The National Transportation Safety Board (NTSB) has reviewed the Pipeline and Hazardous Materials Safety Administration’s (PHMSA) notice of proposed rulemaking (NPRM), Hazardous Materials: Oil Spill Response Plans and Information Sharing for High-Hazard Flammable Trains, published on July 29, 2016. This NPRM addresses issues raised in an August 1, 2014, advance notice of proposed rulemaking (ANPRM) regarding comprehensive oil spill response planning thresholds for high-hazard flammable trains (HHFT) and plan review by local and federal agencies. This NPRM also addresses local agency emergency planning and information sharing and guidance for offerors on classifying crude oil for hazardous materials transportation.
Proposal Topic C—Initial Boiling Point Test
PHMSA proposes to incorporate by reference in the HMR the Standard Test Method for Determination of Light Hydrocarbons in Stabilized Crude Oils by Gas Chromatography, ASTM D7900, for determining the boiling range distribution of hydrocarbons in crude oil for testing Class 3 packing group assignments. Method D7900 is identified in ANSI/API recommended practice 3000 (API RP 3000), Classifying and Loading of Crude Oil into Rail Tank Cars, as an industry best practice. PHMSA proposes not to replace the currently authorized initial boiling point testing methods but to add method D7900 as an alternative.
While not stated in the NPRM, this topic would partly address NTSB Safety Recommendation R-14-006, which was issued as a result of the July 6, 2013, Lac-Mégantic, Quebec, train derailment.
The NTSB safety recommendation reads as follows:
Require shippers to sufficiently test and document the physical and chemical characteristics of hazardous materials to ensure the proper classification, packaging, and record-keeping of products offered in transportation.
In the NTSB’s July 12, 2016, response letter to PHMSA’s action on recommendation R 14 006, we expressed concern that the HMR provides numerous insufficient test methods for shippers to choose from when testing flammable hazardous materials to determine the packing group assignment. Existing section 173.121(a)(2) lists five test methods that are permitted to be used to determine the initial boiling point of Class 3 materials:
• Standard Test Method for Distillation of Petroleum Products at Atmospheric Pressure (ASTM D86)
• Standard Test Method for Distillation Range of Volatile Organic Liquids (ASTM D1078)
• Petroleum products—Determination of distillation characteristics at atmospheric pressure (ISO 3405)
• Petroleum products—Determination of boiling range distribution—Gas chromatography method (ISO 3924)
• Volatile organic liquids—Determination of boiling range of organic solvents used as raw materials (ISO 4626)
Sandia National Laboratories (Sandia) recently reported that due to significant variability in criteria and procedures utilized in selection, acquisition, and analysis of crude oil samples, the available data are of insufficient quality to enable a meaningful comparison of crude oils—either to each other or against a designated standard.
For example, the report noted that the initial boiling point of crude oil is usually tested by Standard Method for Distillation of Petroleum Products at Atmospheric Pressure, ASTM D86, which Sandia criticized because it is not the most accurate or most appropriate for determining packing group. The report suggested that method D86 yields initial boiling point results that characterize substances as less hazardous than the results of other methods such as ASTM D7900. The NTSB further notes that D86 Scope section 1.2 states, “The test method is designed for the analysis of distillate fuels; it is not applicable to products containing appreciable quantities of residual material.” Crude oil is not a distillate fuel. Additionally, test methods ASTM D1078, ISO 3405, ISO 3924, and ISO 4626 are for either finished fuels or specific organic liquids, and they are not appropriate for crude oil and could result in the misclassification of crude oil shipments.
The NPRM proposal adds ASTM D7900 to the test methods offerors may use to determine the initial boiling point of flammable materials. However, adding a new test method without eliminating one or more inappropriate and possibly less accurate method results only in additional data variability among offerors who choose ASTM D7900 and those who select other approved methods.
We are therefore concerned that failure to mandate specific and the most appropriate sampling and testing methods may result in misclassification and data variability that limits its usefulness for research. We urge PHMSA to identify the most appropriate flammable material classification method and require uniform sampling and testing techniques to provide a high level of confidence that materials are being properly classified, packaged, and shipped. The NPRM discussion suggests that PHMSA has already established the bases for exclusively using method D7900 for initial boiling point determination for crude oil:
For the initial boiling point test, the API RP 3000 concluded that for crude oils containing volatile, low molecular weight components (e.g. methane), the recommended best practice is to test using ASTM D7900. This test ensures a minimal loss of light ends because it determines the boiling range distribution from methane through n nonane with an [initial boiling point] defined as the temperature at which 0.5 weight percent loss is observed when determining the boiling range distribution defined in ASTM D7169. This test differs from the boiling point test options currently in the HMR, which do not remove and recover the light ends.
We urge PHMSA to consider the guidelines in API RP 3000 for initial boiling point determination. Currently, the API recommended practice indicates the use of ASTM D7900 and D7169 for determining the initial boiling point in crude oil liquids. However, the Sandia report also notes that although ASTM D7900 and ASTM D7169 will yield a full boiling point distribution and improve on method D86, it still may not yield the most representative initial boiling point value for crude oil because these chromatographic methods are subject to the complexities of the crude oil mixture. Sandia continues to study this issue and will issue findings on more appropriate methods for crude oil testing. API RP 3000 is an industry standard that is widely recognized and readily updated to include future research findings, such as those issued by Sandia.
The Sandia report also pointed out that samples are acquired from a wide range of supply chain points using a variety of open and closed container sampling techniques and characterized using a variety of analysis methods. To minimize initial boiling point measurement variability, API RP 300 First Edition section 184.108.40.206.1 recommends the sampling of crude oil in accordance with ASTM D4057 to minimize the loss of low molecular weight components from crude oil samples. However, since the publication of API RP 300 First Edition, Standard Practice for Manual Piston Cylinder Sampling for Volatile Crude Oils, Condensates, and Liquid Petroleum Products (ASTM D8009,) has been developed for the sampling of crude oil for initial boiling point testing. We urge PHMSA to consider the effects of sampling methods on the variability of crude oil test results, such as initial boiling point, and specify a sampling method that reduces variability.
We urge PHMSA to identify the most appropriate flammable material classification method and require uniform sampling and testing techniques to provide a high level of confidence that materials are properly classified, packaged, and shipped.
The NTSB offers no additional comments to the NPRM. We appreciate the opportunity to comment on the PHMSA NPRM.