Gas Chromatograph Calibration Fundamentals

1. GC Operating Principles

A gas chromatograph separates a gas mixture into its individual components by exploiting differences in how each component interacts with a stationary phase inside a chromatographic column. The separated components are then detected and quantified by one or more detectors. For natural gas analysis, the GC separates methane, ethane, propane, butanes, pentanes, hexanes+, nitrogen, and carbon dioxide, and optionally hydrogen sulfide, oxygen, and other trace components.

Basic GC Process

GC Analysis Sequence: 1. SAMPLE INJECTION A fixed-volume sample loop (typically 0.25 - 1.0 mL) is filled with the gas sample at a known pressure. The loop contents are injected into the carrier gas stream. 2. SEPARATION The carrier gas (helium or hydrogen) carries the sample through the column. Different components travel at different speeds based on their interaction with the stationary phase (column packing or coating). 3. DETECTION As each separated component exits the column ("elutes"), it passes through a detector that produces an electrical signal proportional to the component's concentration. 4. QUANTIFICATION The detector signal is integrated to produce a peak area for each component. Peak areas are converted to concentrations using calibration response factors. Analysis time: C1-C6+, N2, CO2: 5 - 15 minutes (typical process GC) Extended analysis (C9+): 20 - 30 minutes

Carrier Gas

The carrier gas transports the sample through the column without reacting with it. Helium is the traditional carrier gas for natural gas GCs because it has high thermal conductivity (important for TCD sensitivity) and is inert. Hydrogen is increasingly used as an alternative due to helium shortages and cost; hydrogen provides faster analysis times but requires additional safety considerations for explosive atmosphere classification.

Helium Carrier

Standard Choice

High TCD sensitivity, inert, safe. Increasingly expensive. Flow rate typically 20-30 mL/min.

Hydrogen Carrier

Cost-Effective Alternative

Faster separations, lower cost. Requires Class I Div 1/2 considerations for safety. Used with FID detectors.

Nitrogen Carrier

Rarely Used

Poor TCD sensitivity for light hydrocarbons. Only used when N2 is not a required analyte.

2. Detectors (TCD & FID)

Natural gas GCs primarily use two detector types: the Thermal Conductivity Detector (TCD) for universal detection of all components, and the Flame Ionization Detector (FID) for enhanced sensitivity to hydrocarbons.

Thermal Conductivity Detector (TCD)

The TCD measures the difference in thermal conductivity between the pure carrier gas and the carrier gas plus eluted component. Because every gas has a characteristic thermal conductivity, the TCD can detect all components in natural gas including inert gases (N2, CO2). When using helium carrier, which has very high thermal conductivity, even small concentrations of other gases produce measurable signals.

TCD Operating Principle: A heated filament (or thermistor) is maintained at constant power in a flowing carrier gas stream. When an analyte component passes over the filament, the thermal conductivity of the gas mixture changes, altering the filament temperature and therefore its electrical resistance. Signal ~ (TC_carrier - TC_mixture) × Component_Concentration Where TC = thermal conductivity Thermal conductivities (relative to air = 1.0): Helium: 5.84 (carrier gas) Hydrogen: 6.96 Methane: 1.28 Ethane: 0.78 CO2: 0.62 N2: 1.00 Sensitivity: With helium carrier, TCD sensitivity is typically 0.01 - 0.1 mol% for hydrocarbons and inerts.

Flame Ionization Detector (FID)

The FID burns the eluted component in a hydrogen/air flame, producing ions proportional to the number of carbon atoms in the molecule. The FID is roughly 1000 times more sensitive than the TCD for hydrocarbons, but it does not respond to inert gases (N2, CO2, H2O, noble gases) because they do not produce ions when burned.

FID Operating Principle: Component + H2/Air flame → CHO+ ions + electrons The ion current is collected by a polarized electrode and amplified. The signal is approximately proportional to the number of carbon atoms entering the flame per unit time. FID Response: - Equal mass sensitivity for all hydrocarbons (approx.) - Carbon number response: C2 ~ 2x methane per mole - Does NOT detect: N2, CO2, H2O, O2, noble gases, H2S - Cannot be used as sole detector for natural gas analysis FID Sensitivity: Typically 0.001 - 0.01 mol% for hydrocarbons (10-100x better than TCD)

Detector Configuration

Most custody-grade natural gas GCs use a TCD for the main analysis (all components) and may add an FID for improved sensitivity to trace hydrocarbons (C6+, C7+). Some modern process GCs use dual TCD configurations with different column packings to handle both light ends (C1, C2, CO2, N2) and heavy ends (C3+) in parallel channels.

3. Columns & Separation

The chromatographic column is the heart of the GC. It separates the gas mixture based on differences in how strongly each component interacts with the stationary phase. Natural gas GCs typically use packed columns rather than capillary columns because packed columns handle the larger sample volumes needed for process analysis and are more robust in field conditions.

Common Column Configurations

Column Type	Stationary Phase	Components Separated	Typical Dimensions
Molecular Sieve 5A	Zeolite adsorbent	O2, N2, CH4 (permanent gases)	6-12 ft × 1/8 in
Silica Gel / Porapak Q	Porous polymer	CO2, C2, C3 (acid gas + light HC)	6-12 ft × 1/8 in
OV-101 / DC-200	Methylsilicone on Chromosorb	C4, C5, C6+ (heavier HC)	6-20 ft × 1/8 in
Alumina PLOT	Al2O3 on capillary	C1-C6+ (single column approach)	30-50 m × 0.53 mm

Column Switching (Valve-Based)

Natural gas GCs use multi-column configurations with valve switching to achieve the required separation of all components. A typical GPA 2261 analysis uses two or three columns in series with backflush and heart-cut valves to route components through the optimal column for separation. Column switching is controlled by precise timing based on the known elution order of the calibration gas components.

Column aging: Over time, columns degrade due to contamination from liquids, sulfur compounds, and heavy hydrocarbons in the sample. Column performance should be verified during each calibration by checking peak shape, resolution, and retention time stability. Replace columns when resolution drops below the minimum required for separation of critical pairs (e.g., CO2/C2H6).

4. Calibration Gas Standards

Calibration gas is a certified reference mixture with precisely known concentrations of each component. The accuracy of every GC analysis is directly traceable to the calibration gas. For custody transfer applications, calibration gas must be a primary reference standard or traceable to a primary standard from an accredited laboratory.

Calibration Gas Requirements

GPA 2261 Calibration Gas Requirements: 1. Certified by an ISO/IEC 17025 accredited laboratory 2. Component concentrations spanning the expected range of the sample (within 50-150% of expected values) 3. Uncertainty of each component < 1% relative (typical) 4. Cylinder stability: 2-year shelf life for gravimetric blends 5. Balance gas: typically nitrogen or methane Typical Natural Gas Calibration Blend: Methane: 85.00 mol% (balance) Ethane: 5.50 mol% (± 0.02) Propane: 3.00 mol% (± 0.02) i-Butane: 0.80 mol% (± 0.01) n-Butane: 1.50 mol% (± 0.01) i-Pentane: 0.40 mol% (± 0.01) n-Pentane: 0.30 mol% (± 0.01) Hexanes+: 0.20 mol% (± 0.01) Nitrogen: 1.80 mol% (± 0.02) CO2: 1.50 mol% (± 0.02) Total: 100.00 mol%

Calibration Gas Handling

Storage temperature: Store calibration gas cylinders in a temperature-controlled environment (60-100 °F). Extreme temperatures can cause component condensation (especially C5+ and C6+) that changes the gas-phase composition.
Cylinder pressure: Do not use calibration gas below 200 psig cylinder pressure. As pressure drops, heavier components preferentially remain in the liquid phase, enriching the gas-phase composition with lighter components.
Regulators: Use dedicated regulators for calibration gas to prevent cross-contamination. Use stainless steel diaphragm regulators, not brass.
Line purging: Purge calibration gas supply lines for at least 5 minutes before calibration to ensure the sample reaching the GC is representative of the cylinder contents.

Traceability chain: For custody transfer, the calibration gas must have a certificate of analysis from an accredited lab, with stated uncertainty for each component. The certificate number, lot number, and expiration date must be recorded in the calibration log. This traceability chain is audited during custody transfer measurement audits.

5. Calibration Procedure

Calibration establishes the relationship between detector response (peak area) and component concentration. For a properly operating GC, this relationship should be linear over the expected concentration range. GPA 2261 specifies a minimum calibration procedure that includes verification of component identification, response factor calculation, and repeatability checks.

Step-by-Step Calibration

GPA 2261 Calibration Procedure: 1. SYSTEM VERIFICATION - Verify carrier gas supply pressure and flow rate - Check detector baseline stability (drift < 1% full scale/hr) - Verify oven temperature stability (± 0.1 °C) - Verify sample system integrity (no leaks) 2. CALIBRATION GAS ANALYSIS (Minimum 3 runs) - Inject calibration gas through the sample loop - Record the chromatogram - Verify all expected peaks are present and resolved - Integrate peak areas for each component 3. RESPONSE FACTOR CALCULATION For each component i: RF_i = (Peak Area_i) / (Known Concentration_i in mol%) Normalize: RF_normalized_i = RF_i / RF_reference (reference component is usually methane or nitrogen) 4. REPEATABILITY CHECK Compare response factors from 3 consecutive runs: CV (coefficient of variation) must be < 0.5% for major components and < 2% for minor components 5. VERIFICATION Analyze calibration gas as an unknown using the new response factors. All components must agree with certificate values within stated tolerance: - Major components (> 1 mol%): ± 0.5% relative - Minor components (0.1 - 1 mol%): ± 2% relative - Trace components (< 0.1 mol%): ± 5% relative 6. DOCUMENTATION Record: date, technician, cal gas lot#, RF values, repeatability CV, verification results, pass/fail

Calibration Frequency

Application	Calibration Interval	Verification Interval
Custody transfer (AGA/API)	Monthly or per contract	Daily or per analysis cycle
Process monitoring	Quarterly	Weekly
Environmental monitoring	Per regulatory requirement	Per regulatory requirement

6. Response Factor Calculation

The response factor (RF) quantifies the detector's sensitivity to each component. It converts raw peak area (or peak height) to concentration. Response factors must be stable over time; significant changes indicate detector problems, column degradation, or system leaks.

Response Factor Calculation: Absolute Response Factor: RF_i = A_i / C_i Where: RF_i = response factor for component i (area counts per mol%) A_i = peak area for component i (area counts) C_i = known concentration of component i (mol%) Relative Response Factor: RRF_i = RF_i / RF_ref Where RF_ref = response factor of the reference component (usually methane for TCD, propane for FID) Using RF to calculate unknown concentration: C_unknown_i = A_unknown_i / RF_i Normalization: Final mol% = (C_raw_i / Sum_of_all_C_raw) × 100 This ensures the total composition sums to exactly 100%. Example: Calibration gas ethane = 5.50 mol% Peak area for ethane = 275,000 counts RF_ethane = 275,000 / 5.50 = 50,000 counts per mol% Unknown sample ethane peak = 300,000 counts C_ethane = 300,000 / 50,000 = 6.00 mol% (before normalization)

Response Factor Stability

Response factors should be monitored over time by maintaining a control chart. A stable GC will show RF values that vary by less than 1-2% relative from calibration to calibration. Sudden changes in RF indicate problems that must be investigated before the GC is used for custody transfer analysis.

RF Change	Possible Cause	Action
All RFs decreased uniformly	Sample loop leak, low sample pressure	Leak test sample system, verify loop pressure
Heavy component RFs decreased	Column contamination, cold spot in sample line	Bake out column, heat trace sample lines
Single component RF changed	Co-elution, peak integration error	Check resolution, adjust integration parameters
All RFs increased uniformly	Detector sensitivity increase, carrier flow decrease	Verify carrier gas flow, check detector settings

7. Linearity Verification

Linearity verification confirms that the detector response is proportional to concentration over the entire measurement range. A linear detector allows accurate analysis even when the sample composition differs significantly from the calibration gas. GPA 2261 requires linearity verification as part of GC qualification.

Linearity Verification Procedure: 1. Obtain at least 3 calibration gas blends with different concentrations spanning the expected range (e.g., 50%, 100%, and 150% of nominal) 2. Analyze each blend in triplicate 3. Plot peak area vs. concentration for each component 4. Calculate the correlation coefficient (R^2): R^2 = [n × Sum(xy) - Sum(x) × Sum(y)]^2 / [n × Sum(x^2) - (Sum(x))^2] × [n × Sum(y^2) - (Sum(y))^2] Acceptance criteria: R^2 ≥ 0.999 for major components (> 1 mol%) R^2 ≥ 0.995 for minor components (0.1 - 1 mol%) If linearity fails, the detector or column may be degraded and must be serviced before use. Alternative: Multi-point calibration with polynomial curve fit (quadratic) for mildly non-linear detectors, if approved by the measurement agreement.

Practical note: Most custody transfer GCs are calibrated with a single calibration gas (one-point calibration) and rely on verified linearity to extend accuracy across the expected composition range. If the sample composition varies significantly from the calibration gas (e.g., rich gas vs. lean gas), a two-point or multi-point calibration may be needed.

8. Repeatability Requirements

Repeatability is the variation in results when the same sample is analyzed multiple times under the same conditions. It is the most fundamental measure of GC precision and must meet specified limits for the results to be considered valid for custody transfer.

GPA 2261 Repeatability Limits: For consecutive analyses of the same gas: Component Range (mol%) Max Allowed Difference > 10 mol% ± 0.10 mol% absolute 1 - 10 mol% ± 0.05 mol% absolute 0.1 - 1 mol% ± 0.02 mol% absolute 0.01 - 0.1 mol% ± 0.01 mol% absolute Coefficient of Variation (CV): CV = (Standard Deviation / Mean) × 100% Major components: CV < 0.5% Minor components: CV < 2.0% Trace components: CV < 5.0% Heating Value Repeatability: Consecutive Btu/scf calculations from GC analysis should agree within ± 1 Btu/scf for pipeline-quality natural gas (nominally 1000-1050 Btu/scf).

Reproducibility vs. Repeatability

Repeatability measures variation within a single GC under identical conditions. Reproducibility measures variation between different GCs, different operators, or different laboratories analyzing the same gas. Reproducibility limits are approximately 2 to 3 times the repeatability limits. When two custody transfer GCs at the same location produce different results, the reproducibility limits determine whether the difference is within acceptable measurement uncertainty.

Heating Value Impact

The primary purpose of GC analysis at a custody transfer point is to determine the heating value (Btu/scf) of the gas for billing purposes. A typical natural gas composition change of 0.1 mol% in ethane or propane changes the heating value by approximately 1-3 Btu/scf. At a flow rate of 100 MMscf/d and $3.00/MMBtu, a 3 Btu/scf measurement error represents approximately $900/day or $328,000/year in billing error.

Financial impact: GC repeatability directly affects the financial accuracy of custody transfer. For a large-volume metering station measuring 500 MMscf/d, even a 1 Btu/scf systematic error in heating value represents over $500,000/year in billing discrepancy. This is why GPA 2261 and API MPMS Chapter 14.1 set rigorous repeatability requirements.

9. Common Troubleshooting

Field GCs operate in harsh environments and require regular maintenance and troubleshooting. The most common problems fall into categories of sample system issues, column degradation, detector problems, and electronic/software errors.

Diagnostic Table

Symptom	Probable Cause	Corrective Action
All peaks smaller than expected	Sample loop leak, low sample pressure, carrier flow too high	Leak test with soap solution, verify sample loop pressure, check carrier flow
Heavy peaks missing or reduced	Cold spot in sample line, column contamination	Heat trace sample lines, bake out or replace column
Peaks tailing or broad	Column degradation, dead volume in fittings	Replace column, check and tighten all fittings
Baseline drift	Detector contamination, carrier gas impurity, column bleed	Clean detector, replace carrier gas trap, condition column
Retention times shifted	Carrier flow change, oven temperature change, column aging	Verify carrier flow and oven temp, recalibrate retention times
Ghost peaks / extra peaks	Sample carryover, contaminated sample lines, outgassing	Purge sample system, run blank analysis, check for contamination
Poor repeatability	Valve timing issue, sample pressure variation, electrical noise	Check valve actuation timing, regulate sample pressure, check grounding
Methane/ethane not resolved	Molecular sieve column saturated or contaminated with CO2	Bake out molecular sieve at 300 °C, replace if needed

Preventive Maintenance

Sample system: Replace coalescing filters monthly. Check for liquid accumulation in sample probes and lines daily. Verify heat trace temperature on sample lines.
Carrier gas: Replace moisture and hydrocarbon traps per manufacturer schedule (typically annually or when indicator changes color).
Detector: Clean TCD filaments annually. Replace FID jet and igniter as needed. Verify detector temperature.
Valves: Check valve switching timing and actuation. Replace valve seals per manufacturer schedule (typically 1-2 years).
Column: Monitor retention times and resolution. Condition columns periodically by temperature programming to remove accumulated heavy contaminants.

10. Standards & References

Standard	Title	Relevance
GPA 2261	Analysis of Natural Gas and Similar Mixtures by GC	Primary GC analysis standard for natural gas
GPA 2166	Obtaining Natural Gas Samples for Analysis by GC	Proper sampling procedures upstream of GC
GPA 2172	Calculation of Gross Heating Value, Relative Density, etc.	Converting GC composition to heating value
ISO 6974	Natural Gas — Determination of Composition by GC	International GC analysis standard
ISO 6976	Natural Gas — Calculation of Calorific Values	International heating value calculation
AGA Report No. 5	Natural Gas Energy Measurement	Energy determination from composition
AGA Report No. 8	Compressibility Factors (DETAIL equation)	Z-factor from GC composition for flow calculation
API MPMS Ch. 14.1	Collecting and Handling of Natural Gas Samples	Sampling requirements for custody transfer GC
ASTM D1945	Standard Test Method for Analysis of Natural Gas by GC	Alternative GC analysis standard