NTSB Issues Investigations Report on BNSF’s April 26 Hazmat Derailment
What happened?
On April 26, 2024, about 12:37 p.m. local time, westbound BNSF mixed-freight train H-BELPHX1-25 derailed 35 railcars at milepost 178.6 on the BNSF Gallup Subdivision near Manuelito, N.M., according to the report that was issued June 18 (download below). The derailed equipment included six Department of Transportation (DOT) Specification 112A340W (DOT-112) tank cars loaded with UN 1075, liquefied petroleum gas (LPG), a Division 2.1 flammable gas. Four of these tank cars were breached (punctured or torn) during or after the derailment and released lading, which ignited. A fifth LPG tank car was exposed to post derailment fires and released vapor through its pressure relief device (PRD) but was not otherwise breached. About 180,000 gallons of LPG were released from breaches, PRD activity, and flaring operations. (See Figure 1, above)
According to the report, the train was not carrying other hazardous materials. Because tank cars carrying LPG can explode when exposed to fire, the local fire department ordered an evacuation within a two-mile radius of the derailment; 52 people were evacuated; and Interstate 40 was closed in both directions from milepost 10 to milepost 126. Sections of Interstate 40 remained closed for about 48 hours. The local fire department allowed the LPG tank cars to burn overnight, and the fires were extinguished on April 27. The evacuation and road closure were lifted on April 28. No injuries were reported. The weather at the time of the derailment was 78°F with no precipitation. NTSB launched an investigation into the thermal protection performance of the tank cars involved in the release.
According to the report, all six DOT-112 tank cars were equipped with ceramic fiber thermal protection blankets and PRDs. Based on the shipper’s records, each tank car had at least 11% outage (space not filled by liquid lading) as calculated at 110°F. Federal regulations at Title 49 Code of Federal Regulations (CFR) Part 173.31(b) require a minimum of 1% outage at this temperature for LPG. The investigation found no evidence of pre-derailment tank car damage or defects.
Post accident on-scene examinations found damage consistent with mechanical breaches (punctures or tears created by outside forces) on two tank cars: UTLX952565 and UTLX955642. For these two tank cars, the investigation found evidence of lading loss only through mechanical breaches. The released LPG ignited and contributed to post derailment fires near other derailed tank cars, including other LPG tank cars.
Two other derailed LPG tank cars (UTLX 954193 and NATX 40068) were not mechanically breached during the derailment but released lading from shell tears after being exposed to fire, according to the report. The duration of fire exposure before these releases was not recorded. The Federal Railroad Administration (FRA) recovered the panels of tank car shell material containing tears for examination in coordination with the NTSB. Ultrasonic measurements showed that the shell wall had thinned and cracked near the tears. (See Figure 2, below)
According to the report, the NTSB also recovered PRDs from all six derailed DOT-112 tank cars for examination and testing to determine whether they functioned as designed. “Five PRDs were sufficiently intact to test safely; the sixth had sustained thermal damage severe enough that it could not be disassembled or tested. The severely damaged PRD came from a mechanically breached tank car.” Test results are summarized in the table below.
Analysis
Two derailed DOT-112 tank cars were not breached mechanically during the derailment but released material from thermal tears after being exposed to fire, according to the report. The NTSB had not previously observed thermal tears in DOT-112 tank cars required to have thermal protection systems. “There were no indications that tank cars had been improperly loaded or damaged before the derailment, meaning that the design and fabrication of the tank cars were the main factors of interest in understanding the hazardous materials release,” NTSB said. The investigation therefore focused on whether the tank cars’ thermal protection systems and PRDs performed as designed after the derailment. Under federal regulations in 49 CFR Part 179, a thermal protection system is required to prevent lading release from a tank car, except through PRD activity, during 100 minutes of exposure to a pool fire or 30 minutes of exposure to a torch fire. “There was not enough evidence to determine whether the thermally breached tank cars involved in the Manuelito release met this performance standard, so the FRA, in coordination with the NTSB, examined the tank car shell panels containing suspected thermal tears for material or manufacturing defects that could have increased their susceptibility to thermal damage,” according to the report.
According to the report, the laboratory examinations found that “the tears resulted from the shells thinning and cracking.” This, NTSB says, “confirmed the presence of thermal tears, which occur when the pressure within a tank car exceeds the tensile strength of its shell.” As the temperature of a tank car increases, the volatility (and pressure) of the lading tends to increase as the tensile strength of the steel shell and head decreases until the steel tears. “The post-accident examinations and laboratory testing did not indicate that the tank car shells failed as a result of a flaw in design or fabrication. The NTSB’s tests and examinations of the PRDs found no indications that they failed to perform as designed; all testable PRDs discharged near or below the design pressure during tests, and the significantly reduced start-to-discharge pressure of three PRDs was typical of valve springs that have been compressed and subjected to heat during an emergency,” according to the report.
Although two tanks sustained thermal tears, emergency responders did not observe explosions consistent with a boiling liquid expanding vapor explosion (BLEVE). BLEVEs result from fire exposure that raises the temperature and therefore pressure within a tank car until a thermal tear ruptures the shell. Along with the results of the PRD examinations, the absence of a BLEVE “suggests that the thermal tears occurred after the tank cars had released most of their lading through PRD activity before the tank shells failed, in accordance with the DOT-112 design,” according to the report. “The thermal breaches that did occur during this accident resulted from fire exposure, and these fires were fueled in part by mechanically breached LPG tank cars.” This failure mode, in which tank car failures and hazardous material releases cascade beyond the initial mechanical breaches, NTSB says, is a well-documented phenomenon. For trains carrying large amounts of flammable liquids—high-hazard flammable trains (HHFTs)— federal regulations are intended to reduce the probability of derailments and the severity of consequences, including cascading releases. Required precautions include speed and routing limits, information sharing for local emergency responders, use of survivable tank cars, and two-way braking systems.
According to the report, freight train H-BELPHX1-25 did not meet the definition of an HHFT because it contained only six tank cars of hazardous material, and the material was a Division 2.1 flammable gas, not a Class 3 flammable liquid. “The accident still led to a cascading release that breached mechanically intact tank cars, an outcome HHFT regulations are intended to prevent,” according to the report. The NTSB documented similar safety issues following the February 3, 2023, derailment of a non-HHFT mixed freight train in East Palestine, Ohio. In response to the East Palestine derailment, the NTSB recommended that the Pipeline and Hazardous Materials Safety Administration (PHMSA) revise the definition of HHFT to include more trains vulnerable to cascading hazardous materials releases:
“Revise the definition of high-hazard flammable train to account for differences in survivability between tank car specifications and to include hazardous materials other than flammable liquids, such as combustible liquids and Division 2.1 flammable gases, that can contribute to cascading hazardous materials releases; if necessary, obtain legislative authority to act on this recommendation.”
The Manuelito derailment, NTSB says, provides another example of a non-HHFT exhibiting HHFT-like tank car failures following derailment and underscores the importance of the recommendation.
The emergency response to the derailment and fire, including road closures, evacuation of a two-mile area, and the decision to allow the tank cars to burn overnight rather than attempting a vent and burn operation that would have placed personnel near the tank cars, was effective in protecting the public, the NTSB noted.
Lessons Learned
This accident, NTSB says, illustrated that cascading hazardous materials releases occur in trains other than HHFTs. The NTSB has previously recommended that PHMSA expand the definition of HHFT as descried in the safety recommendation above.
The NTSB says it focused its investigation on the performance of the tank cars rather than the emergency response, “which was timely and effective in protecting the public from the danger posed by the derailed hazardous materials tank cars, according to the report. “However, the response included significant road closures, and the after-action report created by the McKinley County Office of Emergency Management noted the risks stranded motorists may face during widespread and prolonged road closures. The office is working with state partners to update its road closure plan.
“The McKinley County Office of Emergency Management’s decision to update its road closure plan even after a successful evacuation is a valuable reminder to emergency response agencies that emergency preparedness is an ongoing task,” NTSB said.
