Flow Regime Affects Erosion Prediction in Multiphase Flow

Topics: Flow assurance

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

Erosion caused by fine solid particles presents one of the greatest threats to oil and gas flow assurance, consequently affecting material selection and wall-thickness design. The authors of this paper conclude that the erosional effect caused by microsized particles is dependent on the flow patterns in the pipe, determined by superficial velocities of each phase. The results of the proposed modeling using multiphase computational fluid dynamics (MCFD) are expected to benefit erosion-impact assessment in multiphase hydrocarbon-production and -piping systems.


Understanding the effect of key parameters that regulate erosion mechanisms is essential. Studies addressing these issues have concluded that particles rebound and that the erosion profile is dependent on particle motion inside oilfield conduits. Researchers also have found that the direct-impingement (semiempirical) model agreed with results achieved with the discrete-phase model implemented in CFD, while the pure empirical model severely underpredicted the erosion, thus emphasizing the importance of flow-behavior modeling. Furthermore, the majority of available erosion-prediction models involve single-phase systems, not taking into account multiphase flow patterns that address the characteristics of individual flow behaviors.

In other studies, the erosional effect caused by microsized particles in single-phase carrier fluid also has been reported. One hypothesis is that the presence of these microsized particles acts as an enabler, which produces homogeneous pits on the surface of metal, increasing significantly the contact surface area upon which chemical and mechanical interactions can take place.

The effect of multiphase flow and its interaction with sand particles, specifically fine solids, is often neglected. Hence, integrating both the erosion and fluid-flow models is essential in enhancing the accuracy of erosion evaluation.

MCFD Model for Erosion Prediction

In a pipeline system, the eroded spots caused by particles are detected typically at locations with restricted flow characteristics or radically changing flow directions, where the bend or elbow is identified as an erosion-susceptible area. Hence, in the present study, a 1-in. elbow is designed with a horizontal inlet and vertical upward outlet orientation. A length of 10 times the elbow internal diameter is placed upstream and downstream of the 90° elbow.

The computational procedure for this work includes MCFD simulation, where an annular flow regime of air and water is modeled in the elbow, incorporated with the division of grids to represent predefined flow conditions (Step 1). By tracking particle impingement within the carrier fluids (Step 2), erosion rate is calculated and analyzed (Step 3).

Sand particles are carried and transported by carrier-fluid solutions. Thus, fluid characteristics contributed to the changes of particle/particle interactions and their behavior, resulting in variations of erosion pattern and magnitude on the target surface. Before assessment of erosion caused by fine particles, annular flow characteristics flowing through the 1-in. 90° elbow are developed with software. In air/water annular flow, sand particles are transported in liquid phase. Velocity increases further through the central core; therefore, particle velocity tends to be accelerated by liquid droplets entrained in high gas streams, causing significant erosion impact on the elbow area.

Air was chosen as the primary phase, and water was chosen as the secondary phase. The air/water mixture flow is simulated to introduce an annular flow regime. The Eulerian/Eulerian multiphase approach is used, considering the interface transfer of heat, mass, and momentum between air and water. The erosion prediction at the elbow is affected significantly by turbulence in the dispersed phase. As such, the shear-stress transport model is used in this study to enhance calculation accuracy of the near-wall turbulent shear stress.

After the annular flow model is established correctly, the next step is to introduce fine-particle injections and the Lagrangian particle-tracking model. The concentration of semiround-shape injected sand of 0.013 vol% with various particle sizes ranging from 50 to 20 μm is modeled. An additional case of 150‑μm sand also is considered for comparison purposes.

Results and Discussion

Flow-Regime Development Affecting Erosion Prediction. Erosion-rate evaluation was begun with the analysis of flow-regime development in the elbow. This analysis was conducted to study the effect of iteration steps on the calculated erosion-rate density. To study the effect of the iteration step, several simulation runs for various particle sizes were performed with a varying number of iteration steps.

Fig. 1 shows the effect of iteration steps on the estimated erosion-rate density for representative particle sizes of 150, 50, and 30 μm in an annular flow regime. It is observed that before 2,500 iteration steps, when the first gas stream hits the elbow, particle trajectories were unstable. When the generated bubble columns associated with liquid slug (disturbance wave) passed through the elbow, followed by annular liquid flow afterward, more-stable particle trajectories approached the elbow, and thus lower erosion-rate density was observed.

Fig. 1—Estimated erosion-rate density for representative cases vs. iteration steps.


Erosion-rate density at the elbow was also predicted for particle sizes of 30 and 150 μm during transient time, from flow developing to flow fully developed. In general, the impingement of fine particles is more focused and concentrated compared with that of 150‑μm particles. This is the result of the turbulent intensity vortex concentrating the fine and light particles while the larger and heavier particles are influenced and distributed to a wider area of erosion hot spots. During the transition time between the 2,500th and 3,500th steps, erosion-rate density decreased considerably, corresponding to the development of annular flow. Global erosion rate may be determined by performing steady-state simulations after the minimum 3,500 iteration steps to achieve fully developed annular-flow conditions, whereas the local erosion phenomena were identified when flow developed between the 2,500th and 3,500th steps.

Particle Size Affecting Erosion Prediction. To validate the ­calculated erosion-rate density using the CFD software, results (for both single-phase-gas only and an annular flow regime) were benchmarked against Tulsa Erosion-Corrosion Research Center Model predictions.

The size of particles impinged onto the elbow surface defines the degree of erosion impact at the elbow in fully developed flow. At smaller particle sizes, lower erosion-rate density was predicted. For single-phase flow, a similar trend for erosion rate calculated with the software model in single-phase fluid and the Tulsa model is observed, although discrepancies also were observed. The software model shows an overpredicted erosion rate for single-phase carrier fluid. While, in a multiphase flow regime, erosion rate decreases with smaller particle size compared with sand impact in single-phase flow, at 50 μm the erosion rate begins to exhibit an increasing trend of decreasing particle size. Above a certain level of particle size (i.e., 150 μm), the electrostatic attraction may exceed the van der Waals attraction, thus causing a higher dispersion and a wider range of the erosion-impact area.

On the basis of the calculations, a more-conservative erosion-rate density was predicted with the software for single phase in comparison with multiphase simulation. Discrepancy between erosion rates predicted using the software in an annular flow regime and those predicted with the Tulsa model is observed. This is because the Tulsa empirical models only factor limited effects of multiphase flow into calculations in which the mixture fluid properties are considered, whereas CFD prediction considers the local and instantaneous sand-particle impact. However, it must be noted that less-empirical models take multiphase-flow regimes into account.

Finer particles (less than 50 μm) cause more-localized erosion compared with particle sizes of 150 μm. Thus, under current flow conditions, erosion impact caused by fine particles is noticeable. Fine particles are highly influenced by turbulence; therefore, simulating an accurate flow regime will affect erosion prediction considerably.

Sand-Mass-Flow Rate Affecting Erosion Prediction. Analyzing the effect of sand mass flow rate on erosion prediction is necessary in the highly turbulent environment of the annular flow regime. Two sensitivity cases were performed, considering an increase in sand-mass-flow rate of fine particles from two to five times, while the sand-mass-flow rate of 150-μm particles remained constant.

Although fine particles are smaller and lighter in weight, as fine-particle-mass-flow rate increased, a significant increase in erosion rate (also two to five times) for fines was reported in comparison with large particles. Turbulent eddies will lead the sand particles to impact the bend. As such, fine particles can speed up easily and more-severe impact can be imposed.


This paper investigated the erosional impact caused by fine particles in an annular-flow regime using an MCFD approach, covering various particle sizes less than 62.5 μm. The integration of the multiphase-fluid-flow model and erosion calculation using the MCFD approach improved the accuracy of the prediction of erosion caused by fine particles. In particular, erosional patterns and the potential for erosion spots at the elbow were predicted, which will be beneficial to material selection and wall-thickness design. Results revealed that erosion prediction is strongly affected by the flow regime. Additionally, fine particles are influenced significantly by turbulence; therefore, increasing the mass-flow rate can increase the speed of fine particles and their subsequent erosional impact.  

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper OTC 28360, “Investigation of Erosion Caused by Fine Particles in Multiphase Flow,” by Tran Nguyet Ngo, SPE, and Deng-Jr Peng, Intecsea, and William Pao, Hamdan Ya, and Rohaizad M. Norpiah, SPE, Petronas, prepared for the 2018 Offshore Technology Conference Asia, Kuala Lumpur, 20–23 March. The paper has not been peer reviewed. Copyright 2018 Offshore Technology Conference. Reproduced by permission.

Flow Regime Affects Erosion Prediction in Multiphase Flow

01 November 2018

Volume: 70 | Issue: 11


Don't miss out on the latest technology delivered to your email weekly.  Sign up for the JPT newsletter.  If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.