1. Overview & Purpose
Meter proving is the process of verifying a flow meter's accuracy by comparing its indicated volume against a known reference volume. The reference device -- the prover -- has a precisely calibrated volume traceable to national standards (NIST). Proving determines a meter factor (MF) that corrects the meter reading to true volume for custody transfer billing.
Image: Meter Proving System Overview
Schematic showing meter station with turbine meter, bidirectional pipe prover in U-loop configuration, detector switches, and four-way diverter valve.
Custody transfer
Billing accuracy
Provers ensure measurement accuracy within 0.02-0.05% for crude oil, NGL, and refined product sales.
Regulatory
API MPMS compliance
API Manual of Petroleum Measurement Standards Chapter 4 defines proving requirements.
Financial impact
Revenue protection
A 0.1% meter error on 100,000 BPD crude at $70/bbl costs $7,000/day or $2.55M/year.
Traceability
NIST standards
Prover base volume is waterdraw-calibrated with traceability to national measurement standards.
Why Meters Need Proving
Flow meters are precision instruments, but their accuracy drifts over time due to mechanical wear, deposits, fluid property changes, and operating condition variations. Proving quantifies this drift as a meter factor and corrects for it:
- Turbine meters: Blade wear, bearing friction changes, and viscosity effects alter the K-factor (pulses per unit volume) over time.
- Positive displacement meters: Clearance wear allows slip flow, changing the displacement per revolution.
- Coriolis meters: Zero drift, tube coating, and density calibration changes affect mass and volume accuracy.
- Ultrasonic meters: Transducer fouling, profile changes, and electronics drift alter velocity measurement.
Proving vs. calibration: Proving determines a correction factor (meter factor) applied to meter readings in service. Calibration adjusts the meter internals to restore original accuracy. Most custody transfer meters are proved regularly (daily to monthly) and calibrated less frequently (annually).
API MPMS Chapter 4 Framework
API MPMS Chapter 4 covers proving systems in several sections:
| Section |
Title |
Coverage |
| Ch. 4.1 |
Introduction |
Proving fundamentals, terminology |
| Ch. 4.2 |
Displacement Provers |
Pipe provers (uni/bidirectional), sphere provers |
| Ch. 4.3 |
Small Volume Provers |
Compact piston provers, design and operation |
| Ch. 4.5 |
Master Meter Provers |
Using a calibrated master meter as reference |
| Ch. 4.7 |
Field Standard Test Measures |
Volumetric test measures (tank provers) |
| Ch. 4.8 |
Operation of Proving Systems |
Procedures, data handling, quality checks |
2. Prover Types
Three main categories of displacement provers are used in the petroleum industry, each with distinct advantages for different applications.
Bidirectional Pipe Prover
A bidirectional pipe prover consists of a U-shaped pipe loop with a displacer (sphere or piston) that travels in both directions between two pairs of detector switches. A four-way diverter valve reverses flow direction through the prover after each pass.
Image: Bidirectional Pipe Prover Schematic
Plan view showing U-loop pipe, detector switch pairs at each end, four-way diverter valve, flow connections to meter run, and sphere launcher/receiver.
Bidirectional Prover Advantages:
- Two data points per round trip (forward + reverse)
- Shorter total proving time than unidirectional
- No sphere return loop needed (sphere reverses in place)
- Better for high-flow-rate applications
Bidirectional Prover Limitations:
- Four-way diverter valve is complex and expensive
- U-loop footprint can be large for high-volume provers
- Diverter valve leakage affects accuracy
- Higher capital cost than unidirectional
Unidirectional Pipe Prover
A unidirectional prover uses a closed loop where the displacer (usually an elastomer sphere) travels one direction between detector switches, then returns through a separate bypass loop to the launch position.
Unidirectional Prover Advantages:
- Simpler design (no four-way valve)
- Lower capital cost
- Easier maintenance
- Good for moderate flow rates
Unidirectional Prover Limitations:
- One data point per round trip
- Longer proving time
- Sphere return loop adds length
- Sphere wear requires periodic replacement
Small Volume Prover (SVP)
A small volume prover uses a precision-machined cylinder with a piston that displaces a small, accurately known volume (typically 0.1 to 50 gallons). Multiple passes are required to accumulate sufficient meter pulses for statistical validity.
Image: Small Volume Prover Cross-Section
Cutaway view showing precision cylinder, piston with seals, optical detector switches, hydraulic actuator for piston return, and flow connections.
SVP Advantages:
- Compact footprint (skid-mounted, portable)
- Quick proving (5-15 minutes total)
- No sphere wear or replacement
- Precision cylinder eliminates waterdraw recalibration needs
- Can be shared among multiple meter runs
SVP Limitations:
- Small displaced volume requires high-resolution pulse interpolation
- Multiple passes needed (5-10 typical)
- More sensitive to meter pulse resolution
- Higher per-unit cost than pipe provers
- Requires pulse interpolation for low-resolution meters
Prover Type Selection Guide
| Factor |
Pipe Prover (Bi) |
Pipe Prover (Uni) |
SVP |
| Flow range |
100-30,000+ GPM |
50-10,000 GPM |
10-5,000 GPM |
| Typical volume |
100-5,000 gal |
50-2,000 gal |
0.1-50 gal |
| Footprint |
Large (50-200 ft loop) |
Large (100-300 ft loop) |
Small (skid-mounted) |
| Proving time |
15-30 min |
30-60 min |
5-15 min |
| Capital cost |
High |
Medium |
Medium-High |
| Maintenance |
Diverter valve, sphere |
Sphere replacement |
Piston seals |
| Best application |
Large terminals, pipelines |
Medium-flow stations |
Multiple meters, portable |
3. Prover Sizing Calculations
Prover sizing determines the minimum displaced volume, pipe diameter, detector spacing, and total length to achieve the required pulse count and timing for accurate meter proving.
Minimum Prover Volume
The prover must displace enough volume to generate sufficient meter pulses for the desired repeatability resolution:
Minimum Displaced Volume:
V_min = N_pulses / K_meter
Where:
V_min = Minimum displaced volume per pass (gallons)
N_pulses = Minimum required pulses per pass
K_meter = Meter K-factor (pulses per gallon)
Pulse resolution and repeatability:
The resolution of one pulse contributes to volume uncertainty:
Resolution = 1 / (K_meter x V_prover) x 100%
For 0.02% resolution (to support 0.05% repeatability):
V_min = 100 / (K_meter x 0.02) = 5000 / K_meter
Example - Turbine Meter:
K-factor: 200 pulses/gallon
Required pulses: 10,000 per pass
V_min = 10,000 / 200 = 50 gallons minimum
With 10% safety factor:
V_design = 50 x 1.10 = 55 gallons
Prover Pipe Diameter
The prover pipe diameter is selected to maintain flow velocity between 1 and 5 ft/s through the prover:
Diameter from Flow Velocity:
A = Q / V_target
D = sqrt(4A / pi)
Where:
A = Required cross-sectional area (ft2)
Q = Maximum flow rate (ft3/s)
V_target = Target velocity (3 ft/s typical)
D = Internal pipe diameter (ft)
Convert flow rate:
Q (ft3/s) = Q (GPM) / 7.48 / 60
Example:
Max flow: 1,000 GPM
Q = 1000 / 7.48 / 60 = 2.23 ft3/s
A = 2.23 / 3.0 = 0.743 ft2
D = sqrt(4 x 0.743 / pi) = 0.97 ft = 11.7 inches
Select: 12" nominal pipe (11.376" ID)
Detector Spacing
The distance between detector switches determines the calibrated prover volume:
Detector Spacing:
L = V / A
Where:
L = Distance between detectors (ft)
V = Prover volume (ft3) = gallons / 7.48
A = Pipe cross-sectional area (ft2)
Example:
Volume: 55 gallons = 7.35 ft3
Pipe ID: 11.376 inches = 0.948 ft
A = pi x (0.948)2 / 4 = 0.706 ft2
L = 7.35 / 0.706 = 10.4 ft between detectors
Total Prover Length
Bidirectional Pipe Prover:
L_total = 2 x L_prerun + L_detector + 2 x L_overtravel
Where:
L_prerun = Prerun length (min 1.5 x diameter, min 3 ft)
L_detector = Detector spacing
L_overtravel = Displacer overtravel for deceleration (min 1.0 x dia, min 2 ft)
Unidirectional Pipe Prover:
L_total = L_prerun + L_detector + L_overtravel + L_return
Where L_return = return loop length (approx L_detector + pi x 2D turn radius)
Small Volume Prover:
L_total = L_stroke + L_ends (typically stroke + 2 ft)
Minimum Pass Time
Each pass through the prover should take at least 30 seconds to ensure accurate detector triggering and flow stabilization:
Pass Time:
t = V_prover / Q_flow x 60 (seconds, where Q is in GPM)
For bidirectional: t_round_trip = 2 x t_single_pass
Minimum time check:
If t < 30 seconds at maximum flow, increase prover volume:
V_min_time = Q_max x 30 / 60 (gallons, for single pass)
Example:
Q_max = 1,000 GPM
V_prover = 55 gallons
t = 55 / 1000 x 60 = 3.3 seconds (too short!)
Required volume: V = 1000 x 30 / 60 = 500 gallons minimum
Sizing trade-off: The minimum prover volume is driven by the LARGER of two requirements: (1) pulse count for repeatability, and (2) minimum pass time of 30 seconds. For high-flow-rate meters with high pulse resolution, the pass time requirement often controls the prover size.
4. Meter Factor Calculation
The meter factor is the ratio of true volume (from the prover) to indicated volume (from the meter). It corrects for systematic meter error.
Basic Meter Factor
Meter Factor:
MF = V_prover_base / V_meter_indicated
Where:
MF = Meter factor (dimensionless, typically 0.9900-1.0100)
V_prover_base = Prover volume corrected to base conditions (60 deg F, 0 psig)
V_meter_indicated = Volume indicated by meter (from pulse count / K-factor)
Corrected prover volume:
V_prover_base = V_prover_cal x Ctsp x Cpsp x Ctlp x Cplp
Where:
V_prover_cal = Calibrated prover base volume (from waterdraw)
Ctsp = Prover steel temperature correction
Cpsp = Prover steel pressure correction
Ctlp = Liquid temperature correction (in prover)
Cplp = Liquid pressure correction (in prover)
Corrected meter volume:
V_meter_base = (Pulses / K-factor) x Ctlm x Cplm
Where:
Ctlm = Liquid temperature correction (at meter)
Cplm = Liquid pressure correction (at meter)
Combined Meter Factor
Combined correction meter factor (CCF):
MF = [V_prover x Ctsp x Cpsp x Ctlp x Cplp] / [V_meter x Ctlm x Cplm]
If meter and prover are at the same temperature and pressure:
Ctlp = Ctlm and Cplp = Cplm (cancel out)
Simplified:
MF = (V_prover x Ctsp x Cpsp) / V_meter
This is why Ctsp and Cpsp are the critical corrections -- they account for
the prover steel changing dimensions at operating conditions vs. the
calibration conditions (typically 60 deg F, 0 psig waterdraw).
Repeatability Requirement
API MPMS Chapter 4 requires that consecutive meter factors from a proving run agree within a specified tolerance:
Repeatability Check:
R = (MF_max - MF_min) / MF_mean x 100%
Where:
R = Repeatability (%)
MF_max = Highest meter factor in the run set
MF_min = Lowest meter factor in the run set
MF_mean = Average of all meter factors
Requirements:
Custody transfer: R ≤ 0.05% (0.0005)
Non-custody: R ≤ 0.10% (0.0010)
Example:
Run 1: MF = 1.0003
Run 2: MF = 1.0005
Run 3: MF = 1.0002
Run 4: MF = 1.0004
Run 5: MF = 1.0003
MF_max = 1.0005, MF_min = 1.0002
MF_mean = 1.00034
R = (1.0005 - 1.0002) / 1.00034 x 100 = 0.030%
Result: 0.030% < 0.05% -- PASS (acceptable repeatability)
Meter Factor Application
| Scenario |
MF Application |
Notes |
| New MF from proving |
Apply prospectively to future volumes |
Most common approach for pipeline operations |
| MF changed > 0.0025 |
Retroactive adjustment from last proving |
API MPMS Ch. 12.2 guideline for significant change |
| MF outside 0.9950-1.0050 |
Investigate meter condition |
May indicate worn meter, incorrect K-factor, or calibration needed |
| MF trending steadily |
Schedule maintenance |
Gradual drift indicates progressive wear |
Meter factor vs. K-factor: The meter factor (MF) is a volume correction applied to meter readings. The K-factor is the meter's pulse-to-volume constant (pulses/gallon). Do not confuse the two. A proving determines MF; the K-factor is a fixed meter characteristic.
5. Volume Correction Factors
Volume correction factors adjust for the effects of temperature and pressure on both the prover steel and the liquid being measured. These corrections are essential for accurate custody transfer measurement.
Prover Steel Corrections
Ctsp -- Temperature Correction for Prover Steel:
Ctsp = 1 / (1 + 3 x alpha x (T_op - T_base))
Where:
alpha = Linear thermal expansion coefficient for carbon steel
alpha = 6.17 x 10^-6 per deg F (API MPMS Ch. 11.1, Table 6C)
T_op = Operating temperature (deg F)
T_base = Base temperature = 60 deg F
Example:
Operating temperature: 120 deg F
delta_T = 120 - 60 = 60 deg F
Ctsp = 1 / (1 + 3 x 6.17e-6 x 60)
Ctsp = 1 / (1 + 0.001111)
Ctsp = 1 / 1.001111
Ctsp = 0.998890
The prover volume at 120 deg F is 0.11% LARGER than at 60 deg F due to
thermal expansion. Ctsp corrects back to base conditions.
Cpsp -- Pressure Correction for Prover Steel:
Cpsp = 1 + (P x D) / (2 x E x t) x (1.25 - nu)
Where:
P = Operating pressure (psig)
D = Prover internal diameter (inches)
E = Modulus of elasticity = 30 x 10^6 psi (carbon steel)
t = Wall thickness (inches)
nu = Poisson's ratio = 0.30 (steel)
Example:
P = 500 psig, D = 11.376", t = 0.406"
Cpsp = 1 + (500 x 11.376) / (2 x 30e6 x 0.406) x (1.25 - 0.30)
Cpsp = 1 + 5688 / 24,360,000 x 0.95
Cpsp = 1 + 0.000222
Cpsp = 1.000222
Pressure expands the prover slightly. The correction is small but
significant for custody transfer accuracy.
Combined Steel Correction
Combined prover steel correction:
CSF = Ctsp x Cpsp
Example:
At 120 deg F, 500 psig:
CSF = 0.998890 x 1.000222 = 0.999112
This means the effective prover volume at operating conditions is
0.089% smaller than the calibrated base volume -- the steel expanded
with temperature but the pressure effect partially offsets it.
Liquid Volume Corrections
Liquid volume also changes with temperature and pressure. These corrections (Ctl, Cpl) are applied per API MPMS Chapter 11.1:
| Correction |
Symbol |
Typical Range |
Standard |
| Liquid temperature |
Ctl |
0.97-1.03 |
API MPMS Ch. 11.1 (Table 6A/6B/6C) |
| Liquid pressure |
Cpl |
0.9995-1.0000 |
API MPMS Ch. 11.2.1M |
| Steel temperature |
Ctsp |
0.9985-1.0015 |
API MPMS Ch. 11.1 (Table 6C coefficients) |
| Steel pressure |
Cpsp |
1.0000-1.0005 |
ASME elastic expansion |
When corrections cancel: If the meter and prover are at the same temperature and pressure (typical for a prover installed directly in the meter run), the liquid corrections Ctl and Cpl cancel between numerator and denominator of the meter factor equation. Only the steel corrections (Ctsp, Cpsp) remain significant.
6. Proving Procedures
A systematic proving procedure ensures repeatable, accurate meter factors. The steps below follow API MPMS Chapter 4.8 guidelines.
Pre-Proving Checklist
| Item |
Verification |
| Flow rate stable |
Flow steady within 5% for at least 10 minutes before proving |
| Temperature stable |
Less than 1 deg F change between prover and meter during proving |
| Pressure stable |
Less than 10 psi change during proving run |
| Prover ready |
Sphere inflated (pipe prover), piston seals inspected (SVP) |
| Valves aligned |
All proving manifold valves in correct position, no bypass leaks |
| Detectors tested |
Both/all detector switches respond to mechanical check |
| Pulse counter zeroed |
Flow computer pulse accumulator cleared for new proving |
| RTDs calibrated |
Temperature sensors verified within 0.5 deg F accuracy |
Proving Run Procedure (Pipe Prover)
Step-by-Step Procedure:
1. Verify stable flow and temperature conditions
2. Launch sphere/piston through prover
3. Record first detector switch actuation (start pulse count)
4. Record second detector switch actuation (stop pulse count)
5. Calculate volume: V_meter = Pulses / K-factor
6. Record temperature at prover and meter
7. Record pressure at prover and meter
8. Calculate meter factor: MF = V_prover_corrected / V_meter
9. Repeat for required number of runs (typically 5)
10. Check repeatability: all MFs within 0.05% range
11. If pass: accept average MF and apply to meter readings
12. If fail: investigate cause, discard outlier or re-prove
Bidirectional prover:
Each round trip gives TWO MF values (forward + reverse).
5 round trips = 10 individual meter factors for averaging.
SVP:
Each run consists of multiple piston passes.
Record accumulated pulses across all passes per run.
5 runs x 5-10 passes/run = 25-50 total piston strokes.
Proving Frequency
| Application |
Typical Frequency |
Trigger for Re-Proving |
| Crude oil custody transfer |
Each shipment or daily |
Product change, flow rate change > 20% |
| NGL pipeline |
Daily or per batch |
Composition change, temperature swing > 10 deg F |
| Refined products |
Each product switch or weekly |
Product grade change, viscosity change |
| LPG |
Daily |
Temperature change, composition change |
| Allocation metering |
Monthly |
Significant flow or composition change |
Troubleshooting Proving Problems
| Problem |
Likely Cause |
Corrective Action |
| Repeatability > 0.05% |
Air/vapor in prover, detector malfunction, unstable flow |
Vent prover, check detectors, stabilize flow before re-proving |
| MF changed > 0.25% |
Meter wear, incorrect K-factor, prover leak |
Inspect meter internals, verify K-factor, pressure-test prover |
| Sphere not detected |
Deflated sphere, fouled detector, wiring issue |
Re-inflate sphere, clean detector, check wiring |
| Erratic pulse count |
Electrical noise, damaged pickup, loose wiring |
Check shielding, replace pickup, tighten connections |
| Temperature difference > 1 deg F |
Sun exposure, inadequate insulation, long piping between meter and prover |
Insulate piping, shade equipment, minimize distance |
Documentation requirement: All proving results must be documented with date, time, operator, flow rate, temperatures, pressures, pulse counts, calculated meter factors, and repeatability. Records must be retained per API MPMS Chapter 12.2 and applicable regulatory requirements (typically 3-7 years).