Investigation of the Severity of Wax Deposition in Bend Pipes Under Subcooled Pipelines Conditions

Nura Makwashi, Kwame Sarkodie, Stephen Akubo, Donglin Zhao, and Pedro Diaz

SPE Europec featured at 81st EAGE Annual Conference & Exhibition

Abstract

Curved pipes are essential components of subsea process equipment and some part of production pipeline and riser. So far, most of the studies on of wax deposition and the possible mitigation strategies have been carried out using straight pipelines, with little attention given to curved pipes. Therefore, the objective of this study is to use an experimental flow loop designed and assembled in the lab to study and understand the mechanisms and variable parameters that affect wax depositional behaviour under the single-phase flow. Series of experiments were carried out with pipes curvatures of 0, 45 and 90-degree at different flow rates (2 and 11 L/min). The sequence in which the bends are incorporated creates non-uniformity of boundary shear, flow separation, and caused isolation of fluid around the bends that affect wax deposition, which depends on flow regimes – Reynolds number along with the radius of curvature of the bend. Prior to the flow loop experiment, the waxy crude oil was characterized by measuring the viscosity, WAT (30℃), pour point (25.5℃), n-Paraffin distribution (C10 - C67), and the saturated/aromatic/resin/asphalte (SARA) fractions

Results of this study shows that the wax deposit thickness decreases at higher flow rate within the laminar (Re<2300) and turbulent (Re>2300) flow regimes. It was observed that the deposition rate was significantly higher in curved pipes, about 8 and 10% for 45 and 90-degree, respectively in comparison to the straight pipe for all flow conditions. Increase elevation of the curved pipe, however, led to a more wax deposition trend; where a higher percentage of wax deposit was observed in 45-degree compared to 90-degree curved pipe. This trend was due addition of gravity forces to the frictional forces - influenced by the physical mechanisms of wax deposition mainly molecular diffusion, shear dispersion and gravity settling. From the results of this study, a new correlation between wax deposit thickness and pressure drop was developed. A relationship was established between wax deposit thicknesses, bend angle in pipes and wax deposition mechanisms with a reasonable agreement with published data, especially for steady state condition. Therefore, this study will enhance the understanding of the wax deposition management and improve predictions for further development of a robust mitigation strategy.

2

IntroductionOil and gas industrial activities in the offshore and deep-water fields have increasingly expanded due to the proven excess reserves in such environment and the decline in production from the onshore and shallow water reservoirs. This is also attributable to the advancement in geological surveying, exploration and production technologies in such complex field (Buller et al., 2002; Bai and Bai, 2005; Lee, 2008; Guan, 2016; Rittirong et al., 2017). In this typically cold environment, hydrocarbons experience drastic temperature changes at seabed, which is usual around 4 – 16℃ (Lee, 2008; Cawkwell and Charles 1989) and the increase in pressure drops from the reservoir to surface facility is extremely large (Hammami et al., 2003; Theyab and Diaz, 2016; Teo, 2016;). Consequently, drops in temperature and instability of pressure among other issues causes unfavourable flow assurance problems during crude oil production, transportation and storage. The key flow assurance concerns affecting oil and gas production as mentioned by Bai and Bai (2005) and Watson et al. (2003) include; (i) system deliverability: pressure drop versus production, pipeline size and boosting. (ii) thermal behaviour: temperature distribution, temperature changes due to start-up and shutdown, insulation option and heating requirements, (iii) production chemistry: hydrates, waxes, asphaltenes, scaling, sand, corrosivity and rheology, (iv) operability characteristics: star-up, shutdown, transient behaviour (e.g. slugging) etc; and (v) system performance: mechanical integrity, equipment reliability, system availability etc.

This work deals with wax precipitation and deposition problem that mimic subsea pipeline condition. Wax (also known as paraffin wax) deposition as reported by Lira-Galeana et al. (1996) is an old problem in the oil and gas industry (Fagin, 1945; Goldman, Marcene S., 1957; Ford et al., 1965) which is still under scrutiny among the researchers and the industrial practitioners (Chala et al., 2018; Wang et al., 2019). Generally, waxes are soluble in crude oil at reservoir condition (70 – 150℃ and pressures of around 50 – 100 MPa). Usually, wax precipitation begins as the solubility of the wax molecules drops due to temperature gradient between the flowing fluid and the surrounding temperature (Rittirong et al., 2017; Zheng et al., 2016). The precipitation process causes the oil viscosity to increase, which increases the pressure drop across the pipeline section. However, in the absence of any mitigation strategy, the precipitation could lead to the deposition of the wax crystal (controlled by several mechanisms) within the inner surfaces of the transport facilities (Zheng et al., 2016; Singh et al., 2000). The later process could ultimately reduce the oil recovery, cause production shutdown, increase operational costs, and negatively affect health, safety, and environment (Lee et al., 2008; Burger et al., 1981; Majeed et al., 1990; Zheng et al., 2013). In the worst case, the deposit could completely block the pipeline (as shown by this study) leading to equipment breakdown, loss of production that requires costly treatment (such as unblocked and removal of wax) or replacing a section of pipeline in order to continue production (Singh et al., 2000; Lee et al., 2008; Adeyemi and Sulaimon, 2012; Rittirong et al., 2017; Rashidi et al., 2016). Some of the fundamental factors that affect wax precipitation and deposition as revealed by Bai and Bai (2005), Rittirong et al. (2015), Rashidi et al. (2016) include: the crude oil composition, the crude oil and the ambient temperature, flow conditions, pipeline geometry and the system pressure.

Previous Experimental Work

Wax deposition problem and possible mitigation has been studied extensively by several researchers; for example published studies have shown that in single-phase crude oil flow investigation of wax deposition has been carried out by Creek et al. (1999), Singh et al. (2000), Gao (2003) Lee (2008),

3

Hoffmann and Amundsen (2010), Dwivedi (2010), Zheng et al. (2013), Bagdat and Masoud (2015), Rashidi et al. (2016), Theyab and Diaz (2016), Chi et al. (2017). Similarly, in two-phase oil and water flow deposition of wax has been studied by Couto (2004), Anosike (2007), Zhang and Liu (2010), Huang et al. (2011), Panacharoensawad and Sarica (2013), Kasumu (2014), Kasumu and Mehrotra (2013), Wang et al. (2019). and Zheng (2017). Whereas in two-phase oil and gas flow, wax deposition study was investigated by Apte et al. (2001), Matzain et al. (2002), Kilincer (2003), Gong et al. (2011), Rittirong (2014), Rittirong et al. (2015, 2016, 2017), Sulaiman et al. (2017). As discovered in the published literatures, most of researches were carried out experimentally using a flow rig – with straight pipeline test section.

As of today, there are few researches that study wax deposition using a flow rig designed with a curved or a bend pipeline (Lee, 2008; Lee et al., 2008; Bagdat and Masoud, 2015; Rashidi et al., 2016) with no concern given to the study of restriction (bends and elbows) and their effect on wax deposition. Similarly, there are lack of experimental database and knowledge on the mechanisms that govern the development of wax deposition through a restricted flow. Therefore, this study is aimed to demonstrate the effects of bends (such as 45 and 90-degree bend) on wax deposition, in order to improve the existing dynamic models for accurate prediction of the wax crystals formation. However, in the other area of flow assurance such as hydrate and multi-phase flow problems, there are several explicit studies that broadly investigates the effects of restriction (e.g. bends and elbows) such as Wang et al. (2004), Xing et al. (2013), Zhao et al. (2017). For instance, study by Xing et al. (2013) incorporated different pipes bend (45 and 90-degree) in form of wave to experimentally study and mitigate severe slugging in three-phase (oil-water-air) flow. The study concluded that the pipe bend (in wavy form) reduces slug length in the pipeline/riser system. In addition, the study disclosed the working principle and the effects of the geometrical parameters and effective location of the wavy pipe in riser system.

Hence, in view of the above, this study is carried out to comparatively investigate the severity of wax deposition in straight pipe, as well as pipeline with 45 and 900 bend. The effort is to elucidate the fluid dynamic behaviour on wax deposition in straight and curved pipe and other major parameters that typically control the deposition of wax. Therefore, the study drives is towards improving the physical understanding of waxing process. It is understood that the presence of a bend causes flow restriction, and significantly affect global and local flow parameters such as pressure drop, advection and particle interaction (Wang et al., 2004; Yadav et al., 2014) and therefore, the physical mechanism of wax deposition. The major mechanisms reported for wax deposition as described by authors such as Bern et al. (1980), Burger et al. (1981), Majeed et al. (1990), Brown et al. (1993), Singh et al. (1999), Azevedo and Teixeira (2003) and Aiyejina et al. (2011) include; molecular diffusion, Brownian diffusion, shear dispersion, gravity settling and shear stripping. As of today, the actual relevance of all the mechanisms and their impact on flow characteristic and dynamics flow behaviour on wax deposition is still questionable. On the other hand, depending on the flow conditions and properties, some mechanisms usually oppose the deposit growth in the pipe. In a straight pipe where pressure drop due to the frictional forces dominates, authors such as Burger et al. (1981), Brown et al. (1993) and Singh et al. (2000) have shown that molecular diffusion is the only main mechanism responsible for the deposition of wax in a laminar boundary sublayer driven by the radial Fickian diffusion of n-alkanes. While other researchers including Bern et al. (1980) and Majeed et al. (1990) believe that Brownian diffusion, shear dispersion, and gravity settling play a pivotal role for specific deposition of wax. On the other hand, Singh et al. (2011) reported that several researchers (e.g. Bott and Gudmundsson, 1977; Venkatesan, 2004)

4

expressed that in highly turbulent flow, the shear stress acting on a wax gel layer may slough off the loosely held wax from the deposit layer, reducing the wax deposit.

Velocity and Pressure Drop Relationship for Wax Deposition Study

Fluid flow in straight pipe experiences no secondary circulation or flow separation at any of the pipe boundary (Murty and Thandaveswara, 2014). Figure 1 shows a simple schematic of straight pipe, as the flow developed, the velocity gradient at the wall reduces and gradually more of the core fluid become sheared. At the inlet of the pipe, the crude oil experiences higher pressure drop similar to the wall shear stress, which is higher and decreases steadily to the fully developed value. As waxy crude oil flow in a fully developed laminar flow, each particle moves in an orderly manner and at a constant axial velocity along the pathline. On the other hand, for a fully developed turbulent flow shown in Figure 1 as well, the fluid particles move in random and rapid fluctuations of swirling eddies throughout. These swirling eddies provides the wax crystals particle with an additional mechanism for mass, momentum and energy transfer that prevent the possible precipitation or deposition, unlike in the laminar flow condition, where the energy transfer is that rapid. In laminar flow, molecular diffusion is the main effect that enhance the mass, momentum and energy transfer across the region (UIO, 2004).

Vavg Vavg Vavg

Irrotational Inlet flow region

Developing velocity profile Laminar velocity

profile

umax

Ƭw Ƭw Ƭw

Turbulent velocity profile

Ƭw Vavg

μݑμݕ ௬ୀ�μݑതμݕ ௬ୀ�

οܲ��

Fully developed region

Entrance Region Flow Re ≤ 2300 Re > 2300

Figure 1– Schematic of straight pipeline with fully developed velocity profiles in laminar and turbulent flow regimes (adapted from UIO, 2004)

In wax deposition experimental studies, pressure drop is one of the key indicators for wax build-up in pipeline (Chen et al., 1997; Bai and Bai, 2005). This is because, as the wax deposit builds-up on pipe wall, the effective diameter of the pipe reduces leading to an increase in pressure drop. A correlation (Eq. 7) has been developed from the pressure drop equation to estimate wax deposit thickness. This is one of the most reliable technique used by several researchers to obtain the wax thickness (Hoffmann and Amundsen, 2010; Singh et al., 2011; Panacharoensawad and Sarica, 2013). Hence, pressure drop in straight pipe is as a result of frictional loss across the pipe section (Chen et al., 1997). As shown in Eq. 2, the pressure drop can be derived from the Darcy-Weisbach as follows:

∆ P f=4 f Ld

ρ2 ( 4Q

π d2 )2

(1)

where ∆ P f and L are the pressure drop and the length of pipe section, d is the hydraulic diameter or effective inside diameter, Q and ρ are the volumetric flow rate and the fluid density, while f is the Fanning friction factor. The above frictional pressure drop is modified as;

5

∆ P f=2cρLd5−n ( μ

ρ )n

( 4 Qπ )

2−n

(2)

This equation shows that any reduction of the effective inside diameter (d) of the pipeline causes an increase in frictional pressure drop across the pipe section. c and n are constant parameters; c = 16 and n = 1 for laminar flow (NRe < 2000), while for turbulent flow c = 0.046 and n = 0.2 (NRe > 2000) (Chen et al., 1997).

On the other hand, fluid flow in a bend pipe is characterised by the presence of a radial pressure gradient generated by the centrifugal forces acting on the fluid (Sreenivas, 2011). Similarly, the flow is primarily associated with flow instability, separation, and strong secondary current, which mainly depend on Reynolds number, bend angle () and the radius of curvature of the bend (Rb) (Dean, 1927; Sreenivas, 2011; Murty and Thandaveswara, 2014; Dutta et al., 2016; Vester et al., 2016). As shown in Figure 2, the crude oil experiences a steady flow without velocity variation within the straight section of the pipe just before the bend. The central portion of the fluid accelerated to continue the constant flow through any cross-section of the pipe. It is obvious that as crude oil flow through a bend pipe (even for a smaller bend curvature ratio, Rc/D≥ 0.5) the adverse pressure gradient near the inner wall and immediately downstream of the bend could lead to flow separation and increases large pressure losses (Idelchik, 1986; Dutta et al., 2016). In this study, two different curvature orientation (in horizontal and incline flow) were used to investigate the effect of pipe bends on wax deposition and to develop better understanding of the physical mechanisms responsible for the deposition of wax.

Therefore, the overall pressure drop in a bend pipe as modified from Eq. 1. is defined (Eq.3) as the sum of the frictional pressure drop created within the straight pipe and the pressure drop due to momentum effect caused by change in flow direction (which depends on the curvature ratio and bend angle) (Dean, 1927; Sreenivas, 2011; Prajapati et al., 2015).

∆ P= f2

ρ u2( π Rb

dθ

180 ° )⏞frictional pressureloss withbend

parameters

+ 12

kb ρu2⏞¿ Pressure lossdue

¿¿

momentum¿¿(3)

Where u is the mean flow velocity; Rb, the bend radius; θ, the bend angle; and kb, the bend loss coefficient obtained from bend loss coefficients Babcock & Wilcox chart. On the other hand, Eq. 3 was

A-A

Uniform flowNo Pulsation or

Velocity variation

Outer Wall flowSeparation/Back mixing

Inner Wall flowSeparation/Back mixing

Spiral Flow

c

d

a

b

Fully developed Velocity profile

Vavg

(a) (b)

Lb = θ

3602 π Rb

 

Lb = θπ Rb

18 0°

Figure 2–A typical velocity profile with a separated flow in pipe bend: (a) Longitudinal and rectangular cross-section around the 900 bend (Adapted from Idelchik, 1986), (b) Bend parameters

6

further modified to obtain overall pressure drop due to gravity. Therefore, the total pressure drop in an inclined pipe bends is expressed as the sum of the frictional pressure drop, the pressure drop due to the momentum and gravitational effect.

Depending on the orientation, Eq 2, and 3 are the basis in which wax deposit thickness is calculated. Generally, pressure drop correlation has been used to obtain wax thickness based of Eq. 5 that is modified by Chen et al. (1997). Whereas, in this study, equation 3 and 4 were further modified (Eq 6 and 7) to obtain the thickness of the deposit wax in a pipe bend.

Experimental Analysis

Crude Oil Characterization

The crude oil samples used in this experimental study was supplied by Roemex Oilfield Service Company – with unknown origins and properties. Hence, the characterization (as described in the subsequent section) was carried out to obtain the properties (shown in Table1) that defined the waxy nature of the sample through various laboratory analyses as brief descriptions in subsequent section.

Table 1–Properties of this crude oil sample and their standard methods of MeasurementProperties Unit Value Method

Appearance −¿ Brown Visual

Density kg/m3 (15℃) 835 Measurement

Sp. Gravity 60℉ /60℉ 0.850 Calculated

API Gravity #̊ API 35 API MethodWax Content wt% 19.7 Modified UOP 46-64Wax Content wt% 19.3 HTGC (n-C17+)Pour Point ℃ 25.5 RheometryPour Point ℃ 27 Modified ASTM D-97WAT at 120 1/s ℃ 30 RheometryWAT at 10 1/s ℃ 35 Rheometry

Viscosity Pa.s (at 40℃>WAT) 0.034 Rheometry

Viscosity Pa.s (at 15℃<WAT) 1.72 Rheometry

Saturates Fraction wt% 73.25Elution

chromatography

Aromatics Fraction wt% 21.2Elution

chromatography

Resins Fraction wt% 5.14Elution

chromatographyAsphaltene

Fraction wt% 0.41Modify ASTM D2007-

80

Rheological Analysis Rheological studies of crude oil were performed using Bohlin Gemini II shear controlled Rheometer to determine the viscosity, wax appearance temperature (WAT) and pour point temperature (PPT) (Singh et al., 2011; Adeyanju and Oyekunle, 2013; and Theyab and Diaz, 2016). A cone with a 4-degree angle and 40 mm diameter is used for the analysis, which has a gap setting of 0.15 mms (0150). The measurements were performed by controlling the cooling rate, shear and temperature. The oil sample is heated above the wax crystallization temperature – at this point, the sample exhibits Newtonian fluid characteristic (Roenningsen et al., 1991). The analysis begins immediately by cooling down the crude oil from 50℃ to -5℃ at a constant rate of 1.0 °C/min. Once the oil sample reached 0℃, the sample is held at a constant temperature for 15 minutes with no shear. The oil sample was analyzed at different

7

shear rate of 10, 60, and 120 1/s, similar to the study reported by Singh et al. (2011), Adeyanju and Oyekunle (2013) and Theyab and Diaz (2016).

SARA Analysis of Crude Oil

Saturates/Aromatics/Resins/Asphalthene (SARA) analysis separates/characterises the crude oil sample into four major fractions of saturates, aromatics, resins and asphaltenes (Jewell, Weber, Bunger, Plancher, & Latham, 1972) As a vital tool used by the oil industries, the analysis is effective for assessing crude's fouling propensity, blending compatibility and asphaltenic stability (Roenningsen et al., 1991; Fan et al., 2002).

A modified ASTM D2549-02 (elution chromatography) method was used in the study. Initially 7 g of crude oil was de-asphaltened with 160 ml of n-heptane according to the ASTM D2007-80. The maltene (filtrate) was first separated from solvent to possible amount using vacuum evaporation before introduce to the open column chromatographic set-up with glass frits. 15 ml of maltene was charged to the column using a pipette. The sequentially elution of each SARA fraction was done according to the studies by Jha et al. (2014). At the beginning, saturate fraction of the crude oil was eluted from the column with 100 ml of n-pentane. While, for elution of aromatic fraction, 100 ml of toluene is added and eluted aromatic is collected. The elution of polars/resin was done in four runs, each elution was collected separately. The sequence of elution began by mixing and adding 100 ml of toluene-methanol solution (50:50), secondly 100 ml of methanol-chloroform solution 50:50, third by using 100 ml of chloroform and finally 100 ml of acetonitrile is added. All the eluted compounds were collected in a separate container and separated from solvent using a rotary vacuum evaporator to recover the pure fraction of saturates, aromatic and resins.

Pour Point Temperature

The PPT is obtained from the viscometric plot, similar to the study by Adeyanju and Oyekunle (2013) and Theyab and Diaz (2016). In addition, a slightly modified ASTM Standard D97-08 method was used to compare the result. In the later technique, an appropriate amount of crude oil was filled in a test jar and preheated to about 50 - 60 ℃ for 30 min. The complete set-up is made of a bath with crushed ice; the test jar sealed with a cork, is placed in the bath. A thermometer is partially immersed in the sample crude through the sealed cork to monitor the temperature changes. The pour point was monitored and checked after every minute. The test Jar is removed and positioned horizontally, if the crude oil remained in that position for a period of 5 seconds without sagging, the temperature reading is recorded as the PPT of the sample. In addition, 3℃ is added to the thermometer reading. According to ASTM D-97 the as the actual pour point is 3℃ higher than the reading observed.

Wax Content and Crystals Morphology of Crude Oil

The wax content is obtained using HTGC as detailed by Singh et al. (2011) and compared by acetone precipitation technique, which is a modified UOP46-64 method as previously reported by researchers such as Hoffmann and Amundsen (2010), Coto et al. (2011) Fan and Buckley (2002), Sarica and Panacharoensawad (2012). In this study 5g of oil was dissolved in 35 cm3c solvent (petroleum ether). The mixture is stirred in beaker for about 10 – 15 min. Acetone is added to the mixture in a ratio of 3:1 vol/vol (acetone: petroleum ether) to precipitated the wax content in the oil. The mixture is then chilled to -20 ℃ for 24 hours. After this, the precipitated waxes are recovered by vacuum filtration using glass Buchner funnel with a vacuum pump connected to the side arm of the filtration flask. The filtered solid is washed with n-heptane to remove asphaltenes and the filtered liquid evaporated. Subsequently, the solid obtained after solvent evaporation contains the n-paraffin – is re-dissolved in n-hexane in order to remove asphaltenes content, and then filtered again. After solvent removal, the final product (wax) was weighted. On the other hand, the microscopic morphology of crystalline waxes molecules in the crude

8

oil was captured and analyze using a Carl Zeiss Axiovert S100 inverted optical microscope equipped with a Motic digital camera under static condition.

Analysis of n–Paraffin Distribution

An industrial standard technique through higher temperature gas chromatography (HTGC) was used to obtain the n-paraffin distribution of the oil sample. This was carried out in collaboration with the University of Plymouth using a proprietary method known as T-SEP® technique (thermal separation technique), which was co-developed by the University of Plymouth and KAT laboratory. The HTGC (Agilent 6890 gas chromatograph) is equipped with flame ionization detector (FID) and a Varian Vf–5ht Ultimetal column (15m x 0.25mm x 0.1μm) to analyse both ‘Whole’ and ‘Topped’ oil samples at 435℃. The oven is programmed to run from 40 to 435 ℃ at 10 ℃/min ramp before holding at 435℃ for 10 min. Helium was used as carrier gas at a flow rate of 5 mL/min. As a proprietary technique, “T-SEP®” allows precise control of hydrocarbon by “topping process”. The method is very reliable, reproducible and effective than the simple conventional techniques employed by most laboratories. However, the details of sample preparation of T-SEP® technique is not included in this paper.

Experimental Flow Rig DesignThe flow loop designed in this study is based on the work by Adeyanju and Oyekunle (2013) and Theyab and Diaz (2016) for the straight pipe, while, the bend pipe section is based on Xing et al. (2013) investigating hydrate formation mitigation and Rashidi et al. (2016). The three test sections (i.e. a straight pipeline and two pipelines with bends) are of equal length (L= 1000mm) with the same internal diameter of the pipe (d = 15mm). They are made of a copper pipe material, which has high thermal conductivity. The fluid flowing inside the pipe can be rapidly cool down to promote the formation of wax deposition in the 15mm straight pipe test section jacketed in a 25mm pipe of the same material. The crude oil flows in the inner pipe while the coolant (mixture of water and glycol of 50% by weight) flows counter-currently in the annulus. On the other hand, the test section with bends are built using standard elbowsjoined together by compression fittings. One of the test sections is shown in Figure 3 with the geometrical parameters including the bend radius bend (R = 60mm), angle of the bend (α=450) and velocity variation within the pipe.

Two of the pipeline test sections with bends were installed in an open tank of 50L capacity. The water in the tank was kept at a desired temperature by a submerged refrigerating cooling coil connected

R 60 mm

R 60 mm

45°

220 mm

220 mm

Straight fittings (x 3)

Buckhead (Tank Connector) (X 2)

Length 1000 mm)

Test Section

uo

α

Outer Wall FlowSeparation/Back mixing

45°

Steady Flow in FullyDeveloped Region

Outer Wall FlowSeparation/Back mixing

Inner Wall FlowSeparation/Back mixing

g

Outer Wall FlowSeparation/Back mixing

Figure 3 – Parameters of bend pipeline test section of three 45-degree bends.

9

with a high capacity chiller. A powerful stirrer operated at 500rpm that can mixes more than 100 Ltrs head capacity was used to keep a uniform water temperature. A three--ports glass flask as shown in Figure 4 serves as crude oil reservoir. The flask can withstand high temperature up to 200ºC. Two ports allow the crude oil to flow in and out of the reservoir, while the third one was used to condense the gaseous component evolving from the crude oil. The crude was circulated by a self-priming screw pump and the flowrate was controlled by an Optidrive variable-frequency drive, which controlled AC motor speed and torque by varying motor input frequency. GC35 pressure transducers are used to monitor pressure drops.

Experimental Methodology and test configurationsThe experimental study was carried out in a single-phase flow. The straight pipe test-section is horizontally positioned throughout the experiment, whereas, the two bend pipeline test sections were studied in horizontal and inclined positions. As shown above (in Figure 4), the flow loop was designed to be flexible, which allowed the test sections to runs independently and to be removed for sampling and analysis of the deposit thickness, mass and volume. Initially, the experiments were carried in horizontal straight pipeline test-section as the base case for comparison with the bend pipes test section of equal dimension with the straight pipe. The data from the initial results are compared with other studies that have similar flow rig configuration and other properties (such as Adeyanju and Oyekunle, 2013; Theyab and Diaz, 2016; Panacharoensawad and Sarica, 2013). On the other hand, there is no published data to directly compare the result of bend pipeline – in wax deposition study. Standard techniques established in the literature are used to quantify the wax deposit thickness in both test sections.

In course of the experiment, the crude oil entered the test section at higher temperature above WAT (15℃ + WAT). The varied parameters included: cooling temperature (15, 20, 25, 30 and 35℃) and flow rates (2, 3, 5, 7, 9 and 11 L/m). The testing procedure consists of the following steps: Start-up: - The crude oil in the reservoir was initially heated to a temperature above WAT (WAT + 15℃), and circulated in the flow loop for about 15 – 30 minute – without cooling. The coolant flowed from the chiller at relatively high temperature (above WAT) within the exchanger shell or in the open tank (for the curved pipes) to avoid possible thermal gradient in the radial direction. Wax deposition testing: -

CONDENSER

Cooling Jacket

TEST SECTION ONE

WATER BATH

GV

TEST SECTION TWO

T1 T2 T3 T4 P2P1

WATER CHILLEROIL RESERVOIR

SCREW PUMP

DATA LOGGER

COMPUTER

T1 T2 T3 T4 P2P1

Coolant out Coolant In

REFRIGERATING COOLING COIL MIXER

FREQUENCY INVERTER

Figure 4 –Facility designed for wax deposition study with two different test sections. FM: flow meter, P1 and P2: inlet and outlet pressure of the test section, T1 and T4: inlet and outlet oil temperature, T3 and T2: the inlet and outlet

coolant (water-glycol mixture) temperature.

10

based on the test matrix; the deposition begins by adjusting the parameters that influenced the formation of wax such as mentioned above. Deposit thickness analysis: - Once a specific test is completed, the deposit, mass, volume and thickness are measured through different methods.

Deposit thickness measurement Four methods were used to comparatively determine the thickness. These methods – which includes the weight correlation, volume correlation, pressure drop correlation, and the film-thickness gauge method, were reported by several researchers (Hoffmann and Amundsen, 2010; Theyab, 2017; Chi et al., 2017; and Rittirong et al., 2017) and are briefly describes below.

Pressure drop correlation: This is used as an online method without interrupting flow and is one of the most reliable method for a single-phase oil flow (Hoffmann and Amundsen, 2010). Once the flow rate, density and viscosity of the crude oil in the test section are determined, the thickness can be calculated. The correlation was developed from Eq. 2 by Chen et al. (1997), and was used by several researchers (Hoffmann and Amundsen, 2010; Singh et al., 2011; Panacharoensawad and Sarica, 2013; Rittirong et al., 2017; and Adeyanju and Oyekunle, 2013) as follows.

(d i−2δw)5−n=2cρL∆ P f

( μρ )

n

( 4 Qπ )

2−n

(4)

Where δ wis the average thickness of the wax deposit. To incorporate pipe bend parameter, a new correlation was developed from Eq. 3 as follows:

Note: Darcy frictional factor ≫ fannig; and for laminar flow f =64 ρQπμd

As the wax deposits, the pipe diameter decreased, which increases the fluid velocity. Therefore, in the above equation the u and d will change when wax deposited. The remaining parameter are assumed to remain constant.

if Qwith out wax=Qwithwax

∴uA=uo Ao

u π d2

4=u

o

π do2

4

Wax thickness was calculated using a numerical method (e.g. Newton–Raphson method). Similarly, for turbulent flow, the expression for the frictional factor (f ) was changed.

Weight–balance correlation: The weight of the deposit is carefully visualised and measured after deposition in the pipe. The technique requires pigging and melting of the deposit wax from the test section after each experiment (pigging operation shown in Error: Reference source not found). Previous studies by Hoffmann and Amundsen (2010); Panacharoensawad and Sarica (2013) and Theyab and Diaz (2016) shows that the thickness obtained by weight correlation (Eq. 8) and pressure drop technique

∆ P f= ρu2( 128 ρQμd2

Rb θ180 °

+ 12

kb) (5)

∆ P f=ρuo2( do

d o−2 δ w)

4

( 128 ρQμ(do−2 δw )2

Rb θ180 °

+ 12

kb) (6)

(do−2 δw)4=ρ uo

2 do4

∆ Pf ( 128 ρQμ(do−2 δ w)

2

Rb θ180°

+12

kb) (7)

PiggingOperation

Wax Deposit

11

produced approximately the same result. Panacharoensawad and Sarica (2013) publish the details of this method.

δ w=R−√R2−M dep

L π ρdep−δ 0 (8)

Wet film-thickness gauge: This technique was effectively been used and reported by Panacharoensawad and Sarica (2013) and Rittirong et al. (2017). The local thickness is measured (Figure 5) by carefully inserting the Wet film gauge (attached with a long handle) at three different location of the pipe section: inlet, outlet and middle. The average of the thickness is taken and are compared with other methods. It should be noted that for the weight and wet film thickness method, the test must be stopped at each time point to analyze the thickness.

Results and Discussion

Crude Oil Rheological Properties

The results from rheological study showed that the viscosity of crude oil varies with temperature and shear rate (10, 60 and 120 1/s). The precipitation of the wax molecules caused changes in the crude oil behaviour from Newtonian to non-Newtonian (as shown in Figure 6). The two major parameters in wax

After experiment

Before experiment

Figure 5- Photo: Left- Wet-film-thickness gauge (scaled 50 – 1200 μm) for deposit thickness measurement. Right- Sample of wax pigging inside the spool piece

12

deposition study (i.e. WAT and pour point) are approximated similar to the study by Singh et al. (2011); Adeyanju and Oyekunle (2013), Paso (2014), Ronningsen et al. (1991), Theyab and Diaz (2016). The WAT was measured instantaneously as the wax crystal precipitated – this defined the beginning of paraffin wax crystallization. As in the Figure 6, WAT (30℃) represented the intersection between the baseline and the tangent fitted from the inflection point. Whereas, the intersection between two tangents drawn as inflection point below WAT defined the pour point temperature (25.5℃). The PPT was compared with the slightly modified ASTM Standard D97-08 method used in this study. The WAT and PPT obtained were used as design parameters to calibrate the flow loop, where the crude oil reservoir temperature was kept constant, between 45 – 50℃ (above WAT) in order to avoid wax crystallization.

n-Paraffin Distribution of Crude Oil and Morphology of the Wax Deposits

The HTGC analysis of the whole oil resulted in a series of resolved and unresolved complex mixtures. Therefore, the unresolved mixtures were merged based on T-Sep technique as shown in Figure7. This produced adjustable values that represented the actual and accurate carbon distribution of the sample. The results showed that the crude oil is highly paraffinic in nature with carbon distribution ranges from n-C15–C67. It was found that the average weight fraction of n-paraffin between n-C15–C35 is 0.90 wt% while n-C35–C67 is 0.0061 wt. In addition, through HTGC analysis, the wax content of 19.3 wt% is obtained, by summing-up the percentage weight of all n-paraffin fraction in the sample. Coto et al. (2011) employed the same method. From the result, it is evidenced that the sample required mitigation plan during transportation, production or storage. The wax content of 19.7 wt% obtained through modified UOP46-64 method was compared with the HTGC method. The two methods produced almost the same results as shown in Table 1.

On the other hand, the crude oil morphology, show a strong interlocking and interactions between the wax crystals that normally form gelled network that complicated problem in oil and gas pipeline. The structural morphology defined as needle-like or rod-like particles are clearly observed in the images (Figure 8). Similar behaviour was reported by other researchers (such as Coto et al., 2014) using different waxy crude oil.

Figure 7-n-Paraffin distribution of the waxy crude oil samples using T-SEP® technique

13

Pressure and Temperature Response in Flow Loop

In wax deposition study, the pressure and temperature sensor responses are paramount. As shown below, the pressure responses are visualized as a voltage to time (trend) plot. Figure 9 shows a correlation between pressure drop signal and temperature of water and crude oil after time t = 26 min of experiment. As shown in the Figure, ΔP t=0 shows a no flow condition in the pipeline (i.e. no pressure difference between channels A and B). Whereas, the observed initial increase in pressure drop, ΔP t at t = 26 min is constant for this period and there is no increase in pressure drop observed. Nevertheless, it is expected that as soon as the wax build-up in pipeline, the pressure drop will increase. This was monitored throughout the experiment and are used for the calculation of the wax thickness for comparative analysis. On the other hand, under the same condition (t = 26 min), the temperature responses for inlet and outlet water and crude oil temperature is constant (as shown below) – 15 and 45℃ respectively (without deposition). However, as the wax precipitated due to the temperature gradient between pipe wall and the flowing fluid, it travelled along with the crude oil and ultimately reached the outlet thermocouple that causes drop in the observed temperature readings. When the wax molecules simultaneously builds up and deposit on the pipe wall – controlled by different mechanisms, the decrease in the outlet temperature rapidly increased and remained constant until the formation is interrupted by putting strategy that increases the crude oil temperature.

Figure 9-Initial pressure and temperature correlation at oil flow rate of Qoil = 7 l/min

Wax Deposition Analysis in Straight Pipe Initially, the effect of two fundamental parameters that control wax deposition are investigated in straight pipeline. The results showed that out of the four methods employed for wax deposition thickness quantification, two of the methods (i.e. pressure drop and weight correlation) produced nearly equal results (Figure 10 and Figure 11). In this study, the straight pipe experiment are conducted and used as the base case for a comprehensive comparison of data obtained in bend pipes. Owing to the fact that

Figure 8- Macroscopic morphology of crude oil sample A, B and C with 0-ppm and 1000-ppm concentration of blend

14

there are no published data for comparison of the bend pipes results. Therefore, the result obtained from the straight pipe test-section experiment are compared with other studies that have similar flow rig configuration, crude oil composition, method of quantifying wax thickness and experimental matrix, which includes; Adeyanju and Oyekunle (2013), Panacharoensawad and Sarica (2013), Theyab and Diaz (2016) and Rittirong et al. (2017) – these study are based on the laboratory scale level. Also, standard offshore case study by Singh et al. (2011) was looked at for understanding and establishing realistic data.

Effect of Thermal Gradient

Effect of thermal gradient was investigated by varying the pipe wall temperature. The study was carried out over two different flow rates (5 and 9 l/m) – that are within the laminar and turbulent flow region. As shown in Figure 10, the experiments were run with constant reservoir temperature (WAT + 15℃) and constant aging period (2 hours) over a range of different cooling temperatures (15, 20, 25, 30, 35 and 40℃). These conditions provided opportunity to study cooling conditions frequently encountered in subsea environment. According to the literature, at below the wax appearance temperature (WAT), crude oil experiences a gel transition at which the non-Newtonian behaviour appeared to be dominant characteristic, whereas, below PP the precipitated wax molecule displayed the rheological properties of a viscoelastic solid (Visintin, Lapasin, Vignati, D’antona, & Lockhart, 2005). Therefore, from Figure 10 and Figure 11, the results clearly showed that wax deposition thicknesses are lower at higher cooling temperature (e.g. at 30℃ and 5 l/m, δ wax ≈ 0.4mm¿ and particularly above the WAT, it was observed that in both laminar (5 l/m) and turbulent (9 l/m) flow there was no wax deposition (i.e. at 35 and 40℃,δwax ≈ 0 mm). This is because if the pipe wall temperature is maintained above WAT, there will be no precipitation of wax molecules. Meaning that the solubility level of the wax molecules at this point is high. Similarly, at higher flow rate (Qoil=¿9 l/min) and cooling at exactly equal to WAT, there is no wax deposition observed (δ wax ≈ 0 mm) as shown in Figure 10. However, small wax thickness (δwax ≈ 0.35 mm) was seen at the same temperature but lower flow rate (Qoil=¿5 l/min) as shown in Figure 10. This could be due to the effect of molecular diffusion mechanism of dissolved wax or the combined molecular diffusion and shear dispersion effect, as also proposed by Bern et al. (1980).

At higher temperatures (e.g. 35 and 40℃) and high heat flux conditions, molecular diffusion is expected to dominate the deposition process, whereas, at lower temperatures and low heat fluxes shear dispersion will dominate (Bern et al., 1980). In shear dispersion mechanism, the literal movement of wax molecules close to the pipe wall leads to the transport of more precipitated wax from the turbulent core to the surface of the pipe, which could resulted to the formation of deposit on the pipe wall or connect with wax already deposited by molecular diffusion (Bern, Withers, Cairns, et al., 1980). Whilst an increase in shear rate usually encourage more wax crystals to diffuse towards the pipe wall, whereas, the corresponding increase in wall shear stress causes more loosely held deposits to be slough-off/stripped from the pipe wall (Venkatesan, 2004; Bott and Gudmundsson, 1977). However, in straight pipe, shear dispersion is less probable mechanism as there is likely no propensity for nucleation to occur than those formed by the molecular diffusion process (Bern et al., 1980) compared to those in bend pipe –discussed in subsequent section.

15 20 25 30 35 400.0

0.5

1.0

1.5

2.0

2.5

Crude Oil Flow Rate = 9 l/m

Dep

osit

Thic

knes

s,

(m

m)

Cooling Temperature (oC)

Guage Method

Pressure Drop

Weight Correlation

Volume Correlation

15 20 25 30 35 400.0

0.5

1.0

1.5

2.0

2.5

Crude Oil Flow Rate = 5 l/m

Guage Method Weight Correlation Volume Correlation Pressure Drop

Dep

osit

Thic

knes

s,

(mm

)

Cooling Temperature (oC)

15

Therefore, as shown in Figure 10, the deposition rate increased with decrease in cooling temperature (pipe wall temperature) and reaches maximum as the temperature near the usual seabed condition (i.e. below the PP of the sample). The decrease in the pipe wall temperature causes the crude oil temperature to drops; this is due to the decrease in both the diffusion coefficient and the radial temperature gradient. Therefore, at the same temperature and under low flow rate the deposit wax blocked the pipeline. A sample of the cross sectional view of the deposit in pipe is shown in Figure 15. Similar result are reported by Adeyanju and Oyekunle (2013) and Theyab and Diaz (2016).

Effect of Crude Oil Flow rate

The effect of different crude oil flow rates (2, 3, 4, 5, 7, 9, 11 l/min) was investigated over two varied pipe wall temperature (15 and 30℃), at constant oil temperature and aging time (2 hours). The flow rates cover the two dynamics flow regimes, i.e. laminar (2, 3, 4 and 5 l/min) and turbulent flow (7, 9 and 11 l/min). This is to fully comprehend the effect of low velocity, high and low momentum diffusion and convection, swirling and lateral mixing in straight pipe.

As shown in Figure 11, wax deposition rate is higher even at high flow rate as the pipe wall temperature is kept at 15℃< PP temperature. This is due to the higher concentration of wax molecules available – which diffuses towards the pipe wall. In this scenario, it is believed that both molecular diffusion and shear dispersion contributed to the deposition process. This implies when temperature of oil and coolant are below the WAT at higher flow rate, shear dispersion mechanism becomes significant to wax deposition. However, as we kept the pipe wall at high temperature, the system experiences very small deposition rate as shown in Figure 11 (cooling at 30℃). This shows that wax precipitation does not necessarily means or lead to the deposition. Therefore, as the pipe wall temperature equals to the wax appearance temperature (WAT) wax precipitation occurred (as showed in Fig. 8). In this experiment, it was observed that at turbulent flow condition (i.e. Qoil=¿9 and 11 l/min) the measured wax thicknessδwax ≈ 0mm. In view of the fact that as the wax crystals precipitated at high flow rate, there is an increase in shear stress between the liquid-deposit interface, which reinforced the intensity of the shear stripping that as a result led to no deposition or less deposition as in the case of Qoil = 7 l/min, δwax ≈ 0.25 mm. In both cases, as shown below the deposit intensity decreased at high flow rate. According to literature, in the laminar flow region, the internal heat transfer coefficient increases with the increase in the oil flow, which increases the radial temperature gradient and consequently, led to an increase in the thickness of wax deposit (Adeyanju and Oyekunle, 2013; Creek et al., 1999), whereas, due to the shear mechanism contrary trend was seen during the turbulent flow.

1 2 3 4 5 6 7 8 9 10 11 120.0

0.3

0.6

0.9

1.2

1.5

1.8

2.1

2.4

Cooling Temperature = 30oC

Dep

osit

Thic

knes

s,

(m

m)

Flow Rate (L/Min)

Gauge Method Pressure Drop Correlation Weight Correlation Volume Correlation

1 2 3 4 5 6 7 8 9 10 11 120.0

0.3

0.6

0.9

1.2

1.5

1.8

2.1

2.4

Cooling Temperature = 15oC

Dep

osit

Thic

knes

s,

(m

m)

Flow Rate (L/Min)

Gauge Method Pressure Drop Correlation Weight Correlation Volume Correlation

Cooling Temperature = 15oC

16

Effect of Bend Pipeline on Wax DepositionIt has been demonstrated in the previous section that wax deposition in horizontal straight pipe

section under the influence of cooling temperature and flowrate are is highly controlled by mechanisms such as molecular diffusion and shear dispersion, whereas shear stripping opposed the deposition process. The remaining two mechanisms i.e. Brownian diffusion and gravity settling are not consider to be significant in the straight pipe based on this experimental conditions. In a gravity settling mechanism, it is expected that the precipitated molecules, which are denser than the surrounding fluid to settle on the pipe wall. Although, in an experimental study by Burger et al. (1981), the gravity settling mechanism was tested by switching a straight pipe test section from horizontal to vertical flow, with no effect of gravity settling is observed. In addition, it was reported through mathematical study that shear dispersion might redisperse the settling of solids wax on the pipe wall either in vertical or horizontal straight pipe (Burger et al., 1981; Kok and Saracoglu, 2000) thereby making gravity settling less significant. On the other hand, the Bronian diffusion mechanism is said to occur as a result of lateral transport by diffusion, however, Burger et al. (1981) and Kok and Saracoglu (2000) reported that two other mechanisms (i.e. molecular diffusion and shear dispersion) also occurred as a result of lateral transport. Therefore, their studies showed that the influence of Brownian diffusion is less compared with the other mechanisms. In another development, Zhu et al. (2008), suggested that “in a laminar flow regime, the shear removal, Brownian dispersion, and gravity settling can be safely ignored with the molecular diffusion the dominate mechanism for wax deposition”.

In this study, the effect of thermal gradient and crude oil flow rate are investigated using the two different bends pipeline test section – in horizontal and incline flow. The crude oil flow rates were varied that covered both laminar flow regimes (i.e. at Qoil= 5 and 7 l/min) and turbulent flow regimes (at Qoil= 9, and 11 l/min) over different range of cooling temperature (15, 25 and 30℃). It is expected, which is also proved (in the straight pipe) that above WAT (i.e. cooling the pipe wall above 30℃) there would not be deposited and at or below WAT and PP the system would experience serious wax deposition rate. Figure 13 shows a comparisons of wax deposition rate (in terms of wax thickness) in horizontal and inclined flow, using a 450 bend pipeline test-section. Clearly, it is observed that the deposition rate is higher, particularly, at condition that was not expected to be, that is at relatively higher cooling temperature (30℃ = WAT) and high flow rate (Qoil= 11 l/min). The thickness measured is, δ wax ≈ 0.2mm in horizontal flow and δwax ≈ 0.45 mm in an inclined flow. Whereas, in straight pipe experiment under same conditions, the deposit thickness δ wax ≈ 0 mm.Therefore, this behaviour could be due to the effect bend in the pipe, where flow segregation and isolation occurred immediately after the bend. In addition, it is believed that the effect of the wall shear rate, the amount of wax precipitated out of the oil and the shape/size of the wax particles may likely be higher in a bend pipe. The isolated crude oil around the bend could easily experience more thermal difference compared to the bulk fluid, which could results in the rapid molecular diffusion process under higher temperature, and a high shear dispersion of particles at lower temperature (Burger et al., 1981; Kok and Saracoglu, 2000). At lower crude oil flow rate (Qoil= 5 l/min) in horizontal flow produced the maximum thickness¿) at 15℃ cooling temperature. Whereas, at the same temperature but higher flow rate, Qoil= 11 l/min the thickness measured is δwax ≈1.4 mm.

Figure 11-Effect of varied crude oil flow rate on: Left – Wax deposit thickness cooling at 15℃. Right – Wax deposit thickness cooling at 30℃

17

Similar deposition behaviour was observed using 900 bends pipeline test-section – in horizontal and inclined flow as shown in Figure 13. Where in the horizontal orientation, thickness of δwax ≈3.5∧1.5mmwere measured at lower (Qoil= 5 l/min) and higher (Qoil= 11 l/min) flow rate under the same cooling temperature of 15℃. In horizontal flow, higher wax thickness was observed using 900 bends pipeline test-section compared to the 450 bends pipeline. This could be related to the effect of bend parameters. According to the literature, the higher the bend radius, the more flow is separated and the more isolated fluid would be around the bend. Consequently, this resulted in more dominated molecular diffusion and shear dispersion mechanisms.

Figure 12-Effect of varied cooling temperature under different flow rates on wax thickness in 450 bend pipe test-section with: Left – Horizontal flow experiment. Right – Inclined flow experiment

4 5 6 7 8 9 10 110.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Dep

osit

Thic

knes

s,

(m

m)

Flow Rate (L/min)

15oC Cooling Temp 25oC Cooling Temp 30 oC Cooling Temp

45-degree curved pipe

5 6 7 8 9 10 110.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Dep

osit

Thic

knes

s,

(m

m)

Flow Rate (L/min)

45-degree curved pipe

15oC Cooling Temp 25oC Cooling Temp 30 oC Cooling Temp

Inclined FlowHorizontal Flow

Figure 13- Effect of varied cooling temperature under different flow on wax thickness in 900 bend pipe test-section: Left – Horizontal experiment. Right – Inclined experiment

5 6 7 8 9 10 110.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

90-degree curved pipe

Dep

osit

Thic

knes

s,

(m

m)

Flow Rate (L/min)

15OC Cooling Temp 25OC Cooling Temp 30OC Cooling Temp

5 6 7 8 9 10 110.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Dep

osit

Thic

knes

s,

(mm

)

Flow Rate (L/min)

90-degree curved pipe

15OC Cooling Temp 25OC Cooling Temp 30OC Cooling Temp

Inclined FlowHorizontal Flow

18

On the other hand, the severity of wax deposition was observed to increase as the crude oil flow through an inclined pipeline test section, especially under very low pipe wall temperature as shown in both cases. As the pipe orientation is switched from horizontal to an inclined flow, an increase in the wax deposition rate was observed in both the two test section.

In Figure 12 it was clearly seen that wax deposition thickness increases from δ wax ≈3.1 mm (in horizontal flow) to δwax ≈ 4.2 mm under inclined flow at the same condition of cooling (15℃) and flow rate of Qoil= 5 l/min for 2 hours aging period. Similarly, in Figure 13, the deposit thickness increased from δwax ≈3.5 mm in horizontal flow, to δ wax ≈3.9 mm under inclined flow at the same conditions above. These results shows that apart from the molecular diffusion and shear dispersion mechanisms, the gravity settling effect is possibly felt in these two cases. Unlike the incident reported by Burger et al. (1981), where no change in the deposition rate was observed when a straight pipe test section was switched from horizontal to vertical flow.

The percentage severity of wax depositional rate in horizontal flow was compared for the three-curvature pipes (0, 45 and 90-degree bend) as shown in Figure 14. This study found that in laminar flow regime of lower flower rate (5 l/min) and at constant cooling temperature of 15℃, the percentage wax deposition in straight pipe (0-degree bend) is 22% compared to 36% in 450 horizontal bend pipe and 42.5% in 900 horizontal bend pipe respectively. This means that at these conditions, 14 and 19% more wax deposition problems would inherit if a curvature of 45 and 900 bend were used instead of a straight pipe. Similarly, at the same temperature but at higher flowrate of 11 l/min, the severity of wax deposition rate decreased. It is seen that only 8 and 10% more wax deposition issues would encountered if 45 and 900 bends pipeline are used instead of a straight pipeline.

Figure 14-Illustration of percentage severity of wax deposition rate at constant cooling temperature and varied flow rate (5, 7, 9 and 11 l/min) for the three test section in horizontal flow

CA B

19

Conclusion The characterized crude oil was found to be highly paraffinic in nature that required standard

management strategy in order to avoid precipitation that could lead to deposition. As expected based on the crude oil properties, related wax deposition problems were observed.

In straight pipeline test section, the high molecular weight crude oil precipitated wax molecules and deposited under the influence of molecular diffusion mechanism at the higher temperature and high heat flux, whereas shear dispersion or combination of both mechanism occur at the lower temperatures and low heat fluxes.

Contrary to the theory established in some literature, gravity settling mechanism was neglected due to possible re-dispersal of the particles by shear dispersion. However, it was observed that the effect of this mechanism ‘gravity settling’ is felt as the bend pipes switched from horizontal to an inclined flow. However, it is obvious that apart from the molecular diffusion, shear dispersion has a hug impact to the wax deposit measured.

The severity of wax depositional rate could be increase from 0% to 14% and 19% respectively if the curvature of the pipe changes from straight pipe to 45 and 900 bend pipes at low temperature and low flow rate. However, if the curvature is inevitable, higher flow at low temperature could reduce the severity.

This study clearly demonstrates that more work is needed on wax deposition to enhance the understanding on deposition mechanism so that a robust mitigation strategy can be proposed.

AcknowledgementsThe authors would like to express sincere appreciation to the Nigerian government through the Petroleum Technology Development Fund for providing the funding of this research and Romex oil company for the supply of crude oil.

Nomenclatured diameter of the pipelinec and n constant parameters f frictional factor∆ P f pressure drop L length of pipe sectionQ volumetric flow rate ρ fluid densityδw average thickness of the wax depositM dep Mass of wax depositρdep Density of wax deposit

High deposit due to low current flow

Low wax deposit due to high current flow

Empty pipe after pigging operation

Sample of deposit in straight pipe

Figure 15-Cross sectional view of the deposit on the pipe wall circumference: A – wax deposit after 450 bend in inclined flow, B – wax deposit at 2l/min and 10℃ and C – Empty pipe

20

ReferencesAdeyanju, O., & Oyekunle, L. (2013). Experimental Study of Wax Deposition in a Single Phase Sub-cooled Oil Pipelines.

Nigeria Annual International Conference and Exhibition, 167515.Adeyemi, B. J., & Sulaimon, A. A. (2012). Predicting Wax Formation Using Artificial Neural Network. In Nigeria Annual

International Conference and Exhibition. Society of Petroleum Engineers. https://doi.org/10.2118/163026-MSAiyejina, A., Chakrabarti, D. P., Pilgrim, A., & Sastry, M. K. S. (2011). Wax formation in oil pipelines: A critical review.

International Journal of Multiphase Flow, 37(7), 671–694. https://doi.org/10.1016/j.ijmultiphaseflow.2011.02.007Anosike, C. F. (2007). Effect of Flow Patterns on Oil-Water Flow Paraffin Deposition in Horizontal Pipes . University of

Tulsa.Apte, M. S., Matzain, A., Zhang, H.-Q., Volk, M., Brill, J. P., & Creek, J. L. (2001). Investigation of Paraffin Deposition

During Multiphase Flow in Pipelines and Wellbores—Part 2: Modeling. Journal of Energy Resources Technology, 123(2), 150. https://doi.org/10.1115/1.1369359

Azevedo, L. F. A., & Teixeira, A. M. (2003). A Critical Review of the Modeling of Wax Deposition Mechanisms. Petroleum Science and Technology, 21(3–4), 393–408. https://doi.org/10.1081/LFT-120018528

Bagdat, M., & Masoud, R. (2015). Control of Paraffin Deposition in Production Operation by Using Ethylene–TetraFluoroEthylene (ETFE). In Proceedings of the International Conference on Integrated Petroleum Engineering and Geosciences Icipeg 2014 (pp. 13–22). https://doi.org/10.1007/978-981-287-368-2

Bai, Y., & Bai, Q. (2005). Wax and Asphaltenes. Subsea Pipelines and Risers (First edit). Great Britain: ELSEVIER. https://doi.org/10.1016/B978-008044566-3.50023-3

Bern, P. A., Withers, V. R., & Cairns, R. J. R. (1980). Wax Deposition in Crude Oil Pipelines. In European Offshore Technology Conference and Exhibition. Society of Petroleum Engineers. https://doi.org/10.2118/206-1980-MS

Bern, P. A., Withers, V. R., Cairns, R. J. R., & Petroleum, B. (1980). EUR 206 WAX DEPOSITION IN CRUDE OIL PIPELINES EUROPEAn OFFSHORE PETROLEUM conFEREnCE &amp;EXHIBITiOn. In European Offshore Technology Conference and Exhibition, 21-24 October, London, United Kingdom.

Bott, T., & Gudmundsson, J. (1977). Depostion of Parafin Wax from Flowing Systems. Institute of Petroleum Technical Paper, IP 77-007.

Brown, T. S., Niesen, V. G., & Erickson, D. D. (1993). Measurement and Prediction of the Kinetics of Paraffin Deposition. In SPE Annual Technical Conference and Exhibition. Society of Petroleum Engineers. https://doi.org/10.2118/26548-MS

Buller, A., Funchs, S., & Klemp, S. (2002). Flow assurance. Statoil’s Research & Technology Memoir No. 1.Burger, E. D., Perkins, T. K., & Striegler, J. H. (1981). Studies of Wax Deposition in the Trans Alaska Pipeline. Journal of

Petroleum Technology, 33(June), 1075–1086. https://doi.org/10.2118/8788-PACawkwell, M. G., & Charles, M. E. (1989). Characterization of Canadian arctic thixotropic gelled crude oils utilizing an

eight-parameter model. J. Pipelines, 7, pp.251-264.Chala, G. T., Sulaiman, S. A., & Japper-Jaafar, A. (2018). Flow start-up and transportation of waxy crude oil in pipelines-A

review. Journal of Non-Newtonian Fluid Mechanics, 251(February 2017), 69–87. https://doi.org/10.1016/j.jnnfm.2017.11.008

Chen, X. T., Butler, T., Volk, M., & Brill, J. P. (1997). Techniques for Measuring Wax Thickness During Single and Multiphase Flow. In SPE Annual Technical Conference and Exhibition. Society of Petroleum Engineers. https://doi.org/10.2118/38773-MS

Chi, Y., Daraboina, N., & Sarica, C. (2017). Effect of the Flow Field on the Wax Deposition and Performance of Wax Inhibitors: Cold Finger and Flow Loop Testing. Energy and Fuels, 31(5), 4915–4924. https://doi.org/10.1021/acs.energyfuels.7b00253

Coto, B., Coutinho, J. A. P., Martos, C., Robustillo, M. D., Espada, J. J., & Peña, J. L. (2011). Assessment and improvement of n -paraffin distribution obtained by HTGC to predict accurately crude oil cold properties. Energy and Fuels, 25(3), 1153–1160. https://doi.org/10.1021/ef101642g

Couto, G. . (2004). Investigation of Two-Phase Oil-Water Paraffin Deposition. University of Tulsa.Creek, J. ., Lund, H. J., Brill, J. P., & Volk, M. (1999). Wax deposition in single phase flow. Fluid Phase Equilibria, 158–

160, 801–811. https://doi.org/10.1016/S0378-3812(99)00106-5Dean, W. R. (1927). XVI. Note on the motion of fluid in a curved pipe. The London, Edinburgh, and Dublin Philosophical

Magazine and Journal of Science, 4(20), 208–223. https://doi.org/10.1080/14786440708564324Dutta, P., Saha, S. K., Nandi, N., & Pal, N. (2016). Engineering Science and Technology , an International Journal Numerical

study on flow separation in 90 ° pipe bend under high Reynolds number by k-ε modelling. Engineering Science and Technology, an International Journal, 19(2), 904–910. https://doi.org/10.1016/j.jestch.2015.12.005

Dwivedi, P. (2010). An investigation of single-phase wax deposition characteristics of South Pelto oil under turbulent flow. University of Tulsa.

Fagin, K. M. (1945). Automatic Scrapers Used in West Edmond Oil Wells. Pet. Eng, 105.Fan, T., & Buckley, J. S. (2002). Rapid and accurate SARA analysis of medium gravity crude oils. Energy and Fuels, 16(6),

1571–1575. https://doi.org/10.1021/ef0201228

21

Fan, T., Wang, J., & Buckley, J. S. (2002). Evaluating Crude Oils by SARA Analysis. Society of Petroleum Engineers Inc. SPE 75228.

Ford, P. ., Ells, J. ., & Russell, R. . (1965). Pipelining high-pour-point crude. Pt. 3. Frequent pigging helps move waxy crude below its pour point. Oil Gas J.;(United States), 63(19).

Gao, C. (2003). Investigation of long term paraffin deposition behavior for South Pelto oil. Thesis.Goldman, Marcene S., and C. C. N. (1957). Prevention of paraffin deposition and plugging. U.S. Patent, No. 2(817,635).Gong, J., Zhang, Y., Liao, L., Duan, J., Wang, P., & Zhou, J. (2011). Wax Deposition in the Oil/Gas Two-Phase Flow for a

Horizontal Pipe. Energy & Fuels, 25(4), 1624–1632. https://doi.org/10.1021/ef101682uGuan, H. (2016). Flow Assurance Risk Evaluation and Chemical Management in a Shale Gas Field (pp. 1–14). NACE

International, Publications Division, 15835.Hammami, A., Ratulowski, J., & Coutinho, J. A. P. (2003). Petroleum Science and Technology Cloud Points: Can We

Measure or Model Them? https://doi.org/10.1081/LFT-120018524Hoffmann, R., & Amundsen, L. (2010). Single-phase wax deposition experiments. Energy and Fuels, 24(2), 1069–1080.

https://doi.org/10.1021/ef900920xHuang, Z., Senra, M., Kapoor, R., & Fogler, H. S. (2011). Wax deposition modeling of oil/water stratified channel flow.

AIChE Journal, 57(4), 841–851. https://doi.org/10.1002/aic.12307Idelchik, I. E. (1986). Handbook of hydraulic resistance (2nd revised and enlarged edition). Washington, DC, Hemisphere

Publishing Corp., 1986, 662 P. Translation. Retrieved from http://adsabs.harvard.edu/abs/1986wdch.book.....IJewell, D. M., Weber, J. H., Bunger, J. W., Plancher, H., & Latham, D. R. (1972). Ion-exchange, coordination, and

adsorption chromatographic separation of heavy-end petroleum distillates. Analytical Chemistry, 44(8), 1391–1395. https://doi.org/10.1021/ac60316a003

Jha, N. K., Jamal, M. S., Singh, D., & Prasad, U. S. (2014). Characterization of crude oil of upper assam field for flow assurance. Society of Petroleum Engineers - SPE Saudi Arabia Section Technical Symposium and Exhibition, 1–9. https://doi.org/10.2118/172226-MS

Kasumu, A. . (2014). An Investigation of Solids Deposition from Two-Phase Wax–Solvent–Water Mixtures.Kasumu, A. S., & Mehrotra, A. K. (2013). Solids Deposition from Two-Phase Wax–Solvent–Water “Waxy” Mixtures under

Turbulent Flow. Energy & Fuels, 27(4), 1914–1925. https://doi.org/10.1021/ef301897dKilincer, N. (2003). Multiphase paraffin deposition behavior of a Garden Banks condensate. Tulsa University, Tulsa, OK.Kok, M. V., & Saracoglu, R. O. (2000). Mathematical Modelling of Wax Deposition in Crude Oil Pipelines (Comparative

Study). Petroleum Science and Technology, 18(9–10), 1121–1145. https://doi.org/10.1080/10916460008949895Lee, H. S. (2008). COMPUTATIONAL AND RHEOLOGICAL STUDY OF WAX DEPOSITION.Lee, H. S., Singh, P., Thomason, W. H., & Fogler, H. S. (2008). Waxy Oil Gel Breaking Mechanisms: Adhesive versus

Cohesive Failure. Energy & Fuels, 22(1), 480–487. https://doi.org/10.1021/ef700212vLira-Galeana, C., Firoozabadi, A., & Prausnitz, J. M. (1996). Thermodynamics of wax precipitation in petroleum mixtures.

AIChE Journal, 42(1), 239–248. https://doi.org/10.1002/aic.690420120Majeed, A., Bringedal, B., & Overa, S. (1990). Model Calculates Wax Deposition for N. Sea Oils. Oil & Gas Journal.

Retrieved from http://www.ogj.com/articles/print/volume-88/issue-25/in-this-issue/pipeline/model-calculates-wax-deposition-for-n-sea-oils.html

Matzain, A., Apte, M. S., Zhang, H.-Q., Volk, M., Brill, J. P., & Creek, J. L. (2002). Investigation of Paraffin Deposition During Multiphase Flow in Pipelines and Wellbores—Part 1: Experiments. Journal of Energy Resources Technology, 124(3), 180. https://doi.org/10.1115/1.1484392

Murty, B. S., & Thandaveswara, B. S. (2014). Hydraulics. http://nptel.ac.in/courses/105106114/. Retrieved from http://doer.col.org/handle/123456789/3973

Panacharoensawad, E., & Sarica, C. (2013). Experimental Study of Single-Phase and Two-Phase Water-in-Crude- Oil Dispersed Flow Wax Deposition in a Mini Pilot-Scale Flow Loop.

Paso, K. G. (2014). Comprehensive Treatise on Shut-in and Restart of Waxy Oil Pipelines. Journal of Dispersion Science and Technology, 35(8), 1060–1085. https://doi.org/10.1080/01932691.2013.833105

Prajapati, D. B., Sahu, A. K., & Student, M. E. (2015). Design of Prototype Internal 20k Helium Gas Purification System for Helium Plant &amp; Test Facility. IJSTE-International Journal of Science Technology & Engineering | (Vol. 1). Retrieved from www.ijste.org

Rashidi, M., Mombekov, B., & Marhamati, M. (2016). A Study of a Novel Inter Pipe Coating Material for Paraffin Wax Deposition Control and Comparison of the Results with Current Mitigation Technique.

Rittirong, A. (2014). PARAFFIN DEPOSITION UNDER TWO-PHASE GAS-OIL SLUG FLOW IN HORIZONTAL PIPES by Ake Rittirong A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Discipline of Petroleum Engine.

Rittirong, A., Exploration, P. T. T., Pttep, P. P., & Panacharoensawad, E. (2015). An Experimental Study of Paraffin Deposition under Two-Phase Gas-Oil Slug Flow in Horizontal Pipes.

Rittirong, A., Panacharoensawad, E., & Sarica, C. (2016). Experimental Study of Paraffin Deposition Under Two-Phase Gas / Oil Slug Flow in Horizontal Pipes, (August 2015), 4–7.

22

Rittirong, A., Panacharoensawad, E., & Sarica, C. (2017). Experimental Study of Paraffin Deposition Under Two-Phase Gas / Oil Slug Flow in Horizontal Pipes, (February), 4–7.

Roenningsen, H. P., Bjoerndal, B., Baltzer Hansen, A., & Batsberg Pedersen, W. (1991). Wax precipitation from North Sea crude oils: 1. Crystallization and dissolution temperatures, and Newtonian and non-Newtonian flow properties. Energy & Fuels, 5(6), 895–908. https://doi.org/10.1021/ef00030a019

Ronningsen, H. P., Bjorndal, B., Hansen, A. B., & Pedersen, W. B. (1991). Wax Precipitation from North Sea Crude Oils. 1. Crystallization and Dissolution Temperatures, and Newtonian and Non-Newtonian Flow Properties. Energy and Fuels, 5(6), 895–908. https://doi.org/10.1021/ef00030a019

Sarica, C., & Panacharoensawad, E. (2012). Review of Para ffi n Deposition Research under Multiphase Flow Conditions.Singh, A., Lee, H. S., Singh, P., & Sarica, C. (2011a). Flow Assurance: Validation of Wax Deposition Models Using Field

Data from a Subsea Pipeline. In Offshore Technology Conference. Offshore Technology Conference. https://doi.org/10.4043/21641-MS

Singh, A., Lee, H. S., Singh, P., & Sarica, C. (2011b). Flow Assurance: Validation of Wax Deposition Models Using Field Data from a Subsea Pipeline. In Offshore Technology Conference. Offshore Technology Conference. https://doi.org/10.4043/21641-MS

Singh, P., Fogler, H. S., & Nagarajan, N. (1999). Prediction of the wax content of the incipient wax-oil gel in a pipeline: An application of the controlled-stress rheometer. Journal of Rheology, 43(6), 1437–1459. https://doi.org/10.1122/1.551054

Singh, P., Venkatesan, R., Fogler, H. S., & Nagarajan, N. (2000). Formation and aging of incipient thin film wax-oil gels. AIChE Journal, 46(5), 1059–1074. https://doi.org/10.1002/aic.690460517

Sreenivas, J. (2011). BENDS (FLOW AND PRESSURE DROP IN). In A-to-Z Guide to Thermodynamics, Heat and Mass Transfer, and Fluids Engineering. Begellhouse. https://doi.org/10.1615/AtoZ.b.bends_flow_and_pressure_drop_in

Sulaiman, S. A., Biga, B. K., & Chala, G. T. (2017). Injection of non-reacting gas into production pipelines to ease restart pumping of waxy crude oil. Journal of Petroleum Science and Engineering, 152(August 2016), 549–554. https://doi.org/10.1016/j.petrol.2017.01.046

Teo, M. (2016). Flow assurance for offshore production. World Pumps, 2016(3), 20–22. https://doi.org/10.1016/S0262-1762(16)70024-1

Theyab, M. A. (2017). Study Of Fluid Flow Assurance In Hydrocarbon Production – Investigation Wax Mechanisms, (April).

Theyab, M. A., & Diaz, P. (2016a). Experimental study of wax deposition in pipeline – effect of inhibitor and spiral flow. International Journal of Smart Grid and Clean Energy Experimental, 174–181. https://doi.org/10.12720/sgce.5.3.174-181

Theyab, M. A., & Diaz, P. (2016b). Experimental study of wax deposition in pipeline – effect of inhibitor and spiral flow. International Journal of Smart Grid and Clean Energy, (September). https://doi.org/10.12720/sgce.5.3.174-181

UIO. (2004). Flow In Pipe. In Fluid Mechanics (pp. 321–343). Retrieved from https://www.uio.no/studier/emner/matnat/math/MEK4450/h11/undervisningsmateriale/modul-5/Pipeflow_intro.pdf

Venkatesan, R. (2004). The deposition and rheology of organic gels. Retrieved from https://deepblue.lib.umich.edu/handle/2027.42/124203

Vester, K., Athanasia, S., Sohrab, S., & Örlü, R. (2016). Combined hot-wire and PIV measurements of a swirling turbulent flow at the exit of a 90° pipe bend. Journal of Visualization, 19(2), 261–273. https://doi.org/10.1007/s12650-015-0310-1

Visintin, R. F. G., Lapasin, R., Vignati, E., D’antona, P., & Lockhart, T. P. (2005). Rheological Behavior and Structural Interpretation of Waxy Crude Oil Gels. https://doi.org/10.1021/la050705k

Wang, C.-C., Youn Chen, I., Yang, Y.-W., & Hu, R. (2004). Influence of horizontal return bend on the two-phase flow pattern in small diameter tubes. Experimental Thermal and Fluid Science, 28(2–3), 145–152. https://doi.org/10.1016/S0894-1777(03)00033-5

Wang, J., Zhou, F., Zhang, L., Huang, Y., Yao, E., Zhang, L., … Fan, F. (2019). Experimental study of wax deposition pattern concerning deep condensate gas in Bozi block of Tarim Oilfield and its application. Thermochimica Acta, 671(June 2018), 1–9. https://doi.org/10.1016/j.tca.2018.10.024

Watson, M., Pickering, P., & Hawkes, N. (2003). The Flow Assurance Dilemma: Risk Versus Cost? THE FLOW ASSURANCE DILEMMA: RISK VERUS COST? Retrieved from http://www.engr.mun.ca/~spkenny/Courses/Undergraduate/ENGI8673/Reading_List/2003_Watson_Flow_Assurance.pdf

Xing, L., Yeung, H., Shen, J., & Cao, Y. (2013). Experimental study on severe slugging mitigation by applying wavy pipes. Chemical Engineering Research and Design, 91(1), 18–28. https://doi.org/10.1016/j.cherd.2012.06.020

Yadav, M. S., Kim, S., Tien, K., & Bajorek, S. M. (2014). Experiments on geometric effects of 90-degree vertical-upward elbow in air water two-phase flow. International Journal of Multiphase Flow, 65, 98–107. https://doi.org/10.1016/J.IJMULTIPHASEFLOW.2014.05.014

Zhang, G., & Liu, G. (2010). Study on the wax deposition of waxy crude in pipelines and its application. Journal of

23

Petroleum Science and Engineering, 70(1–2), 1–9. https://doi.org/10.1016/j.petrol.2008.11.003Zhao, D., Omar, R., Abdulkadir, M., Abdulkareem, L. A., Azzi, A., Saidj, F., … Azzopardi, B. J. (2017). The control and

maintenance of desired fl ow patterns in bends of different orientations. Flow Measurement and Instrumentation, 53, 230–242. https://doi.org/10.1016/j.flowmeasinst.2016.09.003

Zheng, S. (2017). Wax Deposition from Single-Phase Oil Flows and Water-Oil Two-Phase Flows in Oil Transportation Pipelines.

Zheng, S., Saidoun, M., Mateen, K., Palermo, T., & Total, E. (2016). Wax Deposition Modeling with Considerations of Non-Newtonian Fluid.

Zheng, S., Zhang, F., Huang, Z., & Fogler, H. S. (2013). E ff ects of Operating Conditions on Wax Deposit Carbon Number Distribution : Theory and Experiment. https://doi.org/10.1021/ef402051n

Zhu, T., Walker, J. a, Liang, J., & Laboratory, P. D. (2008). Evaluation of Wax Deposition and Its Control During Production of Alaska North Slope Oils . https://doi.org/10.2172/963363