We have discussed the hazards and risks associated with the transferring of hazardous materials in bulk. This case was a terminal to ship transfer, but these basic failures can happen on any transfer. Not following procedures, not utilizing an unloading checklist, and not using the unloading process and the nitrogen purge process as it was DESIGNED all led to this OVER PRESSURE EVENT involving a Category 2 flammable liquid with damages near $1,000,000.
On May 30, 2019, about 0750, a chemical tanker ship experienced an overpressurization of the number 3 port and starboard cargo tanks while discharging liquid hexene at Vopak Terminal in Deer Park, Texas. The overpressurization resulted in damage to the number 3 port cargo tank and the tank top (deck). All cargo was contained onboard the double-hulled vessel, with no pollution or injuries reported.
Damage to the ship was estimated at $750,000, and the contaminated cargo was an estimated $100,000 loss.
The National Transportation Safety Board (NTSB) determines that the probable cause of the overpressurization and rupture of the 3P cargo tank aboard the Fairchem Filly during offloading was the VESSEL AND TERMINAL PERSONNEL INVOLVED NOT FOLLOWING POLICIES AND PROCEDURES RELATED TO CARGO DISCHARGE AND NITROGEN-BLANKETING OPERATIONS. Contributing to the casualty was the lack of effective communication between the vessel and terminal personnel and the decision of the vessel’s PIC to continue discharge operations after being unable to communicate with the terminal.
Hexene tends to absorb oxygen, which negatively affects its purity, and thus needs to be stored and transferred under a “blanket” or “pad” of inert gas. The ship utilized nitrogen (N2) to maintain a blanket of 2 pounds per square inch (psi) over the product while stored onboard the vessel and during the transfer of the hexene to the terminal. During typical transfer operations, the remaining tank volume above the cargo increases as the liquid hexene is pumped off the vessel, requiring additional nitrogen to fill the space. Vopak supplied the ship with nitrogen for the transfer. Vopak received its nitrogen from a shoreside distribution piping network, utilizing a pressure-reducing station to reduce the incoming nitrogen from 170 psi to 90 psi, which would then be distributed to the docks and used by vessels for purging and blanketing tanks. The connection between a ship and the terminal was normally accomplished via either a 2- or 4-inch shore-supplied hose.
At 0430, the dock supervisor, who was acting as Vopak’s Person in Charge (PIC) for the transfer, boarded the ship and met with the chief officer in the cargo control room. The chief officer and Vopak PIC discussed the planned discharge operation and completed the vessel’s “Ship/Shore Safety Checklist” and the terminal’s “Declaration of Inspection” (the paperwork indicated that nitrogen blanketing would be utilized for the hexene discharge). For communications, the Vopak PIC provided the chief officer with a UHF radio that belonged to the terminal, set to channel 4, which they established would be the primary means of communication between the chief officer and the Vopak PIC throughout the discharge operation. They agreed that the secondary means of communication would be verbal communication at the cargo manifold. The ship’s crew used a separate ship’s radio (and channel) when communicating onboard. Once the discussion and paperwork were completed, the Vopak PIC disembarked the vessel.
Utilizing the shoreside crane, Vopak’s dockman and the ship’s pumpman began making hose connections. The 6-inch hexene and MIBK cargo discharge hoses were connected directly to flanges on board the vessel. A 4-inch nitrogen hose was connected via a reducer to the vessel’s 6-inch vapor recovery line. To control the flow of the nitrogen, the shore used a 4-inch gate valve, and the vessel used a 6-inch butterfly valve. The ship’s pumpman was assigned to operate the cargo manifold valves on the ship, including the nitrogen supply valve, and the Vopak PIC was responsible for the shore valves. All hoses were connected by 0540, and shortly afterward, the terminal completed a personnel shift change. The relieving dockman assumed the role of Vopak PIC from the dock supervisor and completed a handover with his counterpart. There was no requirement to inform the vessel of the shift change at that time, and no attempt was made. Over the next hour, terminal employees on the dock made physical checks of the cargo hoses and communicated with the terminal control room to verify that the terminal’s piping system was lined up correctly.
At 0705, after communicating with the terminal, the ship’s chief officer started the 3P and 3S cargo pumps from the cargo control room to begin discharging hexene to the terminal. The hexene in the cargo tanks was already inerted with an approximately 2 psi blanket of nitrogen, and the tanks, therefore, did not require any additional nitrogen supply from the terminal at the start of the operation. At this time, the vessel’s 6-inch nitrogen supply line valve was CLOSED; the dockman stated that he opened the shore valve “maybe a quarter of the way.”
At 0738, the vessel’s crew started the cargo pump for 1S, and the discharge of MIBK commenced. At about the same time, the low inert gas pressure alarms for both the 3P and 3S cargo tanks sounded in the cargo control room, where the chief officer was stationed. These alarms indicated that the tanks’ pressure fell below the low setting of 0.73 psi (5 kPa) and the nitrogen blanket was being depleted.
Over the next several minutes, the chief officer made numerous unsuccessful attempts to contact the Vopak PIC via radio from the vessel’s cargo control room to request the status of the shore-supplied nitrogen. No attempts were made to verbally communicate from the ship’s manifold (by the pumpman) with the Vopak PIC (dockman). Using the ship’s radio, the chief officer radioed the ship’s pumpman, who was stationed by the vessel’s nitrogen control valve. In an effort to increase nitrogen flow and pressure, the chief officer instructed the pumpman to open the ship’s nitrogen valve, which had been CLOSED. Since the chief officer did not see a change in tank pressure, he instructed the pumpman to fully open the ship’s nitrogen valve, thus making the ship entirely dependent on the terminal’s valve to regulate the flow of nitrogen. At this time, the chief officer did not know whether the terminal’s nitrogen valve was open and if open, what percentage. At about the same time, the terminal’s CCTV footage showed the dockman exiting the dock break shack, which was located at the same end of the ship dock as the shore nitrogen supply valve, moving off-camera to an unknown location, and then returning to the shack 3 minutes later.
At 0748, just 10 minutes after the 3P and 3S low-pressure alarms sounded, the same tanks registered an “ERROR” alarm, indicating a pressure of over 3.2 psi (22 kPa). About the same time, both the 3P and 3S cargo tanks pressure relief valves, which were set to open at 2.9 psi (20 kPa), opened. One crewmember stated that there was “so much noise” coming from the area where the valves were located. The crew also stated that the entire vessel “surged.” Ballast water began flowing through the number 3 port wing ballast tank vent and down onto the deck.
At 0749, the ship’s crew turned off the cargo pumps for both the hexene and MIBK and responded to prevent the accumulating ballast water from spilling over the side of the vessel. After several minutes, communication was re-established between the ship’s chief officer and the Vopak PIC (dockman) via radio. The chief officer informed the Vopak PIC that the vessel had experienced an overpressurization of its 3P and 3S cargo tanks and that the 3P cargo tank had ruptured, releasing hexene into the adjacent ballast tank. Both the terminal and the vessel closed their respective valves, and cargo transfer operations ceased.
In addition to individual cargo tank pressure indicators in the cargo control room, the ship was equipped with an independent cargo vapor pressure-monitoring alarm and recording system, which automatically recorded and stored tank pressure data. After the accident, the system data showed a highest recorded pressure of 15.8 psi in the 3P and 3S cargo tanks.
The terminal’s policy for transfers outlined the special equipment needed and the operational steps for a transfer. The policy also stated that nitrogen could not be supplied to a vessel until a “Nitrogen Supply Acknowledgment” form was completed by the chief officer and the Vopak PIC. The form stated the difference between a nitrogen purge (the procedure for removing dangerous and explosive gases from the interior of tanks) and a nitrogen blanket (pad). In addition, the form stated that the ship would be in complete control of the flow of nitrogen and specified the flow rate of nitrogen supply (250,000 ft3 per hour for a purge or 18,000 ft3 per hour for a pad). This form was not completed for the May 30 cargo transfer operation. The terminal also had “work instructions” for each transfer that took place that provided details of the transfer operation and included directives for the quantity to be pumped and the destination of the transfer. There were two “work instructions” for the accident transfer: one for liquid cargo and one for nitrogen blanketing. The liquid cargo “work instruction” stated that a 2-inch nitrogen hose must be used; the associated nitrogen blanket “work instruction” did not specify a hose size. On the morning of the accident, a 4-inch nitrogen hose was connected to the vessel.
Relief valves are fitted to cargo tanks to protect the tanks from an over-pressurization event. The ship’s RV valves were overhauled and pressure-tested in April 2019 and had been inspected 10 days prior to the accident, with no deficiencies noted. Therefore, it is likely that the RV valves performed as designed. The terminal had no recent history of terminal equipment failure.
Relief valves have a maximum flow rate that CANNOT be exceeded, or the pressure will rise in the tank(s) that the RVs protect. The terminal’s “Nitrogen Supply Acknowledgement” form indicated that the terminal could move 250,000 ft3 per hour through the hose for a purge, and their liquid cargo work instruction stated that a 2-inch hose MUST be used for the nitrogen blanketing. However, on the day of this event, a 4-inch hose was used for the nitrogen blanket supply during the liquid hexene cargo transfer. Since the nitrogen pressure at the dock would not change whether completing a purge or a blanket operation, the use of a larger 4-inch hose on the morning of the accident would have likely allowed for a flow rate near the purge maximum. The ship’s cargo tank relief valves had a capacity of 17,057 ft3 per hour, as indicated on the cargo control room placard on board the vessel.
Therefore, this arrangement allowed for a potential nitrogen flow rate well in excess of the maximum capacity of the cargo tank relief valves. Without accurate and ongoing throttling of the nitrogen control valves, the risk of overpressurization was constant. When a low-pressure alarm was activated on the vessel’s hexene tanks, the chief mate repeatedly attempted to contact terminal personnel via a handheld radio to request more nitrogen, but the Vopak PIC (dockman) did not answer. Since the chief officer could not contact the Vopak PIC, he ordered the ship’s pump man to open the ship’s nitrogen supply valve all the way, after which the pressure in the tank should have risen, but it did not. At this point, instead of securing the vessel’s nitrogen valve and stopping the operation until communications could be restored, the chief officer had the pump man fully open the nitrogen valve, effectively removing all shipboard throttling control of the nitrogen coming on board. With the ship’s throttling ability removed, the combined effect of the nitrogen pressure at the dock, the amount that the terminal valve was open, and the larger 4-inch hose (without a flow-reducing orifice installed before the ship’s manifold) resulted in the rapid pressurization of the cargo tanks, exceeding the relief valve capacity and over pressurizing the tanks to 15.8 psi (109 kPa), as recorded on the cargo vapor pressure-monitoring system.
Since the nitrogen hose connection was improperly configured (without a 1-inch hose or orifice), the flow rate of nitrogen had to be controlled by the ship or terminal personnel by manually adjusting the dock or ship valve. Therefore, communication between the ship and terminal personnel was critical. On the morning of the accident, the vessel’s pump man was directed to open the vessel’s nitrogen control valve wide before communication with the terminal was established and the dock’s valve position verified. Since the valves were opened too far, the nitrogen supply rate exceeded the tank relief capacity design limit. Although both the terminal and operator of the vessel had procedures and control measures in place that clearly outlined a nitrogen blanketing operation, the procedures were not followed on the day of the accident. The terminal’s crew had work instructions that required that a 2-inch nitrogen hose be used to transfer liquid cargo, and the ship’s QHSE manual required its crew to use a 1-inch hose or orifice. However, on the day of the accident, a 4-inch nitrogen hose was connected to the vessel with no orifice, which removed engineered controls designed to limit the flow rate of nitrogen to the cargo tanks safely below their relief capacity.
Additionally, the ship’s QHSE guidelines included a “Nitrogen Handling Checklist,” which outlined the flow rate and other information associated with a nitrogen operation, but the ship’s crew did not complete this checklist on the day of the accident. Had the crew referenced this checklist, it is likely that they would have known that the potential flow rate of nitrogen would have exceeded the cargo tanks’ PV valve relief capacity, and they may have reassessed and corrected the configuration by using the correct hose size or a reducing orifice.
CLICK HERE for the full NTSB report