Fire at a fired heater of an hydrocracking unit

This incident occurred at the fired heater of a non-further specified reactor of a refinery (probably a hydrodesulphurisation unit).
A tube of the fired box failed releasing a mixture of hydrogen and hydrocarbons into the firebox. The mixture consisted in approximately 51,000 pounds of diesel, 160 pounds of hydrogen and 560 pounds of methane. It was ignited by the burners, resulting in a large fire involving other part of the unit.
The reason for the failure of the tube was overheating possibly in combination of creep. On the day of the incident, the unit had automatically shut down due to a problem in another part of the process. During this shutdown, the hydrogen and hydrocarbons flowing through the fired heater’s tubes stopped, but the burners continued operating because the fuel gas control valve did not fully close. Without fluid flow through the tubes to remove heat, the tube’s temperature exceeded 1,400℉ (), causing a short-term overheating and degrading tubes’ integrity. When the fired heater was restarted, one of the tubes ruptured.

The INITIATING CAUSE was a failure of a fired heater tube due to overheating, due to the failure of a valve expected to stop the burners.
The ROOT CAUSE was lack of proper knowledge/assessment of the process parameters (temperatures and flows) and consequent failing to optimise the process.
High temperature on tubes had already been measured in the past but disregarded as fallacious. The high values were misinterpreted as indicative of the oxide layer formed at the external surface of the tube, rather than of the tube material. Therefore, the company failed to adjust the burner parameters to avoid temperature excess.

tube, valve, fired heater

Four months before the incident, a contractor performed an infrared scan of the fired heater, finding elevated temperatures in the heater, with one tube section operating above 1,300 degrees Fahrenheit (ca. 705 Centigrade). At the time, the contractor concluded that the high temperatures were measurements of the scale and oxidation on the outside surface of the tubes, not the tube’s metal wall temperature. After the incident, PBF Energy determined that the IR temperature measurements taken before the incident were likely accurate, but they had been misinterpreted. As a result, the infrared temperature data was not used to adjust the operating conditions of the fired heater, which could have lowered the tube temperature within the design limit.

Fires in fired heaters are a recurrent type of incident in the processing industry throughout decades. See for example the HIAD event 656 and 655 occurred respectively in 1970 and 1971, 691 of 1980, and more recently cases reported by the CSB: 1167 , 1168 and 1169 of 2021, 1183 in 2023 and 1178 in 2024 . They are almost all characterised by failure of a radiant tube to overheating, often due to the incapacity to realise a reduced mass flow in the tube and/or to tune the radiant power to different, transitory situation.
For all these cases, improvement recommendations focussed on improving instrumentation for temperature and flow measurements and on a finer management of the burners.
This case is peculiar in the sense that too high temperatures had already been measured by infra-red readers four months before the incident. It would have been the occasion to reduce burners power to avoid degradation of tube materials. On the contrary, the high temperature values were misinterpreted as indicative of the oxide layer formed at the external surface of the tube, rather than of the tube material. Moreover, in absence of burners’ fuel monitoring, the failure of the valve expected to stop fuel to the burner remained undetected. A new peak in temperature, caused by the malfunctioning od fuel valve, was then fatal to the integrity of one of the tube.
The case rises also questions on the relationship between plant operator and contractors. The temperature measures had been performed by a contractor, supposed to be an expert in this type of measurements. Who then decided to disregard the measurement?

After the incident, the company installed larger fired heater viewports to allow for improved infrared scans of the tubes and installed instrumentation to monitor tubes temperature.