CFD: Separating the oil & gas streams
Separators are used throughout the oil and gas industry to split production fluids into components of oil, gas and water, as well as contaminants. On an offshore facility, the equipment is found in many parts of the overall process.
The initial separator, usually referred to as first stage, separates the initial stream into distinct gas, oil and water streams. These streams are then individually processed. Poor separation performance can hinder overall production; in some cases, platforms produce only 50% of design capacity due to poor separation.
The industry has used computational fluid dynamics (CFD) extensively to troubleshoot separation equipment performance with different methodologies.
The most common is segregated single-phase simulation, in which gas and liquid phases are analysed separately. Multiphase volume-of-fluids (VoF) simulations are useful in analysing liquid sloshing behaviour in separators secured to moving platforms.
This sloshing analysis is usually carried out in combination with a user-defined function that adjusts gravity and applies three inertial forces: Coriolis, euler and centrifugal. Historically, fluids neither enter nor leave the vessel.
As new separation equipment becomes smaller and flow rates exceed the design capacity of existing equipment, end users are questioning the accuracy of both the segregated single-phase approach and VoF for sloshing.
Extended use of multi-phase simulation is now possible as a result of enhancements to computer power and ANSYS Fluent CFD capabilities.
Software improvements have led to reduced run times; multiphase and turbulence models have a greater ability to handle primary and secondary phases. The multi-phase method overcomes the limitations of segregated single-phase and VoF approaches. It also allows for detailed analysis of inter-phase interactions, providing more realistic results.
Swift Technology Group has studied two types of separation devices that use the multi-phase method.
Droplet separation is fundamental to good separation. The most common equipment for droplet separation is vertical or horizontal vessels that use gravity as the driving force.
More compact separation equipment often uses cyclones. By spinning the flow, employing a standard tangential inlet, or using more-elaborate swirl elements, cyclones can generate accelerations many times that of gravity to potentially provide more efficient separation in a smaller amount of space. However, many other considerations must be investigated.
Traditionally, cyclonic equipment required exhaustive prototyping and testing to ensure that the many negative consequences were designed out of the final product a lengthy and costly exercise.
In a recent R&D programme for cyclone development, Swift researchers found that the time for each design change cycle was about eight weeks at a cost of around £45,000 per cycle, with seven changes required.
By using CFD, each change can be modelled in two weeks, requiring only one actual test saving a total of more than £300,000. However, it was difficult to quantify the exact benefits of simulation in every case.
There are many examples in which the mixture multiphase model has been used to analyse separation within cyclonic equipment. The model is applicable for dilute-to-moderately dense volume loading, for low-to moderate particulate loading, and for cases in which the Stokes number is less than 1.
The simplified model can be used for hydrocyclones equipment whose main function is to separate final oil droplets from water prior to disposal at sea.
The eulerian multiphase model is applicable to the complex flows that are found in the most common types of separation equipment designed to remove bulk phases as well as re-entrained droplets.
Users can enhance their analysis of cyclonic flows by applying the Reynolds stress turbulence model without limitation for all primary and secondary phases.
An important part of separator analysis is often overlooked: the impact of upstream piping. This system has a large effect on the distribution of fluids within the vessel.
The simulation examples provided horizontal and vertical gravity-driven separators as well as cyclone-based separators incorporated the impact of up-stream piping. It was difficult to accurately validate the simulation results of the installed vertical cyclone and separator.
Simulation has been shown to accurately capture both flow field and separation performance in lab and pilot test rigs. Using these modelling strategies as well as exhaustive testing performed over many years, all the critical aspects of the flow are correctly resolved and indicate the key performance characteristics.
As a result, Swift is changing out the internal components of many vessels based upon simulation results. The main function of a horizontal three-phase separator is to split a feed stream into discrete gas, oil and water streams.
Normally, gas is the primary phase, and the two liquid phases are secondary. These liquid phases form droplets that are entrained in the gas phase, and they produce a film on the pipe walls leading to the separator.
The first component in the separator is the inlet device, whose primary function is to provide a coarse separation of gas and liquid phases. The gas phase continues along the top of the vessel, while the liquids drop to the bottom of the separator. At the bottom of the vessel, the two liquid phases separate, with the water at the bottom and the oil forming a layer between the water and gas phases.
In most cases, perforated baffles are used along the length of the horizontal vessel to control liquid phase flows and to distribute them evenly across the available cross-sectional area of the vessel, minimising axial velocity and maximising separation. The eulerian model is required in this type of simulation because of the number of fluid regime changes.
In a vertical production separator with a vane-type inlet device example, gas and liquid are introduced at the start of the pipe run to the separator vessel. The pipe routing causes the liquid to be biased to one side of the vessel which does not produce optimal separation and, in some cases, can lead to the gross carryover of liquid though the vessel’s gas outlet.
David Stanbridge is managing director of Swift Technology Group, Norwich, UK