Pressure surge or water hammer simulations are performed to evaluate a safe design or operation of oil transfer systems. The speed of sound of the transferred fluid plays a key role within pressure surge simulations. Unlike for steel pipe systems, the speed of sound in hoses is not easily determined. In industry, a value of 450 m/s is frequently used for hoses, but this value is often used independent of conveyed fluid, pressure or the type of hose. For that reason, a project was performed that aimed to experimentally reaffirm existing speed of sound data and to extend its usable range by including tests at several static pressure levels. Speed of sound measurements were performed on two 20”, 12 meter long, water-filled oil offloading hoses. The speed of sound was determined by the initiation of pressure waves in the hoses, while performing pressure measurements on either side of the hose. These experiments were performed at different static pressures up to 18 bar(a). The measured values depend on the applied method to determine the speed of sound. For both applied methods, the speed of sound is higher than the classically assumed 450 m/s. The method that is considered best applicable for the purpose of the experiments, the speed of sound is significantly higher and can reach values up to 1100 m/s. The consequence of this is that the results from pressure surge studies in which 450 m/s was used, are not conservative.

CFD analysis of a high pressure 2” pipe test loop with water-gas flow was undertaken using three different solvers. Multiphase flow induced forces were predicted on the bends at a range of operating pressures between 10 and 80 barg and compared with forces reconstructed from vibration measurements. Overall the three different CFD solvers predicted consistent results. The fluid forces predicted on the bends of the double U-loop test rig have a good range of values compared to the test reconstructed forces. The forces predicted at low pressure were in line with the experimental reconstructed values, whilst at high pressure all three CFD solvers predicted higher forces. The trend of the forces reducing with increased operating pressure, evident in test, was matched by one of the CFD methods, but less well by the other two. At low operating pressure the forces are dominated by the momentum of the liquid in the multiphase flow, whilst at high pressure the pressure fluctuations and turbulent effects will be more important. All three CFD solvers use VOF methods and above about 40 barg it is possible that they struggle to fully resolve the flow behaviour, which will be more influenced by bubble and droplet entrainment and turbulence. Multiphase flow can induce high amplitude vibrations in piping systems, potentially leading to fatigue failures. CFD modelling offers a potentially powerful tool to provide the flow induced forces required for assessing and diagnosing multiphase flow induced pipework vibration problems in hydrocarbon production systems.

Flow induced vibration (FIV) from high velocity multiphase flow is a common source of vibration concern in process piping, potentially leading to fatigue failures and hydrocarbon leaks. FIV screening methods tend to be conservative for multiphase flows and are typically only validated for simple single bends at low pressure. FE can predict the response of a system if a sensible forcing function is provided. CFD can be used to predict realistic forcing functions in complex combinations of bends and tees, typically seen in process piping systems. FIV studies were performed on a topside production system operated by Equinor, carrying multiphase flow at high pressure (∼69 bara) conditions, where significant vibration was measured. The study assessed different vibration simulation methodologies, combining FE analysis with forcing functions based on both correlations and CFD simulations. The aim was to gain a better understanding of the accuracy and limitations of calculation methods typically used to assess fatigue. CFD simulations predicted similar force magnitudes but higher frequency forcing at 69 bara compared to equivalent simulations at atmospheric pressure (at the same liquid and gas superficial velocities). The forcing function correlations used do not predict higher frequency forcing at high pressure, which has a significant impact on the predicted vibration. Care is required when undertaking this type of analysis. It is important to have an accurate FE model of the as-built pipework and supports as well as a forcing function which accurately represents the fluid forces on the bends. For the case simulated here the magnitude and peak frequency of the forcing function had a significant influence on the response of the structure. Forcing functions based on correlated data from tests at low pressure should be used carefully for high pressure systems. In addition, the inclusion of phasing of the forces at each bend can influence the structural response, and simulations performed in the frequency domain do not consider this. A combination of CFD and FE modelling offers a potentially powerful tool for assessing and diagnosing multiphase FIV problems in hydrocarbon production piping systems.

The flow-induced pulsations in corrugated pipes result from a feedback loop between an acoustic resonator and the noise amplification at each shear layer in the axisymmetric cavities forming the corrugations. The quality factor of the resonator is determined by the reflection coefficients at the ends of the corrugated pipe and by the damping in the pipe. In this work, the damping of acoustic waves in a set of smooth and corrugated pipes is measured by a direct method. For these measurements, the tested pipes are placed between two measuring pipes equipped with flush-mounted pressure transducers to allow reconstruction of the acoustic waves with the two-microphone method. Loudspeakers are used to generate acoustic waves, and anechoic terminations allow near reflection-free conditions. The tests are done in air without flow and with flow velocities up to 60 m/s. The results for the corrugated pipes allow to investigate the influence of the corrugations on the damping. Two cases are considered: Without flow, the visco-thermal damping is increased for the corrugated pipe compared to a smooth pipe of same diameter. When there is a flow in the pipe, the damping results depend strongly on the flow velocity. At certain frequencies which depend on the flow velocity, the damping increases or decreases. The regions of increase or decrease are shown to be at constant ranges of Strouhal numbers. The decrease of the damping can in some cases be such that acoustic waves are amplified through the corrugated pipe. This corresponds to the acoustic source behavior of the shear layers. The measured amplification is compared to the computed source term due to the axisymmetric cavities.

In the process systems of offshore installations, welded small-bore side branches can prove vulnerable to high-cycle fatigue failure due to vibrations. This is especially the case for welded connections at tie-in points to the main pipe which are often critical details. International standards and guidelines therefore provide maximum acceptable vibration levels to ensure long term safe operation. In some guidelines, however, these acceptable vibration levels are phrased in terms of screening levels and in practice can be unduly conservative. Process pipework might then unjustly be regarded as unsafe based on measured vibrations in the field. This is especially true for offshore systems, which are characterized by low mechanical damping in the structure. This may result in overdesigned piping or over-conservative operational limits in order to keep vibration levels within the acceptable range. Furthermore, the screening methods and any detailed fatigue assessments typically use established stress-life (S-N) based fatigue design methods where uncertainty exists in the very high-cycle regime. This paper describes a novel and advanced tailor-made fatigue assessment method whereby acceptable vibration levels are based on maximum acceptable stress ranges for individual side branches. The acceptable stress ranges for each critical welded connection are based on a fracture mechanics analysis of fatigue crack growth. This method also minimizes the cantilevered (overhung) mass of small-bore side branches, whilst remaining safe for long-term operation. To illustrate the strength of the assessment methodology in practice, this paper describes the application of the procedure to a 2″ side branch connected to a main piping system. A fracture mechanics model and a detailed 3D finite element model are made. By comparing the stress ranges from the fracture mechanics model with the normalized stress ranges obtained from the dynamic FE analysis, maximum acceptable vibration levels for this particular side branch have been derived. The method is validated with experimental modal analysis and strain gauge measurements.

On December 12, 2007, the second largest oil spill in the history of Norwegian oil exploration occurred on StatoilHydro’s Statfjord Alpha platform. The spill was caused by a snapped 20 ″ oil off-loading hose. Thorough investigations by StatoilHydro [1] and by the Norwegian authorities [2] revealed the chain of events that led to this incident. One of the links in this chain was the unintended fast closure of the shuttle tanker’s bow loading valve during off-loading. This closure initiated a pressure surge in the oil off-loading system. As part of the internal investigation by StatoilHydro, TNO carried out a water hammer analysis of the entire oil off-loading system, including the off-loading hoses to the seabed and further subsea piping up to the platform. These simulations revealed that high pressures could occur in the oil off-loading system due to fast closure of the bow loading valve followed by multiple reflections at diameter changes. The maximum pressures were more than 100 bar above the normal operating pressure of 10 bar. The diameter changes were introduced into the oil off-loading system to maximize the off-loading capacity. The results of the water hammer analysis provided the missing link between the fast closure of the valve and the damaged hose and also showed that this damage most likely occurred within 0.5 second after the closure of the valve. Based on the results of this analysis, also other oil off-loading systems are being reinvestigated to prevent a similar incident to occur in the future.