Worked Example: Custody Transfer Gas Metering Uncertainty & Annual Financial Billing Exposure (ISO 5167, AGA 3 & GUM)

1. Problem Statement & Custody Transfer Context

An interstate natural gas pipeline custody transfer station delivers an average of $65.0\text{ MMSCFD}$ ($76,690\text{ Sm}^3/\text{h}$) to a major regional combined-cycle gas turbine (CCGT) power station. Gas is commercially settled on total energy content ($\text{MMBtu}$) at a contractual tariff rate of $\$3.65\text{ / MMBtu}$.

The fiscal metering facility utilizes a dual-chamber orifice meter tube designed per ISO 5167-2 / AGA Report No. 3, equipped with smart digital transmitters and an on-line process gas chromatograph (GC).

 [Transmission Pipeline] ──> [Orifice Meter: dP, P, T, GC] ──> [CCGT Power Plant]
  Flow = 65.0 MMSCFD           Operating P = 55.0 barg           Contract = $3.65 / MMBtu
                               Operating T = 25.0 °C             Turnover = $89.8 Million/yr

Operating Parameters & Gas Assay

• Daily Volumetric Throughput ($Q$): $65.0\text{ MMSCFD}$ ($1.84 \times 10^6\text{ Sm}^3/\text{day}$)
• Static Line Pressure ($P$): $55.0\text{ barg}$ ($56.01\text{ bar abs} = 812.4\text{ psia}$)
• Flowing Temperature ($T_f$): $25.0\text{ }^\circ\text{C}$ ($298.15\text{ K}$)
• Gas Gross Calorific Value ($GCV$): $38.60\text{ MJ/Sm}^3$ ($1,036.5\text{ Btu/scf}$)
• Gas Energy Factor: $1.0365\text{ MMBtu / MSCF} \implies 67,370\text{ MMBtu/day}$
• Commercial Settlement Tariff: $\$3.65\text{ / MMBtu}$
• Orifice Beta Ratio ($\beta = d/D$): $0.580$

Instrumentation Calibration Uncertainty Assays (Expanded at $k = 2$, 95% Confidence)

Engineering Objectives

1. Determine dimensionless partial sensitivity coefficients ($c_i$) for all input variables per ISO/IEC Guide 98-3 (GUM).
2. Calculate the combined standard uncertainty ($u_c$) and expanded uncertainty ($U_{95\%}$) for volumetric flow rate.
3. Propagate heating value uncertainty to compute the expanded Energy Billing Uncertainty ($U_{\text{energy}}$).
4. Quantify the annual commercial billing turnover and annual dollar financial risk exposure ($\pm \$ / \text{year}$).
5. Construct the Pareto variance budget to identify the primary target for metrological calibration investment.

2. Metrological Modeling & GUM Sensitivity Derivation

2.1 The ISO 5167 Orifice Mass and Volume Flow Equations

The standard volumetric flow rate through an orifice plate is:

$$Q_v = C_d \cdot E_v \cdot Y \cdot \frac{\pi}{4} d^2 \cdot \sqrt{\frac{2 \cdot \Delta P}{\rho_1}} \cdot \frac{\rho_1}{\rho_{\text{std}}}$$

Expressing density in terms of the real gas law $\rho = \frac{P \cdot M_w}{Z \cdot R \cdot T}$:

$$Q_v \propto C_d \cdot d^2 \cdot \left[ \frac{\Delta P \cdot P}{T_f \cdot Z \cdot \rho_{\text{std}}} \right]^{0.5}$$

2.2 Dimensionless Sensitivity Coefficients ($c_i$)

Taking the natural logarithm and differentiating with respect to each independent variable yields the dimensionless sensitivity coefficients:

$$\ln(Q_v) = \ln(C_d) + 2\ln(d) + 0.5\ln(\Delta P) + 0.5\ln(P) - 0.5\ln(T_f) - 0.5\ln(Z) - 0.5\ln(\rho_{\text{std}})$$

Therefore: * Differential Pressure Sensitivity: $c_{\Delta P} = \frac{\partial \ln Q_v}{\partial \ln \Delta P} = \mathbf{+0.50}$ * Static Pressure Sensitivity: $c_P = \frac{\partial \ln Q_v}{\partial \ln P} = \mathbf{+0.50}$ * Temperature Sensitivity: $c_T = \frac{\partial \ln Q_v}{\partial \ln T} = \mathbf{-0.50}$ * Gas Density / GC Sensitivity: $c_{\rho} = \frac{\partial \ln Q_v}{\partial \ln \rho_{\text{std}}} = \mathbf{-0.50}$ * Compressibility Factor Sensitivity: $c_Z = \frac{\partial \ln Q_v}{\partial \ln Z} = \mathbf{-0.50}$ * Discharge Coefficient Sensitivity: $c_{Cd} = \frac{\partial \ln Q_v}{\partial \ln C_d} = \mathbf{+1.00}$

2.3 GUM Variance Propagation Equation

The combined relative standard uncertainty is the root-sum-square:

$$u_c^2(Q_v) = \sum_{i=1}^N c_i^2 \cdot u^2(x_i)$$

3. Step-by-Step Numerical Calculations

Step 1: Input Uncertainties in Relative Percentage

1. Temperature: $$u(T_{\text{rel}}) = \frac{U_T / 2}{T_{\text{abs}}} \times 100\% = \frac{0.125\text{ K}}{298.15\text{ K}} \times 100\% = \mathbf{0.0419\%} \quad (U_{T, \text{rel}} = 0.0839\%)$$

2. Variance Contributions ($c_i^2 \cdot U_i^2$):

3. Discharge Coeff ($C_d$): $(1.0)^2 \times (0.50\%)^2 = 0.250000$
4. Gas Chromatography ($\rho_{\text{std}}$): $(0.5)^2 \times (0.40\%)^2 = 0.25 \times 0.1600 = 0.040000$
5. Differential Pressure ($\Delta P$): $(0.5)^2 \times (0.10\%)^2 = 0.25 \times 0.0100 = 0.002500$
6. Static Pressure ($P$): $(0.5)^2 \times (0.10\%)^2 = 0.25 \times 0.0100 = 0.002500$
7. Compressibility ($Z$): $(0.5)^2 \times (0.10\%)^2 = 0.25 \times 0.0100 = 0.002500$
8. Temperature ($T$): $(0.5)^2 \times (0.0839\%)^2 = 0.25 \times 0.00703 = 0.001758$

Step 2: Combined & Expanded Volumetric Uncertainty

• Total Expanded Volumetric Variance: $$\sigma_{\text{vol}}^2 = 0.250000 + 0.040000 + 0.002500 + 0.002500 + 0.002500 + 0.001758 = \mathbf{0.299258}$$

• Expanded Volumetric Uncertainty ($k = 2$): $$U_{95\%}(Q_v) = \sqrt{\sigma_{\text{vol}}^2} = \sqrt{0.299258} = \mathbf{0.547\%}$$

• Combined Standard Uncertainty: $$u_c(Q_v) = \frac{0.547\%}{2} = \mathbf{0.274\%}$$

Step 3: Energy Billing Uncertainty Propagation

Energy delivery is: $E = Q_v \cdot GCV$. Since volume and GC heating value are measured independently, the combined energy variance is:

$$\sigma_{\text{energy}}^2 = U_{95\%}^2(Q_v) + U_{95\%}^2(GCV) = (0.547)^2 + (0.250)^2 = 0.299258 + 0.062500 = \mathbf{0.361758}$$

$$U_{\text{energy, 95\%}} = \sqrt{0.361758} = \mathbf{0.601\%}$$

Step 4: Commercial Financial Exposure Quantification

• Daily Energy Throughput: $$E_{\text{day}} = 65.0\text{ MMSCF} \times 1,036.5\text{ MMBtu/MMSCF} = \mathbf{67,372.5\text{ MMBtu/day}}$$

• Annual Energy Throughput: $$E_{\text{annual}} = 67,372.5 \times 365.25 = \mathbf{24,607,800\text{ MMBtu/year}}$$

• Commercial Billing Turnover: $$\text{Daily Turnover} = 67,372.5\text{ MMBtu} \times \$3.65 = \mathbf{\$245,910\text{ / day}}\text{Annual Turnover} = 24,607,800\text{ MMBtu} \times \$3.65 = \mathbf{\$89,818,470\text{ / year}} \quad (\approx \$89.8\text{M / yr})$$

• Commercial Financial Billing Exposure ($\pm \$$): $$\text{Daily Financial Risk} = \$245,910 \times \left(\frac{0.601}{100}\right) = \mathbf{\pm \$1,478\text{ / day}}\text{Annual Financial Risk} = \$89,818,470 \times \left(\frac{0.601}{100}\right) = \mathbf{\pm \$540,240\text{ / year}}$$

4. Pareto Uncertainty Budget Analysis

Metrological Insight

• The orifice plate discharge coefficient accounts for $83.5\%$ of the total measurement variance. This is inherent to the empirical Reader-Harris/Gallagher correlation limits in ISO 5167.
• Technology Upgrade Comparison: Upgrading this station to a Multi-Path Ultrasonic Flow Meter (AGA Report No. 9) with high-pressure calibration reduces the primary element uncertainty from $\pm 0.50\%$ down to $\pm 0.20\%$.
• New $U_{95\%}(Q_v) \to \pm 0.320\%$
• New $U_{\text{energy}} \to \pm 0.406\%$
• Annual Financial Exposure Drops from $\pm \$540,240$ to $\pm \$364,660$—an annual risk mitigation of $\$175,580\text{ / year}$, easily justifying the $\$85,000$ CAPEX of an ultrasonic meter retrofit in under six months!

5. ChemProCal Engine Verification & Cross-Check

from engineering_core.gas_thermo import calculate_metering_uncertainty

res = calculate_metering_uncertainty(
    meter_type='orifice',
    flow_rate_mmscfd=65.0,
    operating_pressure_barg=55.0,
    operating_temp_c=25.0,
    gas_gcv_mj_sm3=38.60,
    gas_price_usd_per_mmbtu=3.65,
    u_diff_pressure_pct=0.10,
    u_static_pressure_pct=0.10,
    u_temperature_deg_c=0.25,
    u_gas_chromatograph_pct=0.40,
    u_z_factor_pct=0.10,
    u_meter_factor_pct=0.50,
    u_gcv_pct=0.25
)

Verification Matrix

6. Recommendations for Fiscal Measurement Engineers

1. Focus on Gas Chromatograph Linearity: While pressure transmitters are calibrated to $\pm 0.1\%$, calibration drift on heavy hydrocarbons in the online GC disproportionately corrupts standard density and heating value. Daily automated calibration with certified gravimetric blend bottles is mandatory.
2. Monitor Beta Ratio Ranges: Ensure the orifice beta ratio remains between $0.20 \le \beta \le 0.60$. Above $\beta = 0.60$, the ISO 5167 uncertainty on $C_d$ jumps from $\pm 0.50\%$ to $\pm 0.75\%$, increasing financial risk by $50\%$.
3. Quantify Financial Risk in Tariff Audits: When negotiating custody transfer allocation agreements, always express metrological tolerances in annual dollar exposure ($\pm \$ / \text{yr}$) rather than abstract percentages. This aligns corporate commercial management with instrumentation engineering priorities.