each pass has higher relative uncertainty, API requires minimum 5 runs (vs. 3 for pipe provers). The advantage is portability and smaller footprint.

4. Repeatability Criterion

API MPMS Chapter 4 Repeatability: Repeatability (%) = (MF_max - MF_min) / MF_avg × 100 Criterion: Repeatability must not exceed 0.05% Example: Run 1: MF = 1.0012 Run 2: MF = 1.0015 Run 3: MF = 1.0010 Run 4: MF = 1.0014 Run 5: MF = 1.0011 MF_avg = 1.00124 MF_max = 1.0015, MF_min = 1.0010 Range = 0.0005 Repeatability = 0.0005 / 1.00124 × 100 = 0.0499% Result: PASS (0.0499% < 0.05%) MF to apply: 1.00124 (mean of all 5 runs)

If repeatability exceeds 0.05%, the proving is invalid. The technician must identify and correct the cause before re-proving. Common causes include air or gas entrainment, temperature changes during proving, prover valve leaks, meter malfunction, or insufficient flow stabilization.

5. Outlier Detection

Statistical outlier detection identifies individual proving runs that are significantly different from the others. The Grubbs test is the most commonly used method in custody transfer metering. It compares each value's deviation from the mean to the standard deviation, testing against critical values for the sample size.

Grubbs Test: G = |x_suspect - x_mean| / s Where: G = Grubbs test statistic x_suspect = the value being tested x_mean = sample mean s = sample standard deviation Decision rule: If G > G_critical (from table for N and alpha=0.05), the value is an outlier and may be excluded. Critical values (alpha=0.05, two-sided): N=3: G=1.155 N=5: G=1.715 N=7: G=2.020 N=10: G=2.290 Important: - Only remove outliers with identifiable physical cause - Re-prove after removing outlier run - Never remove more than one outlier from a set - Document reason for exclusion

6. Proving Procedure

Stabilize conditions: Allow flow rate, temperature, and pressure to stabilize before starting. Temperature should not vary more than 1 deg F during a run.
Verify prover: Check prover calibration certificate, valve seal integrity, and detector switch operation before proving.
Run proving: Perform minimum required number of consecutive runs without interruption. Record meter pulses and prover volumes for each run.
Check repeatability: Calculate repeatability after each run. If criterion is met with minimum required runs, proving is valid.
Calculate meter factor: Average all accepted run meter factors. Apply temperature, pressure, and compressibility corrections as applicable.
Document: Record all data, conditions, meter factor, and any anomalies. Retain records per contract and regulatory requirements.

7. Volume Corrections

Volume Correction Factors: V_standard = V_indicated × Ctl × Cpl × Ctm × Cpm × MF Where: Ctl = Temperature correction for liquid (prover) Cpl = Pressure correction for liquid (prover) Ctm = Temperature correction for meter Cpm = Pressure correction for meter MF = Meter factor from proving For gas metering: V_standard = V_actual × (P/P_base) × (T_base/T) × (1/Z) × MF Where Z = compressibility factor (AGA 8 / SGERG-88)

8. Proving Frequency

Application	Typical Frequency	Trigger Events
Custody transfer (liquid)	Monthly or per batch	Product change, meter maintenance, MF shift >0.0025
Custody transfer (gas)	Monthly to quarterly	Calibration, flow profile change, maintenance
Allocation metering	Quarterly to annually	Significant flow or composition change
Check metering	Annually	When discrepancy with primary meter exceeds tolerance

9. Troubleshooting Poor Repeatability

Symptom	Possible Cause	Solution
Repeatability > 0.05%	Air/gas entrainment	Vent prover, check for leaks upstream
Systematic drift in MF	Temperature change during proving	Stabilize conditions, insulate lines
Random scatter	Prover valve leaking	Test prover valve leak rate
One run different	Transient flow disturbance	Exclude outlier (with documentation), re-run
All MFs shifted from previous	Meter wear or contamination	Inspect meter internals, clean, recalibrate

10. Industry Standards

Standard	Title	Relevance
API MPMS Ch. 4	Proving Systems	Prover types, procedures, repeatability
API MPMS Ch. 5	Metering	Meter types and specifications
API MPMS Ch. 12	Calculation of Petroleum Quantities	Volume correction factors
API MPMS Ch. 13	Statistical Aspects of Measuring and Sampling	Outlier detection, data analysis
AGA Report No. 7	Measurement of Gas by Turbine Meters	Turbine meter proving requirements
AGA Report No. 9	Measurement of Gas by Ultrasonic Meters	USM proving and diagnostics
API MPMS Ch. 21	Flow Measurement Using Electronic Metering Systems	Flow computer requirements

MF = V_prover / V_meter

≤ 0.05%

API MPMS Ch. 4 · Ch. 12

Contents

Use this guide when you need to: