Development of an Erosion Dynamics Model and its Application to Wells and Facilities

Wang, Jason (ExxonMobil) | Sami, Muhammad (ANSYS) | Troshko, Andrey (ExxonMobil) | Gallo, Federico (ExxonMobil) | Mayer, Christian (ExxonMobil) | Tenny, Matthew (ExxonMobil)



When producing hydrocarbons from an oil well, managing erosion of both surface and subsurface components caused by solids in the flow stream is critical to maintaining operations integrity in both land and offshore assets. Although component lifetime prediction has advanced in the past few decades, the prediction's accuracy remains a major oil and gas industry challenge. Current computational models only provide an initial erosion rate which is usually assumed constant until equipment failure. However, observed erosional rates vary as a function of time due to the geometrical changes caused by equipment material loss, which result in variations in solid particle impingement velocity [1] thereby either accelerating or slowing the erosional process. The constant rate simplified erosion model often produces inaccurate results that can lead to unexpected equipment failures or unnecessary equipment upgrades depending on whether the rate accelerates or decelerates. Therefore, developing a transient erosion model to capture the variations of erosional rate is needed for an accurate prediction of equipment lifetime.

This paper presents an implementation of an erosion dynamics model in ANSYS FLUENT, a commercial computational fluid dynamics (CFD) software, to capture the progression of transient erosion. The model has the capability to capture the effects of surfaces receding from erosion at each time interval. By dynamically adjusting these surfaces and recalculating the local flow conditions in the area, this method can predict new erosion rates for each time interval and achieve fully coupled geometry-flow-erosion interactions.

This new erosion dynamics model was validated against experimental data from both literature and physical testing, and was determined to have accurately captured the observed erosion trends over time in terms of location and magnitude. The model was then employed to study two real world applications: 1) in evaluating the erosion risk for a high-rate water injector, it predicted the evolution of damage to a coupler designed to connect different diameter pipes, and 2) in analyzing facility piping systems connected to an unconventional well, it predicted the transient erosion trend from proppant flowback, which allowed for pipe geometry optimization to increase in erosional life expectancy.