Flow Measurement

Turbine Meter Sizing & Installation

Size and install turbine meters for natural gas and liquid custody transfer measurement per AGA Report No. 7, API MPMS Chapter 5.3, and ISO 9951. Understand K-factor linearity, Reynolds number effects, and proving requirements.

Gas velocity

15-70 ft/s

Optimal operating range for gas turbine meters per AGA 7.

Rangeability

10:1 to 20:1

Standard to high-performance turndown ratio for custody transfer.

Accuracy

±0.5-1.0%

Custody transfer accuracy when properly sized, installed, and proved.

Use this guide when you need to:

  • Size a turbine meter for gas or liquid service.
  • Design meter piping with proper straight runs.
  • Evaluate meter performance and accuracy.
  • Develop a meter proving program.

1. Operating Principles

Turbine meters measure volumetric flow rate by counting the rotations of a multi-bladed rotor placed in the fluid stream. As fluid flows through the meter body, it impinges on the rotor blades, causing the rotor to spin at a speed proportional to the volumetric flow rate. Each revolution of the rotor corresponds to a fixed volume of fluid passing through the meter.

Image: Turbine Meter Cross-Section

Technical illustration showing flow path, rotor assembly, upstream/downstream bearing supports, pickup coil, nose cone, and tail cone with labeled components.

Rotor assembly

Multi-blade helical rotor

Precision-machined rotor with helical blades converts fluid momentum into rotation proportional to flow velocity.

Pickup coil

Magnetic pulse detection

Non-intrusive magnetic pickup coil generates electrical pulses as rotor blades pass, providing digital flow signal.

Flow conditioning

Internal straighteners

Nose cone and upstream stator vanes condition flow before it reaches the rotor for accurate measurement.

Bearing system

Low-friction support

Ball or journal bearings support the rotor shaft with minimal friction for consistent performance over time.

K-Factor Concept

The K-factor is the fundamental calibration constant of a turbine meter. It defines the number of pulses generated per unit volume of fluid passing through the meter.

K-Factor Definition: K = Pulses / Volume Where: K = Meter factor (pulses/ft³ or pulses/gallon) Pulses = Number of electrical pulses from pickup coil Volume = Volume of fluid passed through meter Flow Rate from K-Factor: Q = f / K Where: Q = Volumetric flow rate (ft³/s or gal/min) f = Pulse frequency (Hz = pulses/second) K = Meter factor (pulses/ft³ or pulses/gal) Example: K = 2,540 pulses/ft³ f = 4,233 Hz (pulses/second) Q = 4,233 / 2,540 = 1.666 ft³/s = 100 ACFM

How Turbine Meters Differ from Other Technologies

Feature Turbine Meter Orifice Meter Ultrasonic Meter
Measurement principle Rotor rotation (velocity) Differential pressure Transit time difference
Moving parts Yes (rotor, bearings) No No
Rangeability 10:1 to 20:1 3:1 to 5:1 20:1 to 50:1
Accuracy ±0.5-1.0% ±0.5-1.0% ±0.1-0.5%
Pressure drop Low-moderate (1-5 psi) Moderate-high (5-50 psi) Very low (<0.5 psi)
Maintenance Bearing replacement periodic Plate inspection periodic Minimal
Cost Moderate Low High
Liquid service Excellent Good Excellent
Key advantage: Turbine meters combine good accuracy (±0.5-1.0%), wide rangeability (10:1 to 20:1), and moderate cost, making them the dominant technology for gas custody transfer in the midstream industry. They are well-proven with decades of field performance data.

Applications in the Midstream Industry

  • Custody transfer: Buy/sell metering at pipeline interconnects, plant inlets/outlets, and delivery points
  • Allocation metering: Proportional flow measurement for multi-party production sharing
  • Check metering: Verification meters installed in parallel or series with primary custody meters
  • Plant measurement: Gas processing plant inlet, residue gas, and NGL product metering
  • Distribution: City gate stations and large commercial/industrial customer meters
  • Liquid hydrocarbons: Crude oil, condensate, NGL, and refined product measurement

2. Meter Sizing Method

Proper turbine meter sizing ensures the meter operates within its linear range at all expected flow conditions. The sizing process converts flow rates to actual conditions, selects an appropriate meter body size, and verifies velocity and Reynolds number requirements.

Step-by-Step Sizing Procedure

Step 1: Convert Standard Flow to Actual Flow

Gas Flow Conversion (MMSCFD to ACFM): ACFM = (MMSCFD × 10&sup6; / 1440) × (Pb / P_act) × (T_act / Tb) × (1 / Z) Where: ACFM = Actual cubic feet per minute MMSCFD = Standard flow in million standard cubic feet per day Pb = Base pressure = 14.73 psia (AGA standard) P_act = Actual operating pressure (psia) T_act = Actual operating temperature (°R = °F + 459.67) Tb = Base temperature = 519.67 °R (60°F) Z = Gas compressibility factor (dimensionless) Example: Flow: 10 MMSCFD Pressure: 500 psig (514.7 psia) Temperature: 80°F (539.67 °R) Z = 0.92 SCFM = 10 × 10&sup6; / 1440 = 6,944 SCFM ACFM = 6,944 × (14.73 / 514.7) × (539.67 / 519.67) × (1 / 0.92) ACFM = 6,944 × 0.02862 × 1.0385 × 1.0870 ACFM = 224.3 ACFM

Step 2: Select Meter Size

Compare the calculated ACFM at all flow conditions (min, normal, max) against standard meter capacity ranges.

Meter Size (NPS) Internal ID (in) Q_min (ACFM) Q_max (ACFM) Rangeability
2" 1.939 10 250 25:1
3" 2.900 25 600 24:1
4" 3.826 50 1,200 24:1
6" 5.761 100 3,200 32:1
8" 7.625 200 6,500 32.5:1
10" 9.564 350 12,000 34:1
12" 11.376 500 18,000 36:1

Note: Actual meter specifications vary by manufacturer. These are representative values for sizing purposes. Always confirm with the manufacturer's data sheet.

Step 3: Verify Flow Velocity

Flow Velocity Calculation: v = Q_actual / A Where: v = Flow velocity (ft/s) Q_actual = Actual volume flow (ft³/s) = ACFM / 60 A = Meter internal cross-sectional area (ft²) = π × (D/24)² D = Meter internal diameter (inches) Optimal Velocity Ranges: Gas service: 15-70 ft/s (custody transfer best at 20-50 ft/s) Liquid service: 3-30 ft/s Example (continuing from Step 1): ACFM = 224.3 ACFM Try 4" meter: ID = 3.826 inches A = π × (3.826/24)² = π × (0.15942)² = 0.0799 ft² Q_actual = 224.3 / 60 = 3.738 ft³/s v = 3.738 / 0.0799 = 46.8 ft/s 46.8 ft/s is within 15-70 ft/s optimal range. Good selection.

Step 4: Check Reynolds Number

Reynolds Number: Re = ρ × v × D / μ Where: Re = Reynolds number (dimensionless) ρ = Fluid density (lb/ft³) v = Flow velocity (ft/s) D = Meter internal diameter (ft) μ = Dynamic viscosity (lb/(ft·s)) Requirements: - Minimum Re > 10,000 for linear response - Preferred Re > 50,000 for best accuracy - Re < 10,000: meter response becomes non-linear, accuracy degrades Example: Gas density at 514.7 psia, 80°F, SG = 0.60, Z = 0.92: ρ = (P × MW) / (Z × R × T) ρ = (514.7 × 17.38) / (0.92 × 10.73 × 539.67) = 1.68 lb/ft³ Gas viscosity ≈ 7.5 × 10&supmin;&sup6; lb/(ft·s) Re = 1.68 × 46.8 × (3.826/12) / (7.5 × 10&supmin;&sup6;) Re = 1.68 × 46.8 × 0.319 / 7.5e-6 Re = 3,340,000 Well above 50,000 minimum. Excellent conditions for turbine meter.

Step 5: Estimate Pressure Drop

Pressure Drop Across Turbine Meter: ΔP = K × ρ × v² / (2 × 144) Where: ΔP = Pressure drop (psi) K = Meter loss coefficient (typically 1.7-3.8 depending on size) ρ = Fluid density (lb/ft³) v = Flow velocity (ft/s) 144 = conversion factor (in²/ft²) Typical Loss Coefficients (K): 2" meter: K = 3.8 4" meter: K = 2.8 6" meter: K = 2.4 8" meter: K = 2.1 12" meter: K = 1.7 Example (4" meter, gas): ΔP = 2.8 × 1.68 × (46.8)² / (2 × 144) ΔP = 2.8 × 1.68 × 2190.24 / 288 ΔP = 35.8 psi At 500 psig operating, 35.8 psi is 7.0% of operating pressure. Acceptable if system can tolerate this loss.

Sizing Decision Matrix

Condition Assessment Action
v_max < 15 ft/s (gas) Under-velocity Select smaller meter size
15 ≤ v_max ≤ 70 ft/s Optimal range Good selection
v_max > 70 ft/s (gas) Over-velocity Select larger meter or parallel runs
Re < 10,000 Non-linear regime Select smaller meter to increase velocity
ΔP > 10% of operating pressure Excessive pressure loss Select larger meter or accept loss
Rangeability requirement not met Insufficient turndown Use parallel meter runs or different technology
Sizing rule of thumb: Target normal flow velocity at 40-60% of the meter's maximum capacity (Q_max). This provides headroom for flow surges while maintaining good accuracy across the operating range. Never size a meter at more than 80% of Q_max for normal operations.

3. Installation Requirements

Proper installation is critical for turbine meter accuracy. The meter must receive a fully developed, symmetrical, swirl-free velocity profile. AGA Report No. 7 provides specific requirements for upstream and downstream piping, flow conditioning, and meter orientation.

Image: Turbine Meter Installation Layout

Plan view showing upstream straight run (10D), flow conditioner location, meter body, downstream straight run (5D), thermometer well, and pressure taps with dimensions.

Straight Run Requirements (AGA Report No. 7)

Minimum Straight Pipe Lengths: Upstream: 10D minimum (10 pipe diameters) Downstream: 5D minimum (5 pipe diameters) With flow conditioner: Upstream of conditioner: 5D minimum Conditioner to meter: 5D minimum Downstream of meter: 5D minimum Without flow conditioner (greater straight runs needed): After single elbow: 20D minimum upstream After two elbows in-plane: 25D minimum upstream After two elbows out-of-plane: 40D minimum upstream After tee (branch flow): 30D minimum upstream After reducer: 15D minimum upstream After valve: 50D minimum upstream (valve must be fully open) Example (6" meter): Minimum upstream: 10 × 6 = 60 inches = 5 feet Minimum downstream: 5 × 6 = 30 inches = 2.5 feet After two out-of-plane elbows without conditioner: 40 × 6 = 240 inches = 20 feet upstream required

Flow Conditioners

When adequate straight run is not available, flow conditioners remove swirl and velocity profile distortion.

Conditioner Type Effectiveness Pressure Drop Application
19-tube bundle Good swirl removal Low (0.2-0.5 psi) Standard AGA 7 requirement; most common
Plate-type (CPA 50E) Excellent (swirl + profile) Moderate (1-3 psi) Tight installations; best accuracy
Vane-type Good swirl removal Low-moderate Retrofit installations
Honeycomb Fair swirl removal Very low Low-pressure drop requirements

Meter Orientation

  • Horizontal: Preferred orientation for gas service. Meter axis horizontal with pickup coil on top.
  • Vertical (flow up): Acceptable for liquid service. Prevents air accumulation.
  • Vertical (flow down): Not recommended. Can cause bearing damage from rotor weight.
  • Inclined: Acceptable if manufacturer confirms. Avoid inclinations greater than 15 degrees without manufacturer approval.

Pressure and Temperature Measurement

Pressure Tap Location: - Upstream of meter: 1D to 2D upstream of meter flange - Downstream of meter: 2D to 5D downstream of meter flange - For custody transfer: measure at both locations Temperature Well Location: - Downstream of meter: 2D to 5D downstream - Must not create flow disturbance upstream of meter - Thermowell insertion: 1/3 to 2/3 of pipe diameter Volume Correction: Q_base = Q_actual × (P_f / P_b) × (T_b / T_f) × Z_b / Z_f Where: Q_base = Volume at base conditions P_f = Flowing pressure (psia) P_b = Base pressure (14.73 psia) T_f = Flowing temperature (°R) T_b = Base temperature (519.67 °R) Z_f = Z-factor at flowing conditions Z_b = Z-factor at base conditions (≈ 1.0)

Strainer and Filter Requirements

  • Upstream strainer: Required for all turbine meter installations to protect rotor and bearings from debris
  • Mesh size: 40-60 mesh (250-420 micron) for gas; 80-100 mesh for liquid
  • Differential pressure indicator: Install across strainer to monitor fouling
  • Cleaning frequency: Clean when differential reaches 5-10 psi; more often during commissioning
  • Basket type preferred: Allows cleaning without removing from line; dual-basket for continuous operation
Critical installation rule: Never install a turbine meter without an upstream strainer. Pipeline debris (weld slag, scale, sand) can damage the rotor within hours, destroying measurement accuracy and requiring expensive repair or replacement.

4. Performance & Accuracy

Turbine meter accuracy depends on maintaining linear K-factor response across the operating flow range. Understanding the factors that affect K-factor linearity is essential for achieving custody transfer accuracy.

Image: K-Factor vs. Flow Rate (Linearity Curve)

Graph showing K-factor on Y-axis vs. flow rate on X-axis, with flat linear region in middle, droop at low flow (bearing friction dominance), and slight rise at high flow (blade tip effects). Annotated with ±0.5% linearity band.

K-Factor Linearity

An ideal turbine meter would produce a constant K-factor across its entire flow range. In practice, K-factor varies slightly due to bearing friction, fluid viscosity, and blade aerodynamics.

Flow Region K-Factor Behavior Cause Accuracy Impact
Low flow (<Q_min) K drops significantly Bearing friction dominates; under-registration Error > 2%, non-linear
Transition (Q_min to 2×Q_min) K rising toward plateau Decreasing friction/velocity ratio ±1-2%
Linear range (2×Q_min to 0.8×Q_max) K essentially constant Fluid forces dominate; friction negligible ±0.25-0.5%
High flow (>0.8×Q_max) K may increase slightly Blade tip effects; compressibility ±0.5-1.0%
Over-range (>Q_max) K unstable Excessive rotor speed; bearing stress Unpredictable; damage risk

Reynolds Number Effects

Reynolds Number and K-Factor Relationship: At high Re (> 50,000): - K-factor is essentially constant - Best measurement accuracy achievable - Preferred operating regime for custody transfer At moderate Re (10,000-50,000): - K-factor remains stable but may show slight viscosity sensitivity - Acceptable for custody transfer with proper calibration - Linearization may improve accuracy At low Re (< 10,000): - K-factor drops below linear value (under-registration) - Non-linear behavior makes correction difficult - Not suitable for custody transfer without special calibration Viscosity Correction (Universal Viscosity Curve): Many turbine meters can be characterized with a universal viscosity curve (UVC) that plots K/K_ideal vs. Re/Re_ref. This allows accurate measurement across viscosity changes by applying Re-based correction factors.

Factors Affecting Accuracy

Factor Effect on K-Factor Magnitude Mitigation
Bearing wear K decreases (under-registration) 0.1-0.5% per year Regular proving; bearing replacement
Blade damage/fouling K changes unpredictably 1-5% possible Upstream strainer; periodic inspection
Swirl in flow K increases (over-registration) 1-3% Flow conditioner; proper straight runs
Non-uniform velocity profile K shifts (direction depends on profile) 0.5-2% Adequate upstream length; flow conditioner
Pulsating flow K increases (always over-registers) 1-10% Move meter away from compressors; add volume
Two-phase flow K erratic; severe over-registration 5-50% Upstream separator; avoid liquid in gas meters
Temperature change K shifts due to thermal expansion 0.01%/°F typically Temperature-compensated K-factor
Pressure change Negligible effect on K directly <0.05% Volume correction handles this

Accuracy Classes

Custody Transfer (±1.0% per AGA 7): - K-factor repeatability: ±0.15% or better - Linearity across 10:1 range: ±0.5% - Proving frequency: monthly or per contractual agreement - Flow conditioner: recommended - Upstream strainer: required - Redundant instrumentation: recommended (dual pickup coils) Check Meter (±2.0%): - K-factor repeatability: ±0.5% or better - Linearity across 5:1 range: ±1.0% - Proving frequency: quarterly or semi-annually - Flow conditioner: optional - Less stringent installation requirements Allocation Metering (±2-5%): - Least stringent accuracy requirements - Proving frequency: annually or as-needed - Simplified installation acceptable
Accuracy reality check: Published meter accuracy specifications assume ideal installation conditions. In field installations with imperfect piping, inadequate straight runs, or poor maintenance, actual accuracy may be 2-3 times worse than the specification. Regular proving is the only way to verify actual field accuracy.

5. Proving & Calibration

Meter proving (in-situ calibration) is essential for custody transfer measurement. Proving determines the actual meter factor under operating conditions and verifies that the meter continues to perform within specifications.

Proving Methods

Method Reference Standard Accuracy Application
Master meter Transfer-proved turbine or ultrasonic meter ±0.25-0.5% Gas meters; portable proving
Bell prover Calibrated volume displacement ±0.1-0.2% Shop calibration; low-pressure gas
Critical flow prover (sonic nozzle) Choked flow through calibrated nozzle ±0.25% High-pressure gas; field or shop
Pipe prover (displacement) Calibrated pipe section with displacer ±0.02-0.05% Liquid meters; permanent installation
Compact prover Small-volume piston prover ±0.02-0.05% Liquid meters; portable or permanent
Gravimetric (weigh tank) Calibrated scale with collection tank ±0.05-0.1% Liquid; shop calibration reference

Proving Procedure (Gas — Master Meter Method)

Master Meter Proving Procedure: 1. Install master meter in series with meter under test 2. Stabilize flow at test point (typically 3-5 flow rates across range) 3. Record simultaneous readings from both meters over fixed time period 4. Calculate meter factor: MF = V_master / V_test Where: MF = Meter factor (correction multiplier) V_master = Volume registered by master meter (corrected to base) V_test = Volume registered by test meter (uncorrected) Acceptance Criteria (AGA 7 Section 7): - Minimum 5 proving runs per test point - Repeatability of proving runs: ±0.25% (maximum deviation from mean) - MF must be within ±2.0% of unity (0.98-1.02) - If MF outside 0.98-1.02: investigate meter condition Proving Frequency (Custody Transfer): - Initial proving: before meter enters service - Routine proving: monthly (common contractual requirement) - After maintenance: re-prove before returning to service - After flow upset: re-prove if pulsation or two-phase event occurred

Linearization

Modern flow computers can store multiple K-factors across the meter's flow range to improve overall accuracy.

Multi-Point Linearization: Instead of a single K-factor, store K vs. flow rate at multiple points: Point 1: Q = 100 ACFM, K = 2,538.2 pulses/ft³ Point 2: Q = 200 ACFM, K = 2,541.5 pulses/ft³ Point 3: Q = 400 ACFM, K = 2,542.0 pulses/ft³ Point 4: Q = 800 ACFM, K = 2,541.8 pulses/ft³ Point 5: Q = 1100 ACFM, K = 2,540.1 pulses/ft³ The flow computer interpolates between points for the actual K at any flow rate. Improvement: Single K-factor accuracy: ±0.5% over 10:1 range Multi-point linearized: ±0.15-0.25% over 10:1 range

When to Recalibrate or Replace

  • K-factor shift > 0.5% from previous proving: investigate bearing condition
  • K-factor shift > 1.0%: bearing replacement likely needed; re-prove after repair
  • K-factor shift > 2.0%: meter may have blade damage; remove for inspection
  • Repeatability degradation: if proving runs show > 0.5% spread, bearing or blade issue
  • Audible bearing noise: replace bearings immediately
  • Vibration increase: rotor imbalance from blade damage or fouling
Proving economics: Regular proving costs far less than the value of measurement error. A 0.5% measurement error on a 50 MMSCFD gas stream at $3.00/MMBTU represents approximately $22,000/month in potential billing error. Monthly proving typically costs $1,000-3,000.

6. Troubleshooting

Common Problems and Solutions

Problem Symptoms Likely Causes Solution
Under-registration K-factor lower than baseline; meter reads low Bearing wear, fouled blades, low velocity Replace bearings; clean rotor; verify sizing
Over-registration K-factor higher than baseline; meter reads high Swirl in flow, pulsation, two-phase flow Install flow conditioner; add pulsation dampener; install separator upstream
Erratic readings Flow indication fluctuates wildly Damaged blade, intermittent two-phase, loose pickup coil Inspect rotor; check for liquid; tighten pickup
No signal Zero flow indication despite flow Pickup coil failure, broken wire, rotor seized Check pickup coil resistance; inspect wiring; remove meter for inspection
Excessive noise Audible grinding, clicking, or squealing Bearing failure, blade contact, debris in meter Remove from service immediately; replace bearings
Poor repeatability Proving runs show >0.5% spread Bearing wear, intermittent obstruction, flow instability Replace bearings; check upstream conditions; verify steady flow during proving
Rotor spin-down Meter continues registering after flow stops Thermal convection, pressure equalization, mechanical vibration Install check valve downstream; verify no flow; add low-flow cutoff

Bearing Life and Maintenance

Bearing Life Estimation: Typical bearing life: 3-7 years (depends on gas cleanliness and velocity) Factors reducing bearing life: - High velocity (> 60 ft/s): life reduces proportional to v² - Dirty gas (no strainer): life reduced to months - Liquid carryover: can destroy bearings within days - Pulsation: accelerates wear from repeated loading Maintenance Schedule (Custody Transfer): Monthly: - Prove meter (verify K-factor) - Check strainer differential pressure - Record bearing run-down time (coast test) Semi-annually: - Clean strainer element - Inspect pickup coil and wiring - Verify flow computer calibration Annually: - Remove meter for visual inspection - Measure bearing clearance - Replace bearings if coast time degraded > 20% Every 3-5 years: - Complete overhaul: new bearings, inspect blades - Recertify with shop calibration - Consider spare meter rotation program

Coast Test (Spin-Down Test)

The coast test is a simple field check of bearing condition. It measures how long the rotor continues spinning after flow is stopped.

Coast Test Procedure: 1. Record the flow rate before stopping flow 2. Close the upstream valve rapidly (within 2-3 seconds) 3. Observe the meter and record the time until the rotor stops spinning 4. Record the number of additional pulses registered after valve closure Interpretation: New meter: 30-90 seconds coast time (at moderate flow rates) Acceptable: > 50% of baseline coast time Warning: 25-50% of baseline Replace bearings: < 25% of baseline Important: - Coast time depends on flow rate before stopping: higher flow = longer coast - Always compare to baseline recorded when meter was new or freshly rebuilt - Coast time is NOT an absolute indicator; it is a trending tool

Pulsation Effects and Mitigation

Pulsating flow from compressors causes turbine meters to over-register because the rotor responds faster to acceleration than deceleration (inertial ratcheting).

  • Error magnitude: 1-10% over-registration depending on pulsation amplitude
  • Root cause: Reciprocating compressors, regulator oscillation, control valve cycling
  • Pulsation index: PI = (Q_peak - Q_trough) / Q_average. If PI > 10%, significant error likely.
  • Mitigation: Move meter 50+ pipe diameters from pulsation source; add volume (pipe, bottle, or tank) between source and meter
  • AGA 7 guidance: Pulsation effects should be evaluated if meter is within 100 diameters of a reciprocating compressor
Pulsation rule: If you can hear pulsation at the meter location, the meter is likely over-registering. Install the meter as far as possible from reciprocating compressors or pressure regulators. A pulsation dampener or large volume pipe section between the compressor and meter is the most effective solution.

Regulatory and Contractual Considerations

  • FERC tariffs: Interstate pipeline tariffs specify measurement accuracy requirements, proving frequency, and dispute resolution procedures
  • State regulations: Many states require specific meter types and accuracy classes for custody transfer
  • Contractual requirements: Gas sales agreements typically specify meter accuracy (±1%), proving frequency (monthly), and meter factor limits (0.98-1.02)
  • Audit requirements: Maintain proving records for 3-7 years minimum (varies by jurisdiction)
  • Dispute resolution: When buyer and seller meters disagree by more than the contractual tolerance, standard procedures determine which meter governs billing