Source EP2. National Transportation Safety Board, Railroad Investigation Report RIR-24-05, dated June 25, 2024; corrected reissue July 23, 2024. The reissue note states corrections concern the description of the TILX402025 lease on report pages 57–58. Excerpts from the corrected PDF. PDF page numbers and report section numbers are supplied for locating the original. THIS IS LATER EVIDENCE, NOT PART OF THE FEBRUARY 23, 2023 BLIND PACKET.

[PDF page 1, report front matter]
NOTE: This report was reissued on July 23, 2024, with corrections to the description of the tank car TILX402025 lease on pages 57–58.

[PDF pages 21–23, sections 1.1.1–1.1.2]
At 8:13 p.m., train 32N passed over the HBD in Salem, Ohio, (MP 69.01) near the surveillance cameras that recorded fire on the hopper car. The HBD recorded a temperature 103°F above ambient on the L1 bearing of the hopper car—a temperature exceeding NS’s 90°F above-ambient threshold for a non-critical alert. The R1 bearing was 20°F above ambient. The HBD transmitted a non-critical alert to the advanced train control (ATC) desk about the temperature of the L1 bearing but, as designed, did not broadcast an alarm to the train’s crew.
About 8:52 p.m. on February 3, 2023, train 32N began to traverse an HBD at MP 49.81 in East Palestine, Ohio. Locomotive event recorder data showed the train was traveling about 43 mph. The maximum authorized speed in the area was 50 mph. Less than a minute later, this HBD recorded a temperature 253°F above ambient on the L1 bearing of the hopper car; the R1 bearing was still 20°F above ambient.
A reading of 253°F above ambient is the maximum reading for the type of HBD installed at MP 49.81; the actual temperature may have been higher.

[PDF page 39, section 1.1.3, Table 3. The columns other than date, time, and OCPX80370 are omitted; rows transcribed exactly, including the original blank date cells. °F here is a tank-car shell measurement, not a bearing reading.]
Date | Time | OCPX80370 Temperature (°F)
2/5/2023 | 4:00 p.m. | 135
 | 5:00 p.m. | 138
 | 6:00 p.m. | 136
 | 7:00 p.m. | 136
 | 8:00 p.m. | 130
 | 9:00 p.m. | 130
 | 10:00 p.m. | 130
 | 11:00 p.m. | 131
2/6/2023 | 12:00 a.m. | 139*
 | 1:00 a.m. | 129
 | 2:00 a.m. | -
 | 3:00 a.m. | -
 | 4:00 a.m. | -
 | 7:00 a.m. | 120
 | 8:37 a.m. | 127
 | 9:30 a.m. | 126
 | 10:30 a.m. | 126
 | 11:30 a.m. | 126
 | 12:30 p.m. | 126
 | 1:30 p.m. | 126
 | 2:30 p.m. | 126
*When interviewed by the NTSB, the EPFD fire chief noted that this increase in temperature coincided with the ignition of a small fire underneath tank car OCPX80370, and that the temperature decreased after the fire was extinguished.

[PDF page 43, section 1.1.3]
By the time of the meetings on February 6, the highest tank car temperatures had fallen from about 138°F the previous afternoon to about 126°F.
According to the incident commander, no Oxy Vinyls employees were invited to these meetings, and none were present. Nor did any Oxy Vinyls employees communicate with the incident commander outside these meetings.

[PDF page 44, section 1.1.3]
The vent and burn was originally scheduled for 1:00 p.m. on February 6, 2 hours ahead of the 3:00 p.m. deadline described to the incident commander by SPSI and SRS. However, the vent and burn was repeatedly delayed for technical and safety reasons: once to rewire the charges and three times because of unauthorized persons entering the evacuation zone. ESI detonated the vent and burn charges at 4:37 p.m. for all five VCM tank cars, puncturing each tank car at two points: the highest point in the vapor space and the lowest point where the liquid had pooled.

[PDF pages 157–158, section 2.5.1]
Oxy Vinyls, which was providing support to NS and its contractors but was not in direct communication with the incident commander, concluded based on the same evidence and its own knowledge of VCM that dangerous polymerization was not occurring. Oxy Vinyls shared its conclusions with NS and NS’s contractors while on the scene; these conclusions were not communicated to the incident commander.
The NTSB obtained expert opinions on VCM polymerization from Oxy Vinyls’ LaPorte plant technical manager and a chemistry professor at Indiana University. Based on their interviews, hearing testimony, and references to scientific literature, stabilized VCM—in the case of East Palestine, VCM packaged in tank cars in a low-oxygen environment—cannot polymerize when exposed to heat without the presence of an initiator (free radicals), and it cannot sustain a polymerization reaction without the continued introduction of free radicals, which are consumed during the polymerization process.

[PDF pages 162–164, section 2.5.1]
In discussions with the incident commander and the NTSB on scene, NS noted a 3°F increase in temperature for tank car OCPX80370 as an indication that polymerization was occurring. This increase to 138°F on the afternoon of February 5 was close to the highest temperature recorded on a tank car shell (139°F, measured during a brief fire under the tank car about midnight on February 5-6).
According to information Oxy Vinyls provided the NTSB after the vent and burn, a runaway polymerization reaction would produce “a rapid, significant, and sustained increase in temperature and pressure.” The dashed red trendline in the figure shows a gradual decline instead.
The NTSB concludes that the observed downward temperature trend in tank car OCPX80370 indicates that polymerization was not occurring within the tank car, contrary to the representation by NS and its contractors.
Evidence available after the vent and burn also does not support polymerization. During the vent and burn, all five VCM tank cars expelled material, indicating that the cars were still under positive internal pressure. Oxygen and other contaminants therefore could not enter the tank cars before the vent and burn, and the VCM remained in a low-oxygen, stabilized environment not conducive to polymerization. As discussed above, PRDs recovered from all five tank cars did not contain obstructions of PVC, and examinations did not identify foreign matter obstructing their operation. Finally, Oxy Vinyls performed laboratory analyses on residue samples recovered from the interiors of the VCM tank cars. While these analyses did not definitively rule out the presence of PVC (VCM and PVC are chemically similar) they did not find evidence of PVC. The NTSB concludes that the VCM within the derailed DOT-105 tank cars remained in a stabilized environment until the vent and burn and did not undergo polymerization; the vent and burn procedure was not necessary to prevent a polymerization-induced tank rupture.

[PDF pages 184–186, sections 3.1–3.2]
26. The observed downward temperature trend in tank car OCPX80370 indicates that polymerization was not occurring within the tank car, contrary to the representation by Norfolk Southern Railway and its contractors.
27. The vinyl chloride monomer within the derailed DOT-105 tank cars remained in a stabilized environment until the vent and burn and did not undergo polymerization; the vent and burn procedure was not necessary to prevent a polymerization-induced tank rupture.
The National Transportation Safety Board determines that the probable cause of the derailment involving Norfolk Southern Railway train 32N was the failure of the L1 bearing on the 23rd railcar in the consist that overheated and caused the axle to separate, derailing the train and leading to a postderailment fire that likely began with the release of a Class 3 flammable liquid from a DOT-111 tank car that was punctured during the derailment.
