Realize Better Risk Characterization of STHE Tube Failure Scenarios Through Relief Systems Dynamics Modeling

The risks of shell and tube heat exchanger (STHE) sudden tube failure scenarios are often considered in relief and flare systems design and evaluation. Scenarios involving the release of high pressure gas following a sudden tube failure, especially where the shell is filled with liquids, require the use of relief and flow dynamics for better understanding of and assessment of risks, risk reduction, and relief requirements.

Historically in most existing STHE installations, the shell side of the STHE was typically protected with a rupture disk where the design basis is all vapor flow allowing for the expansion of gas from the tube pressure to the set point of the rupture disk. Rupture disks were favored over pressure relief valves because they were assumed to have faster opening and response time to the pressure wave experienced by the liquid following the tube failure.

Recent field work by the Energy Institute confirms that the all vapor relief requirements are not sufficient to protect the shell from deformation and potential failure. Furthermore, this recent field work also confirms that STHE shell may also be effectively protected with pressure relief valves since the measured opening time for PRVs can be on the order of 5 milliseconds while many non-fast acting rupture disk opening times can be as long as 100 milliseconds. Additional findings by the Energy Institute indicate that a reasonable tube failure frequency to consider for risk assessments and design is 1/1000 years.

Following the sudden rupture of one tube, between 0.2 and 0.7 milliseconds, high pressure gas will impact the liquid creating a shock with a duration that depends on the length of the heat exchanger shell and the speed of sound in the liquid within the shell. Small amounts of gas present in the liquid shell can temper the magnitude of this initial shock pressure. Once the pressure wave had enough time to be reflected back to the tube failure source/location, high pressure gas will expand into the shell and will cause liquid displacement through the relief device. The relief device will have to have enough capacity to expel enough liquid to create sufficient expansion volume for the high pressure gas. After a brief duration of liquid flow, two-phase flow begins. Finally, all vapor flow occurs. While this entire quasi-dynamic flow time may be on the order on 1 or 2 minutes, the static pressures developed in the shell can cause deformation and potential failure. While relief requirements are typically driven by the two-phase flow regime the relief piping structural support requirements may be driven by the transient reaction force experienced by the piping during the initial acceleration of the first liquid to be vented (slug loading).

Shell internal stress vs. failure stress at 2/3 UTS.

Dynamic relief systems modeling with SuperChems™ component of Process Safety Office® can provide a lot of insight into the risks and the proper design for this scenario. In many instances, the STHE may not be blocked in during the tube failure and relief through the process line is possible in addition to the relief systems flow path. During a worst case scenario where tube failure occurs and the STHE is blocked in, relief systems dynamics can provide a prudent estimate of the pressure reached in the shell with the existing relief device and if that level of pressure can lead to deformation or potential failure. Potential failure is typically considered if the Hoop stress exceeds 2/3 the UTS at the mixed fluid final temperature with consideration of corrosion allowances for new designs or corroded thickness and remaining life for existing designs. Relief systems dynamics modeling provides a clear assessment of consequence vs. failure frequency considering other layers of protection. Such practical analysis enables operating companies to better judge the risk, to define better risk reduction options, and to prioritize when risk reduction measures (if any) can be implemented effectively without creating additional risks for poorly planned or emergency shutdowns.