API Publication 935-1999: Evaluating Petroleum Refinery and Petrochemical Plant Fenceline Monitoring Data

Technical guidance for assessing air quality data at facility boundaries – a comprehensive review

Scope and Objectives of API Publ 935-1999

API Publication 935-1999, formally titled Guide for the Evaluation of Petroleum Refinery and Petrochemical Plant Fenceline Monitoring Data, is a key technical guidance document issued by the American Petroleum Institute (API). It addresses the systematic and scientifically rigorous evaluation of ambient air monitoring data collected at the fenceline of petroleum refineries and petrochemical facilities. The purpose of this guide is to provide facility operators, environmental managers, and regulatory agencies with a consistent framework for interpreting fenceline data to assess potential off-site impacts, to demonstrate compliance with ambient air quality standards, and to support community right-to-know efforts.

The scope of API Publ 935 includes both continuous and intermittent monitoring data, covering a range of volatile organic compounds (VOCs) and other air toxics of concern. While it is not a mandatory standard, it has been widely adopted as a reference for best practices in fenceline monitoring and is often cited in state and federal regulatory contexts.

Key Takeaway: API Publ 935-1999 provides a defensible methodology for evaluating fenceline air monitoring data, enabling facilities to identify potential emission issues, communicate findings to stakeholders, and improve environmental performance.

Technical Methodology and Data Quality Requirements

API Publ 935-1999 recommends a structured approach to data collection, validation, and statistical analysis. The following subsections outline the core technical elements.

Data Quality Objectives (DQOs)

The guide stresses the importance of establishing clear DQOs before monitoring begins. These objectives define the level of uncertainty acceptable for decision-making. Key parameters include precision, accuracy, detection limits, completeness, and representativeness.

Parameter Recommended Target Notes
Precision (relative standard deviation) ≤ 25% Based on duplicate or replicate samples
Accuracy (recovery) 80% – 120% For spiked or audit samples
Method Detection Limit (MDL) ≤ 10% of relevant standard Should be low enough to detect concentrations of health significance
Completeness ≥ 85% of scheduled samples Excluding planned downtime and quality control failures

Sampling and Analytical Methods

The guidance covers both grab sampling (e.g., Summa canisters) and continuous monitoring (e.g., gas chromatography with flame ionization detection, real-time analyzers). It emphasizes the need for thorough quality assurance, including field blanks, trip blanks, and calibration verification. Analytical methods should follow EPA-approved protocols (e.g., EPA Method TO-14A, TO-15, or equivalent).

Data Validation and Treatment

API Publ 935 prescribes procedures for validating raw data, including assessment of holding times, contamination checks, and consistency tests. It specifically addresses non-detect and censored data handling, recommending for example the substitution of one-half the detection limit for simple summary statistics, but cautioning against this for hypothesis tests unless properly justified.

Best Practice: When analyzing fenceline data, always document the rationale for handling non-detects. Consider using robust statistics (e.g., Kaplan-Meier) for datasets with high censoring rates to avoid bias.

Statistical Evaluation

Statistical analysis is central to the API 935 approach. The guide recommends calculation of arithmetic means, percentiles (e.g., 99th percentile), and maximum values for comparison with ambient air standards or health-based benchmarks. Trend analysis using techniques such as linear regression or Mann-Kendall test is advocated to detect changes over time and evaluate the effectiveness of emission controls.

Implementation Strategies and Practical Considerations

Successful implementation of API Publ 935 involves careful planning, stakeholder engagement, and continuous improvement.

Sampling Location and Frequency

Monitoring stations should be sited at the fenceline in directions representative of downwind conditions. The guide recommends continuous or high-frequency sampling during process operations, with additional stations during upwind events to capture background contributions. Seasonal and meteorological variability must be accounted for.

Data Interpretation and Communication

API Publ 935 emphasizes transparent communication of results. Facilities should prepare periodic reports that include clear comparisons to applicable standards, explanation of episodic exceedances, and description of corrective actions. Graphics such as time-series plots, wind roses with concentration overlays, and cumulative probability plots are highly recommended.

Common Pitfall: Interpreting a single exceedance as a violation without considering natural background, meteorological conditions, or known upwind sources can lead to false conclusions. Always conduct a causality analysis before reporting.

Quality System Integration

To maintain credibility, the fenceline monitoring program should be integrated into the facility’s existing quality management system (e.g., ISO 14001). Regular performance audits, inter-laboratory comparisons, and data reviews should be scheduled.

Compliance Considerations and Regulatory Context

While API Publ 935 is a publication rather than a regulatory standard, it holds significant weight in the regulatory environment. The U.S. Environmental Protection Agency (EPA) and state agencies have referenced the guide in consent decrees, Section 114 requests, and Title V permit conditions. Facilities required to perform fenceline monitoring by regulation often adopt the principles of API 935 to ensure defensibility.

Compliance notes include:

  • Recordkeeping: Maintain complete raw data, validation logs, and statistical evaluations for at least the permit-required period (often five years).
  • Reporting: Many regulatory frameworks require semi-annual or annual reports containing summary statistics, exceedance analysis, and corrective actions as described in API 935.
  • Audits: Third-party audits of the monitoring program typically check alignment with API 935 guidance, especially data quality objectives and statistical methods.

Attention: Although API Publ 935 is not mandatory, failure to follow its methodologies when fenceline monitoring is required by permit may lead to regulatory scrutiny; deviation must be justified with equal rigor and documented.

Frequently Asked Questions

Q: What is the main purpose of API Publication 935?
A: API Publ 935 provides a systematic framework for evaluating air quality monitoring data at the fenceline of refineries and petrochemical plants. It guides users in data quality assessment, statistical analysis, and interpretation to demonstrate environmental performance and compliance.
Q: Is API Publ 935-1999 a mandatory standard?
A: No, it is a publication (guidance document), not a consensus standard. However, it is frequently referenced in regulatory permits and enforcement actions as a recommended practice, and adherence is often required by consent decrees.
Q: How does API Publ 935 relate to current EPA fenceline monitoring requirements?
A: EPA regulations such as the Refinery Sector Rule (40 CFR Part 63) have established fenceline monitoring requirements for certain pollutants. EPA guidance often cites API Publ 935 as a source of acceptable data evaluation practices, and many facilities use it to meet these requirements.
Q: Can API Publ 935 be applied to other industrial sectors?
A: While developed for refineries and petrochemical plants, the underlying data evaluation principles (DQOs, validation, statistics) are transferable to other industries performing fenceline or perimeter monitoring. The guide can be adapted with appropriate modifications.


© 2026. This technical article is provided for informational purposes and does not constitute official guidance. Refer to the latest edition of API Publ 935 (1999) for the complete text.

📥 Standard Documents Download

🔒
Please wait 10 seconds, the download links will appear after the ad loads

Leave a Reply

Your email address will not be published. Required fields are marked *