Figure 1. Typical chromatogram of a natural gas under 3,1.5,1/4 columns. A heating procedure was adopted with an initial temperature of 25 °C (65 min), heating at 10 °C min–1 to 120 °C (20 min), then to 200 °C. (Sun et al., 2023)
Define components and calculation
A routine natural-gas composition may include methane, ethane, heavier hydrocarbons, nitrogen and carbon dioxide. Depending on purpose, oxygen, hydrogen, helium or sulfur compounds may also matter. Specify the basis for mole fraction, dry or wet gas, and the intended calculated property such as heating value or density. The component list determines column and detector design; a generic "gas composition" request is incomplete. For process control, determine whether major components or trace impurities govern the decision. Major components require robust calibration and stable response over broad concentration ranges. Trace sulfur requires low-background sampling and suitable selective detection; the main-component detector may not have adequate sensitivity. Check the current applicable method and contractual specification rather than inventing a universal gas limit.
Preserve a representative gas sample
Gas composition can change through condensation, adsorption, diffusion or leakage before injection. Select a sampling location and pressure/temperature conditions that preserve one gas phase where required. Water or hydrocarbon condensate can alter the measured composition. Use appropriate containers and transfer lines, purge adequately and document time between sampling and analysis. A sample cylinder is part of the method, not just shipping equipment. Check seals and lines for leaks and material interactions. Sulfur compounds at trace levels may adsorb onto reactive surfaces, so a suitable inert flow path and recovery challenge are necessary. A major-component standard does not reveal trace loss. Include field and transfer blanks where relevant, and prevent cross-contamination from a high-concentration calibration mixture.
Choose column selectivity
Many gas methods use more than one separation path, with valves or column switching to isolate permanent gases, carbon dioxide and hydrocarbon groups. Stationary-phase and column geometry should resolve the components that must be reported while fitting the instrument's pressure and cycle-time limits. Some analytes may require a molecular-sieve-type stationary phase, but water and carbon dioxide handling must be designed carefully because they can alter retention or be strongly retained. Packed and capillary columns have different capacity, efficiency and switching behavior. A column with good methane/ethane resolution may not separate every permanent-gas pair. Model the actual critical pairs, then verify them using calibrated mixtures. Avoid extrapolating from a hydrocarbon-only demonstration to a mixed gas with oxygen, nitrogen and carbon dioxide.
| Measurement | Separation or detection concern | Useful control |
|---|---|---|
| Major hydrocarbons | Peak resolution and wide response range | Certified multicomponent gas mixture |
| Permanent gases | Appropriate column and detector response | Species-specific calibration |
| Carbon dioxide | Retention and switching path | Check valve timing and mass balance |
| Trace sulfur | Surface loss and low background | Inert-path recovery and blank |
Match the detector to components
A thermal-conductivity detector can respond to many permanent gases and hydrocarbons when the carrier-gas contrast is suitable. A flame-ionization detector is highly responsive to many organic compounds but does not directly quantify nitrogen or carbon dioxide under conventional operation. Dedicated sulfur detection can improve selectivity at low levels, subject to method-specific performance. A single detector may therefore require parallel channels or additional measurements.
Detector-specific calibration
Calibration gases should cover analytes and concentration ranges, with composition and uncertainty documented. Account for carrier-gas choice, detector linearity and switching timing. Verify response factors and calibration stability rather than assuming all hydrocarbons generate equal peak area per mole. A certified standard supports calibration only for components it actually contains and at relevant levels.
Control injection and switching
Fixed-volume sampling valves improve repeatability when pressure and temperature are controlled. Purge the loop sufficiently to avoid memory from the previous gas. Pressure differences between standard and sample can change the amount introduced; equalize or correct according to the instrument method. Document valve states and timing because premature switching can split a peak between flow paths. Carryover can arise in regulators, lines or a sulfur-specific path. Run blanks after high-level mixtures. Monitor retention-time shifts, peak widths and baselines to identify contamination or changing column conditions. Gas leaks may present as noisy baselines or abnormal oxygen/nitrogen peaks, but confirm with a separate leak test. Never use a chromatogram alone as the sole integrity check.
Calculate and qualify composition
Integrate resolved peaks by a consistent rule, apply validated response factors, and review closure of the reported component sum. Normalizing every run to 100% can mask a missing component or an analytical problem; examine unnormalized results first. State treatment of components below reporting limits, water and unmeasured higher hydrocarbons. If heating value or density is calculated, use the specified reference conditions and calculation convention. Replicate injections assess instrument precision, while independent samples assess sampling variability. Include continuing calibration checks and a composition QC mixture. Challenge the method with realistic changes in carbon dioxide and heavier hydrocarbons, because the critical separation may change with matrix composition. If an atypical gas contains hydrogen, helium or reactive species, reassess detector and material compatibility.
Distinguish natural gas from other mixtures
A natural-gas method can be a starting point for biogas, industrial fuel gas or process gas, but sample composition may extend beyond its validated scope. High water, oxygen, hydrogen or corrosive species may require special handling and different column or detector channels. The reporting basis may change from pipeline specification to process control. Verify the method's range and hardware compatibility for each new mixture. Do not assume a chromatogram without peaks means absence of a component; some detectors have weak or no response for particular gases. A sample can also lose trace analyte on an unsuitable transfer surface. Document each component's detector, calibration range and reporting limit. When a new target appears, add it to the method explicitly rather than reinterpreting an existing unidentified signal.
Site services and column selection
The site's natural gas analysis and gas analysis services align with composition and trace-component questions. Custom GC columns and custom silica capillary tubes may support a specified separation or flow-path requirement. Confirm available phases and item-level limits before making a selection. Provide component list, concentration intervals, pressure, moisture, sampling container, detector and intended standard. Define the unresolved pair or matrix problem in existing chromatograms. Column choice must be verified in the assembled method with valves, detectors and sampling hardware.
Establish representative calibration
Use a gas standard with assigned composition and uncertainty for the components and range of interest. One standard cannot reliably span both a dominant methane signal and a trace sulfur target if the detector responses and concentration windows differ greatly. Prepare separate calibration channels or levels where justified. Compare response factors over the claimed range and include a continuing check after a defined number of samples. Calibration mixtures should be handled like samples: connect with suitable regulators and purge the transfer line. A standard in a dry cylinder may behave differently from a wet process gas that condenses in a cold line. Verify that sample conditioning preserves the analytes to be measured. If water is intentionally removed, state whether reported composition is dry-basis and how that conversion was made. When calibrating calculated heating value, remember that GC measures composition while the reported property derives from composition and a specified calculation standard. Composition bias propagates into the calculated number. Keep the calculation version and reference conditions with the report.
Address trace sulfur separately
Trace sulfur compounds can be reactive or adsorptive. Select a sampling and transfer path compatible with the specific sulfur species, minimize uncontrolled contact time and test recovery from cylinder through detector. A sulfur-selective detector may deliver the needed sensitivity, but it requires its own calibration, blank and performance verification. Check whether different sulfur molecules have equal or compound-dependent response under the method. A general hydrocarbon channel will not necessarily detect hydrogen sulfide, mercaptans or other sulfur compounds at the required level. When a contractual specification names a sulfur measure, clarify whether it refers to total sulfur or individual compounds. Those are different measurands. If a sample contains condensate or moisture, assess whether the target partitions away from the gas phase before analysis. Run a high-to-low sequence challenge to detect sulfur carryover. A blank cylinder or line is not automatically sulfur-free; qualify it at the method's reporting level.
Verify peak identity in mixed gases
Unexpected peaks can arise from a new process component, calibration impurity, valve timing or a leak. Compare retention on the appropriate column path and use a reference mixture containing the suspected species. If identity is material to a quality decision, employ a second column condition or detector where available. Do not assign a component from retention time alone when multiple gases coelute under the method. For oxygen and nitrogen, a leak can mimic sample composition change. Inspect connections and run a leak check; compare field and laboratory samples. A helium or hydrogen component may require an appropriate carrier gas and detector configuration to be quantified reliably. Detector suitability depends on the contrast between carrier and analyte, not just the presence of a chromatographic peak. Record how unresolved peaks are treated in component totals. Reporting an "unknown" term is more transparent than silently adding it to a nearby hydrocarbon.
Maintain column and valve performance
Trend a reference mixture's critical-pair resolution, retention and response across the instrument lifecycle. Moisture, heavy hydrocarbons or sulfur can change adsorption surfaces and cause tailing or delayed elution. Follow a column conditioning and cleaning procedure that is compatible with the stationary phase and valve system. A rising retention time may result from flow or pressure change rather than stationary-phase chemistry alone. Replace a switching-valve seal or sample-loop component only with an appropriately rated material and reverify loop volume and timing. A small geometry change can alter the amount injected and the fraction routed to each column. For a transferred method, supply valve timing tables, flow diagrams, gas standards and integration examples along with a column specification. These details are essential to reproduce a reported composition.
Check sampling temperature and phase behavior
If a line cools below the dew point, heavier hydrocarbons and water may condense before the sample loop. The gas reaching the chromatograph then differs from the process gas. Record pressure, temperature and any conditioning upstream of the loop. Heating a line may prevent condensation but must remain within the sampling system's material limits. Where two phases already exist in the process, specify which phase the reported gas composition represents. A highly repeatable GC injection cannot compensate for an unrepresentative sampling point.
Document the calculation audit trail
Keep the raw peak areas, calibration factors, component assignments, unnormalized mole fractions and any renormalized result. State how an unresolved C6-plus fraction, water or components below the reporting limit are treated. For a heating-value result, record the property calculation standard and reference pressure and temperature. If a new gas source introduces components outside the calibration mixture, suspend the calculated claim until the method is extended. This avoids a precise-looking derived value built on an incomplete composition.
Qualify a new sampling location
Changing the sample tap or pressure-reduction point can alter composition even when the GC method stays fixed. Compare simultaneous samples from old and new locations under stable process conditions. Examine condensation, purge volume, line length and trace-gas recovery. If the difference is systematic, establish which location represents the intended process stream. Treat sampling design as a controlled part of the composition method, with a documented diagram and maintenance record.
Connect this guide to a chromatography project
Share the sample matrix, target list, analytical goal, existing extraction, chromatograms, required reporting range and instrument configuration. We can help evaluate separation materials and relevant analytical support.
References
- ASTM International. D1945, Standard Test Method for Analysis of Natural Gas by Gas Chromatography.
- ISO. ISO 6974-1, Natural gas—Determination of composition and associated uncertainty by gas chromatography.
- ASTM International. D5504, Sulfur Compounds in Natural Gas and Gaseous Fuels by GC and Chemiluminescence.
- Sun T, Cao J, Qiu H, et al. Investigation and optimization of methane purification method for natural gas by two-column gas chromatography: A preliminary test for doubly substituted isotopologue (13CH3D) measurements. Front Mar Sci. 2023;10:969567. doi:10.3389/fmars.2023.969567
