Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
API Publication 312 (1990): Methods for Analysis of Natural Gas by Gas Chromatography — Scope, Technical Requirements, and Implementation
Understanding the 1990 American Petroleum Institute recommended practice for compositional analysis of natural gas using gas chromatography
1. Introduction and Scope of API Publ 312-1990
API Publication 312 (1990), formally titled Analysis of Natural Gas by Gas Chromatography, was developed by the American Petroleum Institute’s Committee on Gas Measurement to provide a standardized analytical framework for determining the chemical composition of natural gas. This publication is intended for laboratories engaged in custody transfer, quality certification, and process monitoring of natural gas streams. It addresses both the determination of hydrocarbon components from methane (C1) through pentanes-plus (C5+) and the measurement of fixed gases such as nitrogen, carbon dioxide, and oxygen.
The 1990 edition consolidates prevailing best practices from the natural gas industry and aligns with earlier methods such as GPA 2261 and ASTM D1945. While API Publ 312-1990 is a voluntary publication, it has been widely adopted by gas processors, pipeline operators, and regulatory bodies as a benchmark for accuracy and reproducibility in gas analysis. The document covers gas chromatographic (GC) procedures for samples obtained from pipelines, production wells, and processing plants, assuming a sample pressure of at least 200 kPa and a temperature range of 0–50 °C.
Key exclusions include the analysis of liquefied natural gas (LNG), gas hydrates, and trace sulfur species, which are covered by separate API standards (e.g., API Publ 317 for sulfur compounds). The scope also notes that the method is applicable only to samples that are homogeneous, single-phase gases at the time of analysis.
2. Analytical Requirements and Methodology
2.1 Instrument Configuration
API Publ 312-1990 specifies the use of a gas chromatograph equipped with both a flame ionization detector (FID) for hydrocarbon components and a thermal conductivity detector (TCD) for fixed gases. A single-instrument configuration with multiple columns and column-switching valves is recommended to separate all target analytes within a single run. The publication mandates that the GC oven be capable of temperature programming from ambient to at least 200 °C, with a temperature stability of ±0.1 °C.
Columns are typically packed or capillary; the standard provides specific guidance on stationary phases (e.g., 100% dimethylpolysiloxane for FID, HayeSep or Molecular Sieve for TCD). Carrier gas must be helium with a purity of at least 99.995% (grade 5.0) to avoid interference. The sample introduction system must include a heated injection port set at 150 °C and a gas sampling valve with a calibrated sample loop (volume 0.5–2.0 mL).
2.2 Calibration and Quantification
The calibration procedure in API Publ 312-1990 relies on two primary strategies: external standard for fixed gases and relative molar response (RMR) for hydrocarbons. A multi-component calibration gas mixture certified to within ±0.5 mol% for each component is required. Table 1 summarizes the typical analytical ranges and detector assignments.
| Component | Typical Concentration Range (mol%) | Detector |
|---|---|---|
| Methane (CH₄) | 70–99 | FID |
| Ethane (C₂H₆) | 1–15 | FID |
| Propane (C₃H₈) | 0.1–5 | FID |
| Butanes (C₄H₁₀ isomers) | 0.05–2 | FID |
| Pentanes-Plus (C₅+) | 0.01–0.5 | FID |
| Nitrogen (N₂) | 0.1–5 | TCD |
| Carbon Dioxide (CO₂) | 0.1–3 | TCD |
| Oxygen (O₂) | 0.01–0.5 | TCD |
Data acquisition and processing must be performed by an integrator or chromatography data system that can correct for baseline drift and resolve peaks with a resolution greater than 1.5. The publication recommends using argon as an internal standard for TCD quantification when the sample is known to contain minimal argon (<0.1 mol%).
Tip: Always verify the performance of your GC system with a check standard after every ten sample injections. API Publ 312-1990 recommends that the measured response for each component in the check standard should not deviate more than ±2% from the original certified values.
3. Implementation and Quality Assurance
3.1 Sample Handling and Preparation
Proper sample conditioning is critical for achieving results that conform to API Publ 312-1990. Samples must be collected in stainless steel cylinders (e.g., DOT 3AA or 3E) that have been cleaned and passivated. The standard prescribes a minimum cylinder flush volume of ten times the cylinder’s internal volume to eliminate upstream gas residuals. The sample pressure should be at least 50 % above the dew point to prevent condensation of heavier hydrocarbons. For samples containing more than 0.1 mol% of C5+ components, the entire sample system must be heated to 60 °C to avoid flash condensation.
Warning: Failure to adequately heat sample cylinders when heavy hydrocarbons are present can lead to significant errors in the measured C5+ concentration, often underestimating the actual heating value by 1–3 %. If condensation is suspected, repeat the analysis with a heated transfer line.
3.2 Method Validation and Uncertainty
Laboratories adopting API Publ 312-1990 should perform a validation study covering at least three concentration levels (low, medium, high) for each component. The standard requires that the repeatability (Horwitz ratio) not exceed 30 % for major components (>10 mol%) and 40 % for minor components (<1 mol%). Reproducibility between laboratories should be evaluated through interlaboratory studies; the publication provides reference data from a 1988 round-robin test involving 15 U.S. laboratories.
Table 2 (not shown in full here) in API Publ 312-1990 lists the expected repeatability limits for each component. For methane, the absolute difference between two duplicate results should not exceed 0.5 mol%; for nitrogen, it should be ≤0.1 mol%. The overall combined uncertainty of the mole fraction for methane is estimated at ±1 mol% (95 % confidence level) under ideal conditions.
Best Practice: Maintain rigorous documentation of all calibration events, check standard results, and operator training records. API Publ 312-1990 can be seamlessly integrated into a quality management system such as ISO/IEC 17025. Many accreditation bodies accept it as a valid reference method for natural gas composition.
Critical: The use of non-compliant carrier gas (e.g., helium with purity < 99.99 % or carrier containing trace oxygen) will cause oxidation of the TCD filaments and degrade column efficiency. This can lead to irreversible damage and invalidate all results generated on that system until the issue is resolved.
4. Compliance, Auditing, and Relation to Other Standards
Compliance Demonstrations – Although API Publ 312-1990 is not a regulatory standard, natural gas trading contracts often incorporate it by reference. Laboratories must be able to produce evidence of conformance during audits. This includes (a) the availability of the current 1990 edition in the laboratory, (b) valid calibration certificates for reference gas mixtures traceable to NIST (or equivalent), (c) completed check standard logs, and (d) records of annual system suitability tests.
Interaction with Other Standards – API Publ 312-1990 overlaps significantly with ASTM D1945-14 (Standard Test Method for Analysis of Natural Gas by Gas Chromatography) and GPA 2261-20. However, API’s publication provides more detailed guidance on sample conditioning and the handling of pentanes-plus fractions. Users should check local regulatory requirements; for example, the U.S. Environmental Protection Agency’s Greenhouse Gas Reporting Program (40 CFR Part 98) accepts results generated according to API Publ 312-1990 provided all quality-assurance steps are documented.
Revisions and Supersession – The 1990 edition remains current as of the date of this article, though an updated draft was circulated in 2023 under the project number API 312-2. Users are encouraged to monitor the API Publications Catalog for future editions. When transitioning to a newer version, a gap analysis comparing the new requirements with existing procedures should be performed, and any method modifications must be validated before use.
Frequently Asked Questions
Q: Is API Publ 312-1990 still active or has it been withdrawn?
A: As of 2026, the 1990 edition has not been formally withdrawn by the American Petroleum Institute, though a revision (API 312-2) is under development. The 1990 edition is still widely referenced in custody transfer agreements and is accepted by most regulatory bodies in North America. Always confirm the latest edition with your contract or regulatory requirements.
A: As of 2026, the 1990 edition has not been formally withdrawn by the American Petroleum Institute, though a revision (API 312-2) is under development. The 1990 edition is still widely referenced in custody transfer agreements and is accepted by most regulatory bodies in North America. Always confirm the latest edition with your contract or regulatory requirements.
Q: Can API Publ 312-1990 be used for liquefied natural gas (LNG) analysis?
A: No. The scope of API Publ 312-1990 explicitly excludes LNG and two-phase samples. For LNG composition, refer to API Publ 317 or GPA 2265. Analysis of LNG requires specialized equipment (e.g., vaporization systems and cryogenic injection) that is not covered by this publication.
A: No. The scope of API Publ 312-1990 explicitly excludes LNG and two-phase samples. For LNG composition, refer to API Publ 317 or GPA 2265. Analysis of LNG requires specialized equipment (e.g., vaporization systems and cryogenic injection) that is not covered by this publication.
Q: What is the difference between API Publ 312-1990 and ASTM D1945?
A: Both methods use gas chromatography to measure the same components, but API Publ 312-1990 provides more detailed instructions for sample conditioning, especially for heavy hydrocarbons (C5+). It also recommends specific column configurations and validation criteria. ASTM D1945 is generally considered a generic method, while API’s publication is tailored to pipeline-quality gas with higher repeatability requirements.
A: Both methods use gas chromatography to measure the same components, but API Publ 312-1990 provides more detailed instructions for sample conditioning, especially for heavy hydrocarbons (C5+). It also recommends specific column configurations and validation criteria. ASTM D1945 is generally considered a generic method, while API’s publication is tailored to pipeline-quality gas with higher repeatability requirements.
Q: How often should the calibration be verified according to this standard?
A: API Publ 312-1990 recommends that a check standard be analyzed at the beginning of each analytical sequence and after every ten sample injections, or every four hours, whichever occurs first. The calibration itself should be updated whenever the check standard results exceed the tolerance of ±2 % for any component or when significant changes are made to the GC system (e.g., column replacement, detector repair).
A: API Publ 312-1990 recommends that a check standard be analyzed at the beginning of each analytical sequence and after every ten sample injections, or every four hours, whichever occurs first. The calibration itself should be updated whenever the check standard results exceed the tolerance of ±2 % for any component or when significant changes are made to the GC system (e.g., column replacement, detector repair).
Published: 2026. This article is based on API Publication 312, first edition, 1990. For official requirements, refer to the complete API Publ 312-1990 document available from the American Petroleum Institute.