**3. Full scale experimental and** *in-situ* **monitoring techniques**

#### **3.1 Recirculation tests**

Closed loop pipeline recirculation rigs are a popular choice for those looking to accurately replicate industrial conditions. Tests are conducted within specimen sections which can be inspected periodically. Various techniques have been used to investigate E-C damage within the specimen section. For example, Wood et al. [54] used a combination of gravimetric, ultrasonic thickness and diameter measurements taken using a ball-ended micrometre to quantify material loss within stainless steel test sections. The latter two measurements allowed for damage mapping across the wetted area, which showed good agreement with results from a CFD model, demonstrating that their method could be used to predict erosion damage patterns within pipe sections.

Owen et al. [55] developed a 3D printed pipe elbow to aid with prototyping and flow mapping within pipe sections. Multiple samples of X65 carbon steel were inserted into the elbow. Mass loss measurements at each sample location were used to map erosion damage across the sample section. While their results showed reasonable agreement with those expected from existing literature, it is questionable if their methodology provides any benefit over that used by Wood et al. [54] for the following reasons: the samples were only installed on the inner and outer radius of the elbow, so are incapable of detecting erosion to the elbows sides; it is not certain that the 3D housing, its material, and the geometry of the samples in the housing accurately replicate the flow in real elbows. No measurement of material loss from the housing was provided, so changes in flow condition due to housing erosion could not be ruled out. Rebound characteristics from the 3D printed housing may also be different to a full steel elbow and could affect erosion rates downstream of the first particle-wall impingement. With this being said, due to the ease of 3D printing, this method may be of use in rapid prototyping applications. It also allows for rapid substitution of test material samples for comparison tests.

Elemuren et al. [14] used a continuous recirculation rig with four 1018 steel 90° elbows to investigate the influence of slurry velocity on E-C. A slurry of potash brine with silica sand was used. Gravimetric measurement was used to quantify mass loss from each elbow. The synergy between erosion and corrosion was investigated by comparing results from separate erosion (deoxygenated potash brine w/10% wt. solid) and corrosion (saturated potash brine w/0% wt. solid) tests. It was found that the synergy decreased with slurry velocity, reducing by 45% between 2.5 m/s to 4 m/s. At high velocities, surface impingement became the dominant mechanism of material loss. Erosion-corrosion was found to be a function of solid concentration and flow velocity.

Regions in which flow trajectories are severely altered are found to be the most prone to erosion damage. Such areas include tees, elbows, and contractions, where changes in flow direction mean entrained particles are more likely to cross streamlines and contact the solid walls. Bilal et al. [56] investigated the influence of pipe geometry on wear due to water-sand and water-air-sand flows in pipe bends. They aimed to investigate alternative geometries to standard 90° elbows, where the curvature radius is equal to 1.5 times the pipe diameter, *D*. Geometries of 2.5D and 5D along with a 45° elbow with a radius of 1.5D were investigated. Numerical studies were carried out using commercial CFD code ANSYS FLUENT. Results from the CFD study were compared against experimental results. The inner surfaces of acrylic pipe bends/elbows were painted with black paint primer. Solid particles entrained in the flow remove the paint from the internal surfaces and allow erosion patterns to be observed. The tests were run until no further change occurred in the paint pattern. This gave a qualitative indication of which pipe geometry is least susceptible to erosion. It was found that smoother transitions in pipe geometry reduced erosion rates as particles are less susceptible to hitting pipe walls. Additionally, smaller changes in direction (45°) led to lower erosion. Generally, CFD results were found to corroborate the experimental paint removal study. An exception to this was that CFD failed to predict some of the secondary erosion patterns downstream of elbows. Wall function (modelling of flow in the near wall boundary region) was also found to affect the observed erosion patterns, producing a secondary hotspot on the inside of the elbow that wasn't present in the paint removal study. This was put down to be an artefact of the turbulence model used.

In industry, a wide range of pipe geometries are used in a variety of orientations. It is known that erosion properties can be quite different depending on the orientation of a section. Reportedly, most studies examining erosion in elbows have placed the sections in the horizontal-vertical configuration. Wang et al. [2] studied the effect of orientation on erosion in pipe elbows in the horizontal-horizontal configuration. Two elbows connected in series were analysed. Additionally, the effect of particle concentration and inlet conditions were also examined. CFD simulations were verified against experimental results. The effect of gravity became less noticeable, and the erosion profile became more symmetric across the pipe section as inlet velocity increased. The downstream elbow was found to be severely influenced by the flow profile exiting the previous elbow. The distance between the two had a significant influence on this effect.

To accurately model pipeline decay in real-world operation, factors outside those discussed above should be considered. External factors such as human, geotechnical, and natural hazards are all known to pose a risk to pipeline integrity [57]. External corrosion related to environmental conditions is known to be one of the most significant factors governing pipe life. Methods of monitoring and inspecting

pipelines were reviewed in [58]. The same article proposed a risk-based integrity management model.

Yang et al. [6] carried out a performance investigation for slurry erosion in highpressure hydraulic fracturing pipelines. Unlike other papers which studied slurry erosion, the effects of tensile stress induced due to internal/external pressure was considered. Additionally, the effects of impact angle, particle velocity and concentration, and target material were also considered. Results from an experimental study were used to develop a random forest regression (RFR) algorithm to develop an erosion prediction model for fracturing pipelines. A jetting circulating erosion test rig was used to apply adjustable tensile stress on the specimen during erosion. Test samples consisted of 4 different hypereutectoid steels. The effect of impact angle was evaluated on a sample of unstressed 42CrMo steel. Erosion wear was found to reach its maximum at an impact angle of 30°. At low angles, micro-cutting was found to be the prominent erosion mechanism, while plastic deformation caused erosion at higher impact angles. Erosion rate was found to be proportional to tensile stress. At 30° the increase was approximately 70% between 0 and 500 MPa, while it nearly doubled for the same stress range at impact angles of 90°. This is due to a lower yield force under high stresses being more easily overcome by the kinetic energy of impacting particles.

#### **3.2** *In-situ* **damage monitoring**

Online condition monitoring (OCM) allows for the health of critical components to be continuously monitored. OCM systems generally rely on an array of sensors linked to a processing device to display output parameters to the user. Readings can be used to schedule repairs and detect early signs of failure, reducing costs of manual inspection and safety/economic implications of unplanned downtime. The benefits of OCM are numerous and its implementation has become common practice in several industries. Liu et al. [59] proposed a system to allow for the online monitoring of metal loss in pipelines. Ring pair electrical resistance sensor (RPERS) arrays were used to monitor erosion loss around the diameter of the pipe, while linear polarisation resistance (LPR) was used to measure the corrosion rate in a pipe ring. Electronic measurements were compared to weight loss measurements. The X65 steel test pipe was embedded in a recirculation rig, and CFD studies were used to calculate flow distribution and sand concentration in the section. LPR, RPERS and gravimetric measurements showed good agreement for this configuration. It should be noted that the results cannot be generalised as they only relate to the specific configuration and conditions used.

## **4. Conclusion**

Erosion-corrosion research in the past decade has been dominated by applicationspecific studies. Most of these papers have focused on comparing different materials or pipe geometries under specific flow conditions. Laboratory testing techniques such as slurry pot testers or electrode cells can be used to provide an understanding of factors influencing E-C within several materials. Due to the differences between laboratory and in-service conditions, the applicability of results obtained to realworld scenarios should be checked.

It is well known that pipe geometry has a significant effect on erosion rate. Despite this, relatively little research has been done to investigate alternative pipe

### *Erosion-Corrosion in Pipe Flows of Particle-Laden Liquids DOI: http://dx.doi.org/10.5772/intechopen.107231*

geometries. Work by Li et al. [50] exploring the influence of internal bumps in a pipe elbow showed that erosion rate could be reduced with well-designed internal surface modifications. This could be implemented in particularly vulnerable pipe sections and should be explored further. As demonstrated by Wang et al. [2] pipe orientation and flow conditions entering pipe elbows can heavily influence the resulting erosion pattern. Despite this, few papers focus on the role of pipe geometry upstream of subject sections on erosion. By altering the upstream configuration, it is conceivable that one could alter the erosion pattern to increase component lifetime with relatively little difficulty. Perhaps this could be the subject of future work.

Recent improvements in the area indicate a promising future for CFD erosion modelling. Improved impact models, dynamically deforming geometry and more accurate particle tracking have all helped improve the accuracy of CFD erosion predictions. Despite these advancements, more work is required before CFD can be heavily relied upon. Errors in near-wall erosion models have been identified and require rectification. Loosely representative empirical erosion models are still heavily relied upon for wall erosion modelling and limit the applicability of the technique for a wide range of materials. While CFD-DEM allows the modelling of more dense slurry flows, the computational expense limits its use to small-scale modelling. To reduce computational expense, coupled DEM-DPM appears promising, but its application to liquid-solid flows is yet to be thoroughly explored.
