Pipeline Operations — Measurement & Metering

Gas Measurement Fundamentals

Accurate natural gas measurement underpins every financial transaction in the midstream value chain. From wellhead allocation to custody transfer at pipeline interconnects, measurement determines revenue, royalties, taxes, and contractual compliance. This guide covers the principal meter technologies governed by AGA Reports 3, 7, 9, and 11, along with gas analysis, sampling, flow conditioning, and uncertainty quantification. Whether you are designing a new meter station or auditing an existing one, these fundamentals provide the engineering basis for defensible, accurate measurement.

Launch Calculator Meter Selection

Custody Transfer

≤ 1.0% Uncertainty

Typical contractual accuracy requirement for fiscal-quality gas measurement.

Meter Technologies

7 Primary Types

Orifice, turbine, ultrasonic, Coriolis, rotary, diaphragm, and V-cone meters.

Key Standards

AGA 3 · 7 · 9 · 11

Primary calculation and installation standards for natural gas metering.

Contents

Use this guide when you need to:

  • Select the right meter technology for a given application.
  • Understand AGA calculation standards for custody transfer.
  • Design sampling systems and GC installations.
  • Quantify measurement uncertainty and error budgets.

1. Introduction to Gas Measurement

Natural gas measurement serves three fundamental purposes in the midstream industry: custody transfer (determining the quantity and quality of gas changing ownership), allocation (distributing commingled volumes back to individual producers or shippers), and regulatory compliance (reporting volumes for royalties, severance taxes, and emissions tracking). The financial stakes are enormous: a 0.5% measurement bias on a 200 MMscf/d pipeline at $3.00/MMBtu results in over $3 million per year in misallocated revenue.

Measurement Hierarchy

The gas industry distinguishes between measurement tiers based on the financial and regulatory significance of the measurement point:

Custody Transfer

Fiscal Quality – Highest Accuracy

Ownership change points between buyer and seller. Governed by contractual accuracy requirements, typically ±0.5% to ±1.0% of reading. Subject to periodic proving, calibration, and audit. Uses AGA-compliant primary elements with redundant instrumentation.

Allocation

Check Metering – Moderate Accuracy

Distributes commingled production among multiple interest owners. Accuracy requirements typically ±1.0% to ±2.0%. Monthly allocation statements reconcile individual well meters against the custody transfer meter. Imbalances are resolved per the allocation agreement.

Operational

Process Measurement – General Purpose

Used for process control, compression monitoring, flare measurement, and facility balancing. Accuracy of ±2% to ±5% is generally acceptable. May use non-AGA meter types (annubar, vortex, thermal mass) where simplicity and cost are prioritized.

Energy-Based Measurement

Natural gas is sold on an energy basis (MMBtu or GJ), not a volumetric basis. This requires two independent measurements at the custody transfer point:

Energy determination: Energy (MMBtu) = Volume (Mscf) × Heating Value (BTU/scf) × 10−3 Volume at standard conditions: Qstd = Qactual × (Pf / Pb) × (Tb / Tf) × (1 / Zf) × Zb Where: Qstd = volume at standard (base) conditions Qactual = volume at flowing conditions Pf = flowing pressure (psia) Pb = base pressure (14.73 psia per AGA) Tb = base temperature (519.67 °R = 60 °F) Tf = flowing temperature (°R) Zf = compressibility at flowing conditions (AGA 8) Zb = compressibility at base conditions (≈ 1.0) Heating value: Determined by gas chromatograph per GPA 2261 / GPA 2145 Gross (superior) heating value used for billing in North America
Base conditions matter: Different states, contracts, and countries use different base (standard) conditions. The US gas industry predominantly uses 14.73 psia and 60°F (per ANSI/API 2530), while some older contracts use 14.696 psia. Canada uses 14.696 psia and 60°F. The difference between 14.73 and 14.696 psia represents a 0.23% volume difference, which can amount to significant dollars on large-volume custody transfer points. Always verify the contractual base conditions before configuring flow computers.

2. Orifice Meters

The orifice meter is the most widely installed gas measurement device in North America, with an estimated installed base exceeding one million units. Its longevity stems from a century of empirical research, well-defined standards (AGA Report No. 3 / API MPMS Chapter 14.3), and the ability to verify calibration by physical inspection of the orifice plate and meter tube. Orifice meters measure flow by creating a differential pressure across a precisely machined restriction (the orifice plate) inserted in the gas stream.

Operating Principle

AGA 3 orifice flow equation: Qv = C' × √(hw × Pf) Expanded form: Qv = Cd × Ev × Y1 × Fb × Ftb × Fpb × Ftf × Fgr × Fpv × d2 × √(hw × Pf) Where: Cd = discharge coefficient (Reader-Harris/Gallagher equation) Ev = velocity of approach factor = 1 / √(1 − β4) Y1 = expansion factor (gas compressibility correction) Fb = basic orifice factor Ftb = flowing temperature base factor Fpb = pressure base factor Ftf = flowing temperature factor Fgr = real gas relative density factor Fpv = supercompressibility factor d = orifice bore diameter (inches) hw = differential pressure (inches H2O) Pf = flowing pressure (psia)

Discharge Coefficient

The discharge coefficient Cd is the most critical parameter in the orifice flow equation. AGA 3 Part 1 (2012) specifies the Reader-Harris/Gallagher (RHG) equation, which replaced the earlier Buckingham equation. The RHG equation accounts for Reynolds number effects, beta ratio, and tap location:

Reader-Harris/Gallagher equation (simplified form): Cd = C∞ + b/ReDn Where C∞ is a function of β and tap type, and b/ReDn accounts for Reynolds number effects. Validity range: β (beta ratio d/D): 0.10 to 0.75 ReD (pipe Reynolds number): ≥ 4,000 Pipe diameter D: 2" to 30" (50 mm to 760 mm) Typical Cd values (flange taps): β = 0.20: Cd ≈ 0.5988 β = 0.40: Cd ≈ 0.6045 β = 0.50: Cd ≈ 0.6067 β = 0.60: Cd ≈ 0.6097 β = 0.70: Cd ≈ 0.6151 Uncertainty in Cd: ± 0.5% for β ≤ 0.6 ± (0.5 + 0.6β2)% for β > 0.6

Beta Ratio Selection

Beta Ratio Differential Range Advantages Disadvantages
0.20 – 0.40 High hw at low flow Good low-flow sensitivity, lower Cd uncertainty High permanent pressure loss, limited capacity
0.40 – 0.55 Moderate hw Best balance of accuracy and range Moderate pressure loss
0.55 – 0.65 Lower hw at high flow Higher capacity, lower pressure loss Higher Cd uncertainty, sensitive to edge sharpness
0.65 – 0.75 Low hw Maximum capacity Highest uncertainty, requires frequent plate inspection

Chart Recorders vs. Electronic Flow Measurement (EFM)

Legacy orifice meter installations used circular chart recorders to trace differential pressure and static pressure over time. The charts were integrated manually or by optical scanners to determine daily volumes. Modern installations use electronic flow measurement (EFM) devices that sample differential pressure, static pressure, and temperature at intervals of 1 second or faster, computing instantaneous flow rate and accumulating volume digitally.

Feature Chart Recorder EFM / Flow Computer
Calculation frequency Integrated (average over chart period) Every 1 second (typical)
Accuracy ± 1.0% to ± 2.0% (integration errors) ± 0.25% to ± 0.5% (properly configured)
Data resolution Continuous pen trace (limited readability) 1-second samples, hourly/daily archives
Audit trail Physical charts (storage required) Digital logs with event timestamps
Multi-plate/multi-run Manual plate changes, separate charts Automatic stacking, run switching logic
Remote access None – requires site visit SCADA, cellular, satellite communication

Orifice Plate Inspection

  • Edge sharpness: The upstream edge of the orifice bore must be sharp and free of nicks, burrs, and wear. AGA 3 requires the edge radius to be less than 0.0004 × d (bore diameter). A worn edge increases Cd and causes the meter to over-register.
  • Bore diameter: Measured with a micrometer at four equally spaced points. Must agree within 0.001 inch for plates up to 1 inch bore, and within 0.0005 × d for larger bores.
  • Flatness: The orifice plate must be flat within 0.010 inch per inch of dam height (D − d) / 2 to prevent differential pressure bias.
  • Concentricity: The orifice bore must be concentric with the meter tube within prescribed tolerances to ensure symmetric flow contraction.
Permanent pressure loss: The orifice meter creates a permanent pressure loss of approximately 40% to 80% of the measured differential pressure, depending on the beta ratio. For a 100" H2O differential at β = 0.50, the permanent loss is approximately 73" H2O (2.6 psi). On high-volume transmission pipelines, this pressure loss translates directly to compression costs. This is one of the primary reasons ultrasonic meters have replaced orifice meters on large-diameter trunk lines.

3. Turbine Meters

Gas turbine meters use a free-spinning rotor mounted in the flow stream to measure volumetric flow rate. The rotational speed of the rotor is proportional to the gas velocity, and each revolution sweeps a known volume. AGA Report No. 7 (Measurement of Gas by Turbine Meters) governs the calibration, installation, and use of turbine meters for custody transfer applications.

Operating Principle

Turbine meter flow equation: Q = Pulses / K Where: Q = volume at actual (line) conditions per unit time Pulses = number of electrical pulses from pickup coil K = meter K-factor (pulses per unit volume) Volume at standard conditions: Qstd = Qactual × (Pf / Pb) × (Tb / Tf) × (Zb / Zf) K-factor determination: K = f / Qactual Where: f = pulse frequency (Hz) Qactual = actual volumetric flow rate (ft3/hr or m3/hr) The K-factor is determined by calibration against a reference standard (bell prover, sonic nozzle, or transfer-standard turbine meter) and is characterized across the operating flow range.

K-Factor and Linearity

The K-factor of a turbine meter varies across the flow range due to bearing friction, fluid drag on the rotor blades, and tip clearance effects. A well-designed turbine meter maintains K-factor linearity within ±0.5% over a 10:1 turndown ratio (Qmax:Qmin). Some designs achieve 20:1 or greater with linearization corrections applied in the flow computer.

Parameter Typical Value Notes
Rangeability (turndown) 10:1 to 25:1 Depends on design and pressure
Repeatability ± 0.1% to ± 0.25% At stable flow conditions
Linearity ± 0.5% to ± 1.0% Over calibrated range, before linearization
Accuracy (after linearization) ± 0.25% to ± 0.5% Across 10:1 calibrated range
Pressure rating ANSI 150 to ANSI 2500 Full pipeline pressure capability
Size range 2" to 12" (common) Larger sizes available from some manufacturers

Proving Requirements

Turbine meters for custody transfer must be proved (calibrated in service) at regular intervals, typically monthly or quarterly, depending on the contract. Proving is performed by comparing the meter output against a reference standard under actual operating conditions. Common proving methods include:

  • Sonic nozzle prover: A critical-flow nozzle installed in series with the turbine meter. Flow through the nozzle at critical conditions is calculable from upstream pressure and temperature. Accuracy: ± 0.25%.
  • Transfer-standard meter: A calibrated master meter (often another turbine meter with a traceable calibration) installed temporarily in series.
  • In-situ spin test: Verifies mechanical condition by measuring free-spin coast-down time. Not a flow calibration, but detects bearing wear and drag changes.
Pressure effect on capacity: Turbine meter capacity increases with operating pressure because the gas density (and therefore mass flow) increases at higher pressures while the volumetric velocity that drives the rotor remains within its mechanical range. A 4-inch turbine meter rated for 5,000 acfh at atmospheric pressure can handle over 30,000 Mscf/d at 1,000 psig. Always verify the meter capacity at the actual operating pressure, not just the rated acfh.

4. Ultrasonic Meters

Ultrasonic flow meters (USMs) have become the preferred technology for large-volume custody transfer measurement on transmission pipelines. AGA Report No. 9 (Measurement of Gas by Multipath Ultrasonic Meters) provides the calculation standard. USMs measure gas velocity by transmitting ultrasonic pulses through the flowing gas and measuring the transit-time difference between upstream and downstream propagation.

Transit-Time Principle

Transit-time velocity measurement: For a single acoustic path at angle θ to the pipe axis: v = (L / 2 cosθ) × (1/tAB − 1/tBA) Where: v = gas velocity along the path L = acoustic path length between transducers θ = path angle relative to the pipe axis tAB = transit time in the downstream direction tBA = transit time in the upstream direction Speed of sound from transit times: c = (L / 2) × (1/tAB + 1/tBA) This provides an independent check: the measured speed of sound must agree with the theoretical value (calculated from composition and conditions per AGA 10) within ± 0.2%. A discrepancy indicates contamination, incorrect gas composition, liquid in the meter, or transducer issues. Volume flow rate: Q = A × Σ(wi × vi) Where: A = meter bore cross-sectional area wi = weighting factor for path i vi = measured velocity on path i

Multipath Configurations

Custody-transfer USMs use multiple acoustic paths (typically 4 to 6) arranged at different positions across the pipe cross-section. Each path samples the velocity at its location, and the paths are combined using numerical integration (Gaussian quadrature or Chebyshev methods) to determine the average flow velocity. More paths provide better integration accuracy, especially for non-ideal velocity profiles.

Configuration Paths Profile Sensitivity Typical Application
Single-path 1 High – assumes axisymmetric profile Flare gas, check metering (not custody transfer)
Dual-path 2 Moderate – detects asymmetry Allocation metering
4-path (chordal) 4 Low – Gaussian integration Custody transfer (AGA 9 compliant)
4+1 or 5-path 4–5 Very low – swirl detection High-accuracy custody transfer
6-path (crossed) 6 Lowest – full profile characterization Premium custody transfer, audit-grade

Diagnostic Capabilities

One of the most significant advantages of USMs over other meter types is their built-in diagnostic capability. The following diagnostics are continuously available:

  • Speed of sound validation: Compare measured SOS against AGA 10 calculated value. Agreement within ±0.2% confirms correct gas composition and clean transducers.
  • Path velocity ratios: Comparison of velocities on inner versus outer paths reveals asymmetric flow profiles that indicate upstream piping disturbances.
  • Signal quality (SNR): Signal-to-noise ratio on each path indicates transducer condition, liquid contamination, or acoustic interference. SNR > 15 dB is typical for healthy operation.
  • Gain and turbulence: Automatic gain control levels and velocity turbulence intensity provide information about flow regime and transducer performance.
  • Profile factor: Ratio of centerline velocity to average velocity indicates the Reynolds number and can detect flow disturbances.
Zero permanent pressure loss: Ultrasonic meters are full-bore, spool-piece designs with no obstruction in the flow path. This means zero permanent pressure loss across the meter, a critical advantage over orifice meters on large-volume trunk lines where even a few psi of pressure drop translates to substantial compression costs. A 30-inch USM on a 500 MMscf/d pipeline that replaces an orifice meter can save $500,000 or more per year in fuel gas alone.

5. Coriolis Meters

Coriolis meters measure mass flow rate directly by exploiting the Coriolis effect on fluid flowing through vibrating tubes. Unlike volumetric meters that require pressure, temperature, and compressibility corrections to convert to standard conditions, a Coriolis meter provides mass flow directly, simplifying the measurement chain and reducing uncertainty sources. AGA Report No. 11 governs Coriolis meter applications for natural gas measurement.

Operating Principle

Coriolis effect: When fluid flows through a vibrating tube, the Coriolis force creates a measurable phase shift (time delay) between the inlet and outlet of the vibrating section. Fc = −2m × (ω × v) Where: Fc = Coriolis force vector m = mass of fluid element ω = angular velocity vector of vibrating tube v = velocity vector of fluid Mass flow rate: ṁ = Ks × Δt / (Kf) Where: ṁ = mass flow rate (lb/s or kg/s) Ks = stiffness factor of the vibrating tubes Δt = measured phase shift (time delay) Kf = shape factor depending on tube geometry Density measurement: ρ = K1 × fn−2 + K2 Where: ρ = fluid density fn = natural frequency of the vibrating tube K1, K2 = calibration constants The tube resonant frequency decreases as the fluid density increases, providing simultaneous density measurement with mass flow.

Performance Characteristics

Parameter Typical Gas Performance NGL/LPG Performance
Mass flow accuracy ± 0.35% to ± 0.5% of reading ± 0.1% to ± 0.15% of reading
Repeatability ± 0.1% to ± 0.2% ± 0.05%
Density accuracy ± 0.5 kg/m3 ± 0.2 kg/m3
Rangeability 10:1 to 20:1 50:1 to 100:1
Pressure drop Moderate (bent-tube designs) Moderate (bent-tube designs)
Max size (gas service) 6" to 12" (practical limit) 6" to 16"

Applications

  • NGL and LPG custody transfer: The primary application for Coriolis meters in midstream. Mass flow measurement eliminates the need for density compensation on liquids whose density changes significantly with temperature and pressure.
  • Compressed natural gas: High-pressure gas applications where mass flow is preferred over volumetric measurement.
  • Fuel gas and flare gas: Low-pressure gas applications where simplicity and self-diagnostics are valued.
  • Multiphase and wet gas: Some Coriolis meter designs can tolerate entrained liquids, although accuracy degrades. Not recommended for high GVF (gas void fraction) streams without specialized algorithms.
Gas density effect on accuracy: Coriolis meter accuracy is fundamentally tied to the ratio of fluid density to tube density. At low gas pressures (low density), the Coriolis force is weak relative to tube stiffness, and measurement uncertainty increases. For gas service, Coriolis meters perform best at pressures above 300 psig where gas density provides adequate signal strength. Below 100 psig, measurement uncertainty may exceed ±1.0% and Coriolis meters become less competitive with turbine or ultrasonic alternatives.

6. Displacement Meters

Positive displacement (PD) meters measure gas flow by mechanically isolating and counting discrete volumes of gas as they pass through the meter. Each revolution or cycle of the measuring element displaces a precisely known volume, making PD meters inherently volumetric devices. They are the standard metering technology for low-pressure gas distribution systems, residential and commercial billing, and some industrial applications.

Rotary Meters

Rotary meter operation: Two figure-eight (or lobe-shaped) impellers rotate in opposite directions within a precision housing. Each revolution captures and transfers a fixed volume of gas from inlet to outlet. Volume per revolution = Vrev (calibration constant) Total volume = Nrev × Vrev Performance: Accuracy: ± 1.0% over 10:1 to 25:1 range Pressure: Up to ANSI 300 (720 psig) Sizes: 2" to 12" Capacity: Up to 15,000 acfh (12" size) Pressure correction: Volume at standard conditions requires P and T correction: Qstd = Qactual × (Pf / Pb) × (Tb / Tf) × (Zb / Zf)

Diaphragm Meters

Diaphragm meters (also called bellows meters) use flexible diaphragms in multiple chambers to alternately fill and empty, creating a reciprocating displacement action. A mechanical index (dial register) counts the cycles and displays total volume. Diaphragm meters are the standard for residential and small commercial gas service worldwide.

Feature Rotary Meter Diaphragm Meter
Pressure range Low to high (up to 720 psig) Low pressure only (< 5 psig typical)
Typical sizes 2" to 12" 3/4" to 4" (residential/commercial)
Accuracy ± 1.0% over rated range ± 1.0% to ± 2.0%
Rangeability 10:1 to 25:1 100:1 or greater
Moving parts Impellers, bearings, gears Diaphragms, valves, linkage
Application Industrial, distribution, allocation Residential, small commercial
Output Pulse + mechanical register Mechanical register (pulse optional)
Displacement meters and low flow: PD meters have an inherent advantage at low flow rates because they physically trap and count gas volumes regardless of velocity. This gives diaphragm meters exceptional rangeability (> 100:1) and makes them ideal for intermittent or highly variable loads like residential heating. However, their mechanical complexity and limited pressure rating make them unsuitable for high-pressure transmission or gathering applications where orifice, turbine, or ultrasonic meters are preferred.

7. Gas Chromatography & BTU Analysis

Gas chromatography is the analytical backbone of natural gas measurement. At every custody transfer point, a gas chromatograph (GC) determines the composition of the gas, from which heating value, relative density, and compressibility factor are calculated. The financial value of the gas transaction depends equally on the accuracy of the GC analysis and the accuracy of the flow meter.

GC Operating Principle

A gas chromatograph separates the components of a gas mixture by passing the sample through a chromatographic column containing a stationary phase (packing material or coating). Each component travels through the column at a different rate depending on its molecular interactions with the stationary phase. As each component exits (elutes from) the column, a detector measures its concentration.

GC analysis per GPA 2261: Components analyzed: C1, C2, C3, iC4, nC4, iC5, nC5, C6+, N2, CO2 Heating value calculation (GPA 2145 / GPA 2172): GHVideal = Σ (yi × HVi) Where: yi = mole fraction of component i (from GC) HVi = ideal gross heating value of component i (from GPA 2145 physical constants table) Real gas correction: GHVreal = GHVideal × (1 / Zb) × Zmix,b Where Zb and Zmix,b account for non-ideal gas behavior at base conditions (small correction, typically < 0.3% for pipeline-quality gas). Relative density (specific gravity): G = Mmix / Mair = Σ(yi × Mi) / 28.9625 Example calculation: 95% CH4 (1,010.0) + 3% C2H6 (1,769.7) + 1% C3H8 (2,516.1) + 0.5% CO2 (0) + 0.5% N2 (0) GHV = 0.95(1010.0) + 0.03(1769.7) + 0.01(2516.1) = 959.50 + 53.09 + 25.16 = 1,037.8 BTU/scf

Online vs. Laboratory Analysis

Feature Online Process GC Laboratory GC
Analysis time 3 to 8 minutes per cycle 15 to 30 minutes per analysis
Components C1–C6+, N2, CO2, H2S (standard) C1–C12+ extended, sulfur speciation
Calibration Automated with reference gas blend Manual with multiple calibration standards
Accuracy (major comp.) ± 0.1 mol% (C1, C2) ± 0.05 mol% (C1, C2)
Heating value accuracy ± 1 to 3 BTU/scf ± 0.5 to 1 BTU/scf
Cost $30,000 – $80,000 installed $80,000 – $200,000+ (instrument only)
Environment Field-hardened enclosure Climate-controlled laboratory

C6+ Characterization

The C6+ fraction reported by a standard GPA 2261 analysis lumps all components heavier than pentane into a single value. For accurate heating value and compressibility calculation, the C6+ fraction must be characterized by its molecular weight and specific gravity. This is typically done by:

  • Assumed characterization: Using default values (M = 86, SG = 0.664 for hexane) when the actual C6+ content is small (< 0.1 mol%).
  • Laboratory characterization: Measuring the C6+ molecular weight and density from a representative liquid sample collected at the sampling point.
  • Extended analysis: Performing a GPA 2286 analysis that resolves individual components through C12+, eliminating the need for C6+ characterization.
BTU accuracy impact: A 1 BTU/scf error in heating value determination on a 100 MMscf/d custody transfer point at $3.00/MMBtu represents approximately $109,500 per year in misallocated revenue. Proper GC calibration with NIST-traceable reference gas blends, correct C6+ characterization, and adherence to GPA 2261 analysis procedures are essential for minimizing heating value measurement error.

8. Gas Sampling

The most accurate GC in the world produces meaningless results if the gas sample does not represent the actual flowing stream. Gas sampling is arguably the weakest link in the measurement chain, and sampling errors of 5 to 10 BTU/scf are common when proper procedures are not followed. GPA 2166 (Obtaining Natural Gas Samples for Analysis by Gas Chromatography) and API MPMS Chapter 14.1 provide the definitive procedures for sample collection and handling.

Sampling Methods

Method Standard Description Best Application
Spot sampling GPA 2166 Single grab sample at a point in time using a sample cylinder Periodic quality checks, lab analysis
Composite sampling GPA 2166 Proportional-to-flow collection over hours or days into an accumulator Custody transfer where composition varies
Online GC GPA 2261 Continuous automated analysis with sample conditioning system Primary custody transfer, process control

Sample Probe and System Design

GPA 2166 sampling probe requirements: 1. Probe location: Center third of pipe cross-section 2. Probe orientation: Facing into the flow (upstream) 3. Minimum distance from upstream disturbance: 5D 4. Sample line: 1/4" or 3/8" OD stainless steel tubing 5. Sample line length: As short as practical (< 25 feet) 6. Sample line heating: Required if Tline approaches hydrocarbon or water dewpoint 7. Flow rate through sample system: Sufficient to maintain turbulent flow and minimize lag time Sample cylinder preparation: - Cleaned and evacuated OR purged with sample gas - Cylinder pressure rating ≥ 1.5 × sample pressure - Cylinder material: stainless steel (DOT 3E-1800 or equivalent) for sulfur-bearing gas - Piston cylinders preferred over fixed-volume cylinders for high-pressure or wet gas sampling Critical rule: Sample pressure and temperature must ALWAYS remain above the cricondentherm (hydrocarbon dewpoint) and water dewpoint throughout the entire sample path. If any point drops below either dewpoint, liquid dropout contaminates the sample and biases the analysis toward lighter components.

Phase Envelope Considerations

The single most common source of sampling error is inadvertent condensation of heavier hydrocarbons in the sample line, probe, or cylinder. This occurs when the sample pressure or temperature falls below the hydrocarbon phase envelope at any point between the sample tap and the analyzer or cylinder. The result is a sample that is depleted in C3+ components, yielding a lower heating value than the actual flowing gas.

  • Pressure reduction: Never reduce sample pressure through a single-stage regulator without heating. Use a heated regulator or multi-stage pressure reduction with intermediate heating.
  • Dead legs: Eliminate dead-end sections in sample tubing that can accumulate liquid condensate.
  • Sample conditioning: Online GC sample systems typically include filters, coalescers, pressure regulators, and flow controllers designed to maintain single-phase gas throughout.
Sampling bias quantification: Studies by GPA and AGA have shown that improper gas sampling can introduce heating value errors of 5 to 20 BTU/scf, far exceeding the uncertainty of either the flow meter or the GC itself. On a 50 MMscf/d stream at $3.00/MMBtu, a 10 BTU/scf sampling bias costs approximately $54,750 per year. Proper probe design, heated sample lines, and adherence to GPA 2166 procedures are the most cost-effective investment in measurement accuracy.

9. Flow Conditioning

All primary flow meters assume a known, well-developed velocity profile in the measurement section. Upstream piping disturbances (elbows, tees, valves, headers) create swirl, asymmetry, and profile distortion that degrade meter accuracy. Flow conditioning addresses these disturbances by either providing sufficient straight pipe for the profile to recover naturally, or by installing a flow conditioner to actively reshape the velocity profile.

Straight-Run Requirements

Minimum upstream straight pipe (diameters D): Orifice Turbine USM (AGA 3) (AGA 7) (AGA 9) Single elbow: 30D 10D 10D Two elbows (same): 50D 15D 15D Two elbows (diff): 80D 25D 20D Partially closed valve: 80D 25D 25D Tee (flow from branch): 80D 30D 30D Header/manifold: 100D+ 50D+ 50D+ Downstream straight pipe (all meters): Minimum 5D downstream of primary element. Note: These are minimum requirements without a flow conditioner. With an approved flow conditioner installed at the proper location, upstream requirements are typically reduced to 10D–15D upstream of the conditioner plus 2D–5D between the conditioner and the meter.

Flow Conditioner Types

Conditioner Type Mechanism Pressure Drop Application
19-Tube bundle Parallel tubes eliminate swirl only Low (0.2 – 0.5 psi) Orifice meters (AGA 3 approved), swirl removal
CPA 50E Perforated plate with specific hole pattern Moderate (0.5 – 2 psi) Orifice and turbine meters, profile reshaping
Laws plate Perforated plate with radially graded holes Moderate (0.5 – 2 psi) All meter types, compact installations
Gallagher conditioner Anti-swirl vanes + perforated plate Moderate (1 – 3 psi) USM installations per AGA 9
Vortab Tab-type vortex generators Low (0.3 – 0.8 psi) General flow conditioning, retrofits
Zanker plate Perforated plate with stepped holes Moderate (0.5 – 2 psi) Orifice meters, profile and swirl control

Flow Conditioner Placement

The location of the flow conditioner relative to the upstream disturbance and the meter is critical. If placed too close to the disturbance, the conditioner cannot effectively reshape the distorted profile. If placed too close to the meter, the conditioner's own wake may not have dissipated.

  • Upstream of conditioner: Minimum 2D to 5D from the last upstream fitting, depending on the conditioner type and manufacturer's recommendations.
  • Between conditioner and meter: Typically 5D to 10D for tube bundles, 2D to 5D for perforated plates (which produce a more uniform profile with shorter recovery length).
  • Tube bundle length: AGA 3 specifies a minimum tube bundle length of 2D with tubes having an inside diameter no larger than D/4.
Conditioner fouling: Flow conditioners with small holes or tubes can accumulate pipeline debris, liquids, and corrosion products over time. Fouling increases the pressure drop across the conditioner, changes the velocity profile it produces, and can shift the meter factor or bias. Regular inspection and cleaning of flow conditioners should be included in the meter station maintenance program, particularly downstream of pig launchers or in dirty-gas service.

10. Measurement Uncertainty

Measurement uncertainty quantifies the range within which the true value is expected to lie, given all known and estimated error sources. For custody transfer gas measurement, the uncertainty analysis follows ISO/IEC Guide 98-3 (GUM – Guide to the Expression of Uncertainty in Measurement) and provides the mathematical framework for determining whether a measurement system meets contractual accuracy requirements.

Uncertainty Propagation

Combined standard uncertainty (RSS method): uc = √[ Σ(ci × ui)2 ] Where: uc = combined standard uncertainty ci = sensitivity coefficient for input i ui = standard uncertainty of input i Sensitivity coefficients: ci = ∂Q / ∂xi × (xi / Q) This is the partial derivative of the flow equation with respect to each input variable, expressing how much the output changes for a unit change in each input. Expanded uncertainty: U = k × uc Where: k = coverage factor (k = 2 for 95% confidence level) Orifice meter uncertainty example (AGA 3): Source ui ci (ciui)2 Discharge coefficient Cd: 0.50% 1.0 0.2500 Orifice bore d: 0.05% 2.0 0.0100 Pipe diameter D: 0.10% 0.3 0.0009 Differential pressure hw: 0.20% 0.5 0.0100 Static pressure Pf: 0.10% 0.5 0.0025 Temperature Tf: 0.10% 0.5 0.0025 Compressibility Z: 0.10% 0.5 0.0025 ________________________ RSS: uc = √(0.2784) = 0.53% Expanded (k=2): U = 1.06%

Error Budget Components

Uncertainty Source Orifice Turbine USM Coriolis
Primary element / meter ± 0.50% ± 0.25% ± 0.20% ± 0.35%
Differential pressure ± 0.20% N/A N/A N/A
Static pressure ± 0.10% ± 0.10% ± 0.10% ± 0.10%
Temperature ± 0.10% ± 0.10% ± 0.10% ± 0.10%
Compressibility (AGA 8) ± 0.10% ± 0.10% ± 0.10% ± 0.10%
Gas composition (GC) ± 0.10% ± 0.10% ± 0.10% ± 0.10%
Heating value ± 0.15% ± 0.15% ± 0.15% ± 0.15%
Installation effects ± 0.30% ± 0.15% ± 0.10% ± 0.05%
Combined (k=2, 95%) ± 1.1% ± 0.7% ± 0.5% ± 0.6%

Compliance Thresholds

Custody transfer contracts specify maximum allowable measurement uncertainty. Common thresholds include:

  • FERC-regulated interstate pipelines: ± 1.0% on volume at the custody transfer point.
  • High-volume interconnects: ± 0.5% negotiated accuracy, typically requiring ultrasonic or turbine meters with online GC.
  • Allocation metering: ± 2.0% to ± 5.0%, depending on contract terms and production volumes.
  • International (ISO 17089): ± 0.5% to ± 1.5% depending on meter class (A, B, or C).
Dominant uncertainty source: In a properly installed and maintained orifice meter system, the discharge coefficient Cd is the dominant uncertainty source, contributing approximately 70% of the total variance. For turbine and ultrasonic meters, the dominant source is typically the meter factor calibration. For all meter types, installation effects (flow profile distortion, vibration, pulsation) are the most common cause of unexplained measurement bias. A comprehensive uncertainty analysis should always include installation effects, not just the manufacturer's stated accuracy.

Meter Technology Selection Guide

Selecting the right meter technology depends on the application, flow rate range, pressure, accuracy requirements, installation constraints, and total cost of ownership. The following comparison summarizes the key factors for the most common gas meter types:

Criterion Orifice Turbine USM Coriolis Rotary PD
Typical accuracy ± 0.5–1.0% ± 0.25–0.5% ± 0.2–0.5% ± 0.35–0.5% ± 1.0%
Rangeability 3:1 to 5:1 10:1 to 25:1 30:1 to 100:1 10:1 to 20:1 10:1 to 25:1
Pressure loss High Moderate None Moderate Moderate
Max pipe size 30" 12" 56"+ 12" 12"
Moving parts None Yes (rotor) None None (vibrating) Yes (impellers)
Diagnostics Limited Limited Extensive Good Limited
Dirty gas tolerance Good Poor Moderate Good Poor
Installed cost (6") $15K–$30K $10K–$25K $50K–$100K $30K–$60K $8K–$20K
Best application General purpose, legacy Mid-range custody transfer Large pipeline CT NGL/LPG, mass flow Distribution, low flow

11. Industry Standards Reference

AGA Reports

Standard Title Scope
AGA Report No. 3 Orifice Metering of Natural Gas and Other Related Hydrocarbon Fluids Orifice meter calculation, installation, and plate inspection requirements
AGA Report No. 7 Measurement of Gas by Turbine Meters Turbine meter calibration, installation, and K-factor determination
AGA Report No. 8 Compressibility Factors of Natural Gas and Other Related Hydrocarbon Gases Z-factor calculation for volume correction at standard conditions
AGA Report No. 9 Measurement of Gas by Multipath Ultrasonic Meters USM calculation, installation, diagnostics, and performance requirements
AGA Report No. 10 Speed of Sound in Natural Gas and Other Related Hydrocarbon Gases SOS calculation for USM validation and diagnostics
AGA Report No. 11 Measurement of Natural Gas by Coriolis Meter Coriolis meter application, installation, and performance for gas

API MPMS (Manual of Petroleum Measurement Standards)

Standard Title Scope
API MPMS Ch. 14.1 Collecting and Handling of Natural Gas Samples for Custody Transfer Sample probe design, cylinder handling, composite sampling
API MPMS Ch. 14.3 Concentric, Square-Edged Orifice Meters (= AGA 3) Orifice meter calculation and installation
API MPMS Ch. 5.6 Measurement of Liquid Hydrocarbons by Coriolis Meters Coriolis meter proving and operation for NGL/LPG
API MPMS Ch. 21.1 Electronic Gas Measurement Flow computer configuration, audit trail, data handling

GPA and ISO Standards

Standard Title Scope
GPA 2145 Table of Physical Constants of Paraffin Hydrocarbons Component heating values, molecular weights, densities
GPA 2166 Obtaining Natural Gas Samples for Analysis by Gas Chromatography Spot and composite sampling procedures
GPA 2172 Calculation of Gross Heating Value, Relative Density, Compressibility Heating value and physical property calculation from composition
GPA 2261 Analysis of Natural Gas and Similar Gaseous Mixtures by GC Standard GC analysis method (C1–C6+, N2, CO2)
GPA 2286 Tentative Method of Extended Analysis for Natural Gas Extended GC analysis through C12+ for HCDP calculations
ISO 17089 Measurement of Fluid Flow – Ultrasonic Meters for Gas International USM standard with meter class definitions
ISO/IEC Guide 98-3 (GUM) Guide to the Expression of Uncertainty in Measurement Framework for measurement uncertainty calculation
ISO 5167 Measurement of Fluid Flow by Differential Pressure Devices International orifice, nozzle, and Venturi meter standard

Ready to use the calculator?

→ Launch Orifice Meter Calculator