In the chemical and energy industries, liquid products such as LNG, ammonia, or naphtha are commonly transferred between ships and onshore storage tanks. These transfers are performed through dedicated loading arms connected to long pipelines. When the process involves offloading—i.e., transferring product from a ship to a storage facility—there is a particular risk associated with water hammer events due to the length of the piping, the high flow rates, and the emergency shutdown scenarios.
These pipelines often span several kilometers from the quay to the storage tanks, crossing infrastructure and featuring elevation changes and expansion loops. These geometric details are essential to accurately capture dynamic behaviors such as water hammer. Unlike short process lines in a plant, these longer pipelines are more susceptible to large pressure transients because of their mass, length, and the potential for trapped gas or vapor pockets.
At Dynaflow Research Group, we often build detailed 3D geometrical models of such systems. Unlike in process engineering, where 2D schematics may suffice due to the lack of mechanical stress considerations, our models account for both flow and mechanical response. These 3D representations are essential for understanding the forces acting on the system and conducting accurate stress analyses.
The model includes the full transport pipeline, which may feature multiple expansion loops and elevation changes such as bridges—elements commonly encountered in real-world installations to cross roads or other infrastructure. Including these elevation changes is critical to accurately capture pressure variations due to height differences, which directly influence the behavior of transient events like water hammer.
In steady-state operation, flow velocities are commonly around 3 m/s. Faster loading and unloading reduce ship docking times and operational costs, which incentivizes high flow rates. However, this also increases the risk and severity of water hammer. When a valve is closed—especially quickly—the momentum of the moving fluid must be abruptly stopped. This deceleration requires a force, which manifests as a pressure surge that travels through the liquid at the speed of sound in the fluid, typically around 1000 m/s. The faster the valve is closed, the less time the fluid has to decelerate, resulting in a larger pressure spike. This transient pressure wave can cause significant mechanical stress in the piping system, making accurate modeling and mitigation strategies critical.
The flow rate determines the initial momentum of the fluid in the pipeline. The faster the fluid is moving, the more momentum it carries—and the harder it is to stop when a valve closes. When the valve closes rapidly, the change in velocity (Δv\Delta vΔv) over a short time leads to a sharp pressure spike.
This surge can be estimated using the Joukowsky equation:
A higher flow rate results in a greater, and therefore a larger, pressure surge. This is why high flow operations—while economically efficient—require special attention to water hammer risk.
The length of the pipeline plays a crucial role in how long the pressure surge lasts and whether it becomes amplified. This is due to the concept of critical time, which refers to the time it takes for the pressure wave generated at the valve to travel to the tank (or other boundary) and reflect back.
Where:
If the valve closes faster than the critical time, the pressure wave has no time to return before the closure is complete. In this case, the full Joukowsky pressure is realized because the entire momentum is abruptly stopped.
If the valve closes more slowly than the critical time, the returning pressure wave can interact with the valve motion, potentially softening the pressure surge—a form of self-damping due to wave interference.
In long pipelines, the critical time is longer, meaning that rapid valve closures are more likely to fall below this threshold, triggering the full intensity of water hammer. Also, long lines contain more moving fluid mass, increasing the potential energy that must be absorbed or redirected.
Valves within these systems can be located at multiple points along the pipeline, depending on operational preferences and safety requirements. For instance, if the tank owner prioritizes safeguarding the storage facility, a valve may be positioned close to the tank.
However, there is no universally optimal valve location; each system must be evaluated individually. Ideally, multiple valves should be installed throughout the line to compartmentalize sections and reduce the forces generated during a water hammer event.
One of the most critical valve types in these systems is the Emergency Shut Down (ESD). ESD valves are typically installed in pairs and are designed to close rapidly during emergency situations. Due to the need for fast actuation, these valves are often ball valves, which require lower actuation torque compared to other types like butterfly valves. Butterfly valves, while compact, need greater force to rotate under pressure, making them less suitable for emergency applications.
To analyze the system’s response to ESD valve closure, we examine two simulation cases: one idealized and one physically realistic. In the non-physical (idealized) case, cavitation is ignored—this means absolute pressures are allowed to fall below the vapor pressure of the fluid, resulting in unrealistic negative pressures as if the liquid could be “pulled” apart. In the realistic case, pressure is bounded below by the vapor pressure, preventing cavitation.
At the moment the Emergency Shut Down (ESD) valve begins to close (defined here as time zero), the flow rate initially remains almost unchanged, resulting in only a slight pressure variation. However, as the valve approaches full closure—typically between 70% and 100%—the remaining fluid momentum is abruptly halted. This sudden deceleration generates a sharp pressure surge, which is the hallmark of a water hammer event.
In the red (non-physical) curve, once the valve is fully closed, the pressure wave oscillates harmonically—rising and falling as it travels back and forth along the pipeline. These oscillations gradually decrease in amplitude due to frictional damping. However, because the flow rate is relatively low, the friction forces are weak, and damping is slow.
The black curve represents the realistic case -Which can be analyzed using a software such as BOSfluids – where the pressure is constrained by the vapor pressure limit. When this limit is reached, vapor cavities form in the liquid. This typically occurs at high points in the system, where pressure is naturally lowest and further reduced by the rarefaction (sub-pressure) wave.
Here’s what happens in detail:
As the pressure wave causes a sudden deceleration of the fluid column (which is no longer being “pushed” toward the tank), flow velocity decreases. Eventually, the flow may reverse, forming a vapor cavity. Once enough pressure builds back up—due to reflection or compression—the vapor rapidly condenses back into liquid. This violent phase change results in a sharp pressure spike, or implosion, visible as a sudden jump in the black curve. This process is both damaging and difficult to predict without accurate transient modeling.
In steady-state flow, the pressure difference between two points in a pipeline—such as across an elbow—is primarily due to shear forces acting between the moving fluid and the pipe wall. These shear forces (friction) cause a gradual pressure drop along the pipe. Across elbows, the flow changes direction, but the forces on either side are typically balanced, resulting in no significant net mechanical force acting on the elbow or its supports.
However, during a transient event—such as the sudden closure of a valve—this balance is disrupted. When the valve closes, a pressure surge is generated and propagates upstream through the fluid. As this pressure wave reaches the first elbow upstream of the valve, it creates a sudden and asymmetric force on the bend.
This happens because:
As a result, a net unbalanced force acts on the elbow, which was not present under steady-state conditions. This force can be large enough to stress supports, anchor points, or even cause mechanical damage if the system was not designed to handle such dynamic loads.
Understanding these unbalanced forces is essential for structural integrity assessments and for designing restraint systems capable of withstanding transient loads caused by water hammer.
In conclusion, the analysis of water hammer phenomena in offloading arm systems reveals critical insights into the challenges faced during liquid transfer operations in the chemical and energy industries. The study emphasizes the significant risks associated with high flow rates, long pipeline lengths, and rapid valve closures, all of which can lead to severe pressure transients capable of causing mechanical stress and potential damage to the infrastructure.
The findings highlight the necessity for advanced modeling techniques, particularly 3D simulations, to accurately predict and address the impact of transient events. These models provide invaluable information for understanding the dynamic behavior of the system, allowing for the optimization of valve configurations and the implementation of effective mitigation strategies to reduce the risks associated with water hammer.
Moreover, the importance of valve placement, particularly for Emergency Shut Down (ESD) systems, is underscored as a key factor in managing pressure surges. By compartmentalizing the pipeline with strategically located valves, operators can better control the forces generated during transient events, enhancing the overall safety and reliability of the liquid transfer process.
Ultimately, a thorough understanding of water hammer effects and their implications for pipeline design and operation is essential for maintaining the integrity of offloading systems. By adopting comprehensive analysis and design approaches, stakeholders can ensure the safe and efficient transportation of liquid products from ships to storage facilities, mitigating the risks associated with water hammer and enhancing operational efficiency.
Phone: +31 (0)85 058 0046
E-mail: infoaanvraag@dynaflow.com
Laan van Oversteen 20
6th floor
2289 CX Rijswijk
The Netherlands
© Dynaflow Research Group BV
