13.2.1 Hydrocarbon Composition

The petroleum fluids from reservoirs normally are multiphase and multicomponent mixtures, primarily consisting of hydrocarbons, which can be classified into the following three groups:

In addition to hydrocarbons, water (H₂O), nitrogen (N₂), carbon dioxide (CO₂), hydrogen sulfide (H₂S), salts, and solids are often found in petroleum mixtures.

Table 13-1 lists some typical physical properties of the main components of petroleum. The boiling point of hydrocarbon components increases with an increase in the carbon number of the component formula. If the carbon number of a component is less than 5, the component is in the gas phase at atmospheric pressure. When the mixture contains larger molecules, it is a liquid at normal temperatures and pressures. A typical petroleum fluid contains thousands of different chemical compounds, and trying to separate it into different chemical homogeneous compounds is impractical. A composition of gas condensate is shown in Table 13-2. In the usual composition list of hydrocarbons, the last hydrocarbon is marked with a “+,” which indicates a pseudo-component, and lumps together all of the heavier components.

Table 13-1. Physical Properties of Main Petroleum Components [1]

Table 13-2. Typical Composition of a Gas Condensate [2]

The characterization of a pseudo-component including its molecular weight, density, pseudo-critical properties, and other parameters should be calculated ahead of the flow assurance analysis (see Chapter 12). The characterization should be based on experimental data and analyses. As shown in Table 13-2, hydrocarbon isotropic compounds with the same formula and carbon atom numbers may have very different normal boiling points and other physical properties.

13.2.2 Equation of State

For oil and gas mixtures, the phase behavior and physical properties such as densities, viscosities, and enthalpies are uniquely determined by the state of the system. The equations of state (EOS) for petroleum mixtures are mathematical relations between volume, pressure, temperature, and composition, which are used for describing the system state and transitions between states. Most thermodynamic and transport properties in engineering analyses are derived from the EOS. Since 1873, when the first EOS for representation of real systems was developed by the van der Waals, hundreds of different EOSs have been proposed and they are distinguished by Leland [3] into four families:

The cubic equations of the van der Waals family are widely used in the oil and gas industry for engineering calculations because of their simplicity and relative accuracy for describing multiphase mixtures. The simple cubic equations of state, Soave-Redlich-Kwong (SRK) and Peng-Robinson (PR), are used in flow assurance software (e.g., PIPESIM). They require limited pure component data and are robust and efficient, and usually give broadly similar results.

The SRK Equation[4]

(13-1)

where p is the pressure, T is the temperature, v is the molar volume, R is the gas constant, and a and b are the EOS constants, which are determined by the critical conditions for a pure component:

And m = 0.480 + 1.574 ω − 0.176 ω², where ω is an acentric factor. For a mixture, a and b are found as follows:

where z_i and z_j are the mole fraction of components i and j, respectively, and , where k_ij is a binary interaction coefficient, which is usually considered equal to zero for hydrocarbon/hydrocarbon interactions, and different from zero for interactions between a hydrocarbon and a nonhydrocarbon, and between unlike pairs of nonhydrocarbons.

PR Equation[5]

(13-2)

The EOS constants are given by

and m = 0.3764 + 1.54226 ω − 0.26992 ω², where ω is an acentric factor.

13.2.3 Hydrocarbon Properties

Oil and gas are very complex fluids composed of hydrocarbon compounds that exist in petroleum in a wide variety of combinations. Physical properties change with the component composition, pressure, and temperature. No two oils have the same physical properties. In considering hydrocarbon flow in pipes, the most important physical properties are density and viscosity.

13.3.1 General

A separate water phase in a pipeline system can result in hydrate formation or an oil/water emulsion under certain circumstances. Emulsions can have high viscosities, which may be an order of magnitude higher than the single-phase oil or water. The effect of emulsions on frictional loss may lead to a significantly larger error in pressure loss than calculated. In addition to affecting pipeline hydraulics, emulsions may present severe problems to downstream processing plants [7].

An emulsion is a heterogeneous liquid system consisting of two immiscible liquids with one of the liquids intimately dispersed in the form of droplets in the second liquid. In most emulsions of crude oil and water, the water is finely dispersed in the oil. The spherical form of the water globules is a result of interfacial tension, which compels them to present the smallest possible surface area to the oil. If the oil and water are violently agitated, small drops of water will be dispersed in the continuous oil phase and small drops of oil will be dispersed in the continuous water phase. If left undisturbed, the oil and water will quickly separate into layers of oil and water. If any emulsion is formed, it will exist between the oil above and the water below. In offshore engineering, most emulsions are the water-in-oil type.

The formation of an emulsion requires the following three conditions:

The agitation necessary to form an emulsion may result from any one or a combination of several sources:

The greater the amount of agitation, the smaller the droplets of water dispersed in the oil. Water droplets in water-in-oil emulsions are of widely varying sizes, ranging from less than 1 to about 1,000 μm. Emulsions that have smaller droplets of water are usually more stable and difficult to treat than those that have larger droplets. The most emulsified water in light crude oil, that is, oil above 20° API, is from 5 to 20 vol % water, while emulsified water in crude oil heavier than 20° API is from 10 to 35%. Generally, crude oils with low API gravity (high density) will form a more stable and higher percentage volume of emulsion than will oils of high API gravity (low density). Asphaltic-based oils have a tendency to emulsify more readily than paraffin-based oils.

The emulsifying agent determines whether an emulsion will be formed and the stability of that emulsion. If the crude oil and water contain no emulsifying agent, the oil and water may form a dispersion that will separate quickly because of rapid coalescence of the dispersed droplets. Emulsifying agents are surface-active compounds that attach to the water-drop surface. Some emulsifiers are thought to be asphaltic, sand, silt, shale particles, crystallized paraffin, iron, zinc, aluminum sulfate, calcium carbonate, iron sulfide, and similar materials. These substances usually originate in the oil formation but can be formed as the result of an ineffective corrosion-inhibition program.

13.3.2 Effect of Emulsion on Viscosity

Emulsions are always more viscous than the clean oil contained in the emulsion.

Figure 13-1 shows the tendency of viscosity with the increase of water cut in different emulsion conditions. This relationship was developed by Woelflin [8] and is widely used in petroleum engineering. However, at higher water cuts (greater than about 40%), it tends to be excessively pessimistic and may lead to higher pressure loss expectations than are actually likely to occur. Guth and Simha [9] presented a correlation of emulsion viscosity:

(13-13)

Figure 13-1 Woelflin Viscosity Data

where μ_e is the viscosity of emulation (in cP or mPa·s), μ_o is the viscosity of clean oil (cP or mPa·s), C_w is the volume fraction of the water phase, it is a ration of water volumetric flow rate compared to the volumetric flow rate of total liquids, including oil, water and other liquids.

This correlation is similar to that of Woelflin in that it predicts the emulsion viscosity to be some multiple of the oil viscosity. The factor is determined solely as a function of the water cut. Up to a water cut of about 40%, the two methods will give almost identical results. Above that, the effective viscosity rises much more slowly with the Guth and Simha [9] correlation, and computed pressure losses will thus be lower at higher water cuts.

Laboratory tests can provide useful information about emulsions, but the form of the emulsion obtained in practice is dependent on the shear history, temperature, and composition of the fluids. Limited comparison of calculated emulsion viscosity with experimental data suggests that the Woelflin correlation overestimates the viscosity at higher water fractions. The following equation, from by Smith and Arnold [10] can be used if no other data are available:

(13-14)

13.3.3 Prevention of Emulsion

If all water can be excluded from the oil and/or if all agitation of hydraulic fluid can be prevented, no emulsion will form. Exclusion of water in some wells is difficult or impossible, and the prevention of agitation is almost impossible. Therefore, production of emulsion from many wells must be expected. In some instances, however, emulsification is increased by poor operating practices.

Operating practices that include the production of excess water as a result of poor cementing or reservoir management can increase emulsion-treating problems. In addition, a process design that subjects the oil/water mixture to excess turbulence can result in greater treatment problems. Unnecessary turbulence can be caused by overpumping and poor maintenance of plunger and valves in rod-pumped wells, use of more gas-lift gas than is needed, and pumping the fluid when gravity flow could be used instead. Some operators use progressive cavity pumps as opposed to reciprocating, gear, or centrifugal pumps to minimize turbulence. Others have found that some centrifugal pumps can actually cause coalescence if they are installed in the process without a downstream throttling valve. Wherever possible, pressure drop through chokes and control valves should be minimized before oil/water separation.

13.4.1 Black Oils

Black oil is liquid oil and consists of a wide variety of chemical species including large, heavy, and nonvolatile molecules. Typical black oil reservoirs have temperatures below the critical temperature of a hydrocarbon mixture. Point C in Figure 13-2 represents unsaturated black oil; some gases are dissolved in the liquid hydrocarbon mixture. Point C1 in the same figure shows saturated back oil, which means that the oil contains as much dissolved gas as it can take and that a reduction of system pressure will release gas to form the gas phase. In transport pipelines, black oils are transported in the liquid phase throughout the transport process, whereas in production flowlines, produced hydrocarbon mixtures are usually in thermodynamical equilibrium with gas.

13.4.2 Volatile Oils

Volatile oils contain fewer heavy molecules than black oil, but more ethane through hexane. Reservoir conditions are very close to the critical temperature and pressure. A small reduction in pressure can cause the release of a large amount of gas.

13.4.3 Condensate

Retrograde gas is the name of a fluid that is gas at reservoir pressure and temperature. However, as pressure and temperature decrease, large quantities of liquids are formed due to retrograde condensation. Retrograde gases are also called retrograde gas condensates, gas condensates, or condensates. The temperature of condensates is normally between the critical temperature and the cricondentherm as shown in Figure 13-2. From reservoir to riser top, as shown in points B to B2 in the figure, the fluid changes from single gas and partial transfer to condensate liquid and evaporate again as gas, causing a significant pressure drop.

For a pure substance a decrease in pressure causes a change of phase from liquid to gas at the vapor–pressure line; likewise, in the case of a multicomponent system, a decrease in pressure causes a change of phase from liquid to gas at temperatures below the critical temperature. However, consider the isothermal decrease in pressure illustrated by a line B–B1–B2 in Figure 13-2. As pressure is decreased from point B, the dew-point line is crossed and liquid begins to form. At the position indicated by point B1, the system is 5% liquid. A decrease in pressure has caused a change from gas to liquid. Table 13-3 lists the composition of a typical condensate gas. Curve A–A2 illustrates the variations of pressure and temperature along the flowline from reservoir to the separator on a platform.

13.4.4 Wet Gases

A wet gas exists in a pure gas phase in the reservoir, but becomes a liquid/gas two-phase mixture in a flowline from the well tube to the separator at the topside platform. During the pressure drop in the flowline, liquid condensate appears in the wet gas.

13.4.5 Dry Gases

Dry gas is primarily methane. The hydrocarbon mixture is solely gas under all conditions of pressure and temperature encountered during the production phases from reservoir conditions involving transport and process conditions. In particular, no hydrocarbon-based liquids are formed from the gas although liquid water can condense. Dry gas reservoirs have temperatures above the cricondentherm. Tables 13-3 and 13-4 list the hydrocarbon composition and properties, respectively, of typical reservoirs.

Table 13-3. Hydrocarbon Composition of Typical Reservoirs [11]

Table 13-4. Hydrocarbon Properties of Typical Reservoirs [11]

13.4.6 Computer Models

Accurate prediction of physical and thermodynamic properties is a prerequisite to successful pipeline design. Pressure loss, liquid hold up, heat loss, hydrate formation, and wax deposition all require knowledge of the fluid states.

In flow assurance analyses, the following two approaches have been used to simulate hydrocarbon fluids:

The accuracy of the compositional model is dependent on the accuracy of the compositional data. If good compositional data are available, selection of an appropriate EOS is likely to yield more accurate phase behavior data than the corresponding black-oil model. This is particularly so if the hydrocarbon liquid is a light condensate. In this situation complex phase effects such as retrograde condensation are unlikely to be adequately handled by the black-oil methods. Of prime importance to hydraulic studies is the viscosity of the fluid phases. Both black-oil and compositional techniques can be inaccurate. Depending on the correlation used, very different calculated pressure losses could result. With the uncertainty associated with viscosity prediction, it is prudent to utilize laboratory-measured values.

The gas/oil ratio (GOR) can be defined as the ratio of the measured volumetric flow rates of the gas and oil phases at meter conditions (multiphase meter) or the volume ratio of gas and oil at the standard condition (1 atm, 60°F) in units of scf/STB. When water is also present, the water cut is generally defined as the volume ratio of the water and total liquid at standard conditions. If the water contains salts, the salt concentrations may be contained in the water phase at the standard condition.

13.5.1 General

The complex mixture of hydrocarbon compounds or components can exist as a single-phase liquid, a single-phase gas, or as a multiphase mixture, depending on its pressure, temperature, and the composition of the mixture. The fluid flow in pipelines is divided into three categories based on the fluid phase condition:

The pipelines after oil/gas separation equipment, such as transport pipelines and export pipelines, generally flow single-phase hydrocarbon fluid while in most cases, the production flowlines from reservoirs have two- or three-phase fluids, simultaneously, and the fluid flow is then called multiphase flow.

In a hydrocarbon flow, the water should be considered as a sole liquid phase or in combination with oils or condensates, since these liquids basically are insoluble in each other. If the water amount is small enough that it has little effect on flow performance, it may be acceptable to assume a single liquid phase. At a low-velocity range, there is considerable slip between the oil and water phases. As a result, the water tends to accumulate in low spots in the system. This leads to high local accumulations of water and, therefore, a potential for water slugs in the flowline. It may also cause serious corrosion problems.

Two-phase (gas/liquid) models are used for black-oil systems even when water is present. The water and hydrocarbon liquid are treated as a combined liquid with average properties. For gas condensate systems with water, three-phase (gas/liquid/aqueous) models are used.

The hydraulic theory underlying single-phase flow is well understood and analytical models may be used with confidence. Multiphase flow is significantly more complex than single-phase flow. However, the technology to predict multiphase-flow behavior has improved dramatically in the past decades. It is now possible to select pipeline size, predict pressure drop, and calculate flow rate in the flowline with an acceptable engineering accuracy.

13.5.2 Single-Phase Flow

The basis for calculation of changes in pressure and temperature with pipeline distance is the conservation of mass, momentum, and energy of the fluid flow. In this section, the steady-state, pressure-gradient equation for single-phase flow in pipelines is developed. The procedures to determine values of wall shear stress are reviewed and example problems are solved to demonstrate the applicability of the pressure-gradient equation for both compressible and incompressible fluid. A review is presented of Newtonian and non-Newtonian fluid flow behavior in circular pipes. The enthalpy-gradient equation is also developed and solved to obtain approximate equations that predict temperature changes during steady-state fluid flow.

13.5.2.2 Friction Factor Equation

The Darcy-Weisbach equation is one of the most general friction head loss equations for a pipe segment. It is expressed as:

(13-22)

(13-23)

where f is the friction factor, Re = VDρ/μ is the Reynolds number, and ε/D is the equivalent sand-grain roughness (relative roughness) of the pipe, which is decided by the pipeline material. Figure 13-4 shows the graphical representation of the friction factor at different flow regions, which is also called as Moody diagram.

Figure 13-4 Moody Diagram [13]

Table 13-5 lists common values of absolute roughness for several materials. For steel pipe, the absolute roughness is 0.0018 in. unless alternative specifications are specified. The absolute roughness of 0.0018 in. is for aged pipe, but that is accepted in design practice because it is conservative. Flexible pipe is rougher than steel pipe and, therefore, requires a larger diameter for the same maximum rate. For a flexible pipe, the roughness is given by ε = ID/250.0 unless an alternative specification is given. The roughness may increase with use at a rate determined by the material and nature of the fluid. As a general guide, factors of safety of 20% to 30% on the friction factor will accommodate the change in roughness conditions for steel pipe with average service of 5 to 10 years.

Table 13-5. Pipeline Material Roughness [14]

Material	ε, Equivalent Sand-Grain Roughness (in.)
Concrete	0.012–0.12
Case iron	0.010
Commercial or welded steel	0.0018
PVC, glass	0.00006

As shown in Figure 13-4, the function f (Re, ε/D) is very complex. A laminar region and a turbulent region are defined in the Moody diagram based on the flow characteristics.

Laminar Flow

In laminar flow (Re < 2100), the friction factor function is a straight line and is not influenced by the relative roughness:

(13-24)

The friction head loss is shown to be proportional to the average velocity in the laminar flow. Increasing pipe relative roughness will cause an earlier transition to turbulent flow. When the Reynolds number is in the range from 2000 to 4000, the flow is in a critical region where it can be either laminar or turbulent depending on several factors. These factors include changes in section or flow direction. The friction factor in the critical region is indeterminate, which is in a range between friction factors for laminar flow and for turbulent flow.

Turbulent Flow

When the Reynolds number is larger than 4000, the flow inside the pipe is turbulent flow; the fiction factor depends not only on the Reynolds number but also on the relative roughness, ε/D, and other factors. In the complete turbulence region, the region above a dashed line in the upper right part of a Moody diagram, friction factor f is a function only of roughness ε/D. For a pipe in the transition zone, the friction factor decreases rapidly with increasing Reynolds number and decreasing pipe relative roughness. The lowest line in Figure 13-4 is the line representing the friction factor for the smoothest pipe, where the roughness ε/D is so small that it has no effect.

Although single-phase flow in pipes has been studied extensively, it still involves an empirically determined friction factor for turbulent flow calculations. The dependence of this friction factor on pipe roughness, which must usually be estimated, makes the calculated pressure gradients subject to considerable error and summarizes the correlations of Darcy-Weisbach friction factor for different flow ranges. The correlations for smooth pipe and complete turbulence regions are simplified from the one for a transitional region. For smooth pipe, the relative roughness term is ignored, whereas for the complete turbulence region, the Reynolds term is ignored.

Table 13-6 summarizes some Darcy-Weisbach friction factor correlations for both laminar flow and turbulent flow.

Table 13-6. Darcy-Weisbach Friction Factor Correlations [12]

Flow Region	Friction Factor, f	Reynolds Range
Laminar		Re < 2100
Smooth pipe		Re > 4000 and ε/D →0
Transitional, Colebrook-White correlation		Re > 4000 and transitional region
Complete turbulence		Re > 4000 and complete turbulence

13.5.3 Multiphase Flow

13.5.4 Comparison of Two-Phase Flow Correlations

Many two-phase flow correlations have been developed in past several decades, most of which are based on water/air systems, and every correlation has its limit because of the data used. Most software used in flow assurance has several options for models or correlations for two-phase flow. Baker Jardine [19] conducted a major data matching study to identify the most suitable two-phase flow correlations for oil wells, condensate wells, oil pipelines, and gas/condensate pipelines, and the results of the analysis are summarized in following sections.

13.6.1 General

The occurrence of slug flow in a transportation pipeline can cause many problems in design and operation processes, which include kinetic force on fittings and vessels, pressure cycling, control instability, and inadequate phase separation. Slugging greatly affects the design of receiving facilities. In gas-condensate systems, larger lines result in more liquid being retained in the pipeline at low rates. When the flow rate is increased, much of the liquid can be swept out, potentially overwhelming the liquid handling capability of the receiving facilities. The facilities can be flooded and damaged if the slugs are larger than the slug catcher capacity. Therefore, quantifying the slug size, frequency, and velocity is necessary prior to equipment design.

The pressure at the bottom of the riser can vary if the holdup in the riser is not about the same as that in the line feeding it. If the riser holdup is too large and the gas velocity is too small to provide continuous liquid lift, too much of the liquid reverses and flows downward. Liquid accumulates at the base, causing an unstable pressure situation. This is relieved by large liquid slugs periodically leaving the riser at a high velocity. The changes in liquid amount and the corresponding pressure changes can be dramatic. A large slug catcher installation can be provided onshore, but it is not economical to place it on the platform. This is one of the practical reasons why a pipeline section immediately ahead of the riser should be horizontal or have a slightly upward slope of 2° to 5°. The section length probably should be several times the riser height. The upward incline eliminates a possible “sump” effect and serves to decrease pressure/holdup instabilities. Severe slugging in the riser can be enhanced by a negative pipeline inclination just prior to it. Actually, severe slugging is unlikely if there is a positive inclination.

Figure 13-9 shows the effect of the mass flow rate of two-phase flow on flow stability. A higher flow rate helps to decrease the slug and increase flow stability in the flowline. Higher system pressures also increase the tendency for a stable flow. Choking, at the top of the riser can be used to minimize severe slugging. The flow in a riser may differ from that in a wellbore, which has a relatively long horizontal flowline at the end of it. Holdup and surging from that horizontal flowline are transmitted to the relatively short riser. The riser may have to handle far more liquid than a well because the flowline can feed it liquid surges that far exceed those possible by gas-lift or reservoir mechanisms.

Figure 13-9 Riser Flow Stability versus Flow Rate

In many oil and gas developments that incorporate multiphase flowlines, the possibility of slugs or surges is one of the most important flow assurance concerns due to the excessive demands large changes in oil/gas flow rates place on the processing facilities. Multiphase surges come in three forms:

13.6.2 Hydrodynamic Slugging

Hydrodynamic slugs are initiated by the instability of waves on the gas/liquid interface in stratified flow under certain flowing conditions. Figure 13-10 shows the formation process of hydrodynamic slugging from a stratified flow. In Figure 13-10a, the gas/liquid interface is lifted to the top of the pipe when the velocity difference between gas phase and liquid phase is high enough. This wave growth is triggered by the Kelvin-Helmholtz instability. Once the wave reaches the top of pipe, it forms a slug (Figure 13-10b). The slug is pushed by the gas and so travels at a greater velocity than the liquid film, and more liquid is then swept into the slug. In Figure 13-10c, entrainment reduces the average liquid holdup in the slug, increasing the turbulence within the slug.

Figure 13-10 Formation of Hydrodynamic Slugging [34]

A main part of the frictional pressure drop in multiphase flow is thought to be due to the turbulent region within the slug. Thus, the size of the turbulent region can have a significant effect on the frictional pressure losses in a pipeline. Two-phase flow pattern maps indicate hydrodynamic slugging, but slug length correlations are quite uncertain. Tracking of the development of the individual slugs along the pipeline is necessary to estimate the volume of the liquid surges out of the pipeline.

Slugging simulations need to be performed over the flow rates, water cuts, and GORs for the field life. The effects of any artificial lift should be included in the simulations. In general, simulation results are presented as liquid and gas flow rates at the separator, slug lengths at the base and top of the riser, and pressure at key locations as a function of time. The key locations in a well are upstream of the well flow control devices and bottomhole. If processing equipment is included in the model, the separator pressure, separator levels, and outlet gas and liquid rates from the separator as a function of time are presented.

In cases where the predicted slugging causes liquid or gas handling problems, the effects of additional choking upstream of the separator should be determined. The evolution of slugs is very sensitive to the pipe inclination and changing the inclination by less than a degree can be sufficient to change the balance, causing a flow regime transition. Thus, peaks and troughs along the pipeline profile of relatively small elevation change may have a very significant effect.

Figure 13-11 compares the flow patterns of a simple horizontal topography and one with some undulations. The multiphase flow in undulate flowlines switches between the stratified and slug flow regimes, implying that not only could the slug sizes differ markedly, but the pressure drops could be very different too. Few pipelines have constant inclinations; most undulate following the natural terrain. When modeling multiphase flow in lower flow rates, it is important to represent these undulations as faithfully as possible. At higher flow rates, undulations may not have much impact on predictions.

Figure 13-11 Effect of Pipeline Topography to Flow Pattern [34]

13.6.3 Terrain Slugging

Figure 13-12 shows the formation process for a terrain slug in a pipeline-riser system:

Figure 13-12 Formation of Riser Slugging

Liquid accumulates at the base of the riser, blocking the flow of gas phase. This liquid packet builds up until the gas pressure is sufficiently high to overcome the hydrostatic head and blow the liquid slug from the riser. Slug lengths can be two to three times the riser height.

Terrain slugs can be very severe, causing large pressure variations and liquid surges out of a pipeline. Terrain slugging is a transient situation that requires a dynamic model to predict and describe. When the minimum flow rate is defined as the terrain slugging boundary, a region without severe slugging should be determined as a function of the water cut and including gas lift. In cases where the predicted slugging causes liquid or gas handling problems, the effect of additional choking upstream of the separator should be determined.

It may be difficult to design a slug catcher to cope with the magnitude of terrain slugs. If the transportation system terminates in a vertical riser onto a receiving platform, the passage of the slug from the horizontal pipeline to the vertical riser results in cyclic flow effects. As the slug decelerates into the vertical riser, the following gas bubble is compressed. Compression continues until sufficient energy is generated to accelerate the slug from the riser. When the wellhead pressure is limited, the vertical riser from the seabed into the platform may form a backpressure due to slugs, and it will limit well production. In such circumstances the vertical pressure loss can be reduced if a slug catcher is located on the seabed, and the gas and liquid phases are separated. The liquid is pumped to the surface with the gas free flowing to the platform through a separate riser. An alternative is to inject gas at the base of the riser, which will lighten the fluid column and minimize vertical pressure losses.

13.6.4 Start-Up and Blowdown Slugging

Slugs form in the start-up operation process because of transformation from a steady state to a transient process. The start-up simulations should be performed starting from shutdown conditions to different representative operating conditions throughout the field life. A range of start-up rates, consistent with reservoir management constraints, should be evaluated. If necessary, artificial lifting to mitigate start-up slugs should be evaluated. For gas lifting, the required gas-lift rate should be determined. If lift gas is unavailable until it is obtained from the production, this operating constraint should be included in the simulations.

13.6.5 Rate Change Slugging

When the flow rate is increased, the liquid holdup in the line decreases. This change in holdup can either exit the line as a steady flow with increased liquid production, or it can come out in the form of a slug, depending on the flow rate change. The rate change slugs can occur in gas/condensate flowlines when the rates are increased. The flowline may be in a steady flow pattern, such as stratified flow, at both the initial and final flow rates but will slug during the transition period until the line reequilibrates at the higher rate.

As with start-up slugs, it is impossible to predict whether slugs will occur when rates are changed using steady-state or hand methods. The flowline must be dynamically simulated using a transient flow program.

13.6.6 Pigging

Pigs are run through pipelines for a variety of reasons, including:

Figure 13-13 shows the flow rate variations of liquid and gas at a riser base due to pigging. When the slug is big enough, the full riser base will be filled with liquid phase. The simulations of the pigging operation should be performed where pigging is required. The time for the pigging operation and the pig velocity as a function of time should be reported. For a round trip pigging, the inlet and outlet temperature, pressure, and flow rate as a function of time should be analyzed. For a round trip pigging with liquid, the backpressure necessary to maintain single-phase flow behind the pig should be analyzed.

Figure 13-13 Flow Rate Variations Due to Pigging

13.6.7 Slugging Prediction

The slugging prediction can be carried out using PIPESIM (Schlumberger), OLGA (Scandpower Petroleum Technology), ProFES (Aspentech), or TACITE (Simulation Science). Steady-state multiphase simulation software such as PIPESIM can predict hydrodynamic slug distributions along the flowline and riser slugs. However, only OLGA 2000 (standard version), the OLGA 2000 slug tracking module, and ProFES are good at simulating transient multiphase flow and predicting the liquid holdup variations along the flowlines, terrain slugs and start-up/shutdown transient slugs.

Ramp-up slugs are of primary importance for gas/condensate systems where the increased flow rate can sweep out large volumes of liquid. Simulations of flow rate ramps should be performed from turndown rates to intermediate and full production rates over the life of the field. OLGA 2000 slug tracking [35] is generally not required for ramp-ups in gas/condensate systems except in cases with hilly terrain or high liquid loadings (>50 bbl/mmscf). Results may be presented as outlet liquid and gas rates as a function of time.

The production system should be modeled starting at the reservoir using inflow performance relationships provided by reservoir modeling. At the outlet, pressure control is recommended. If this is not possible, a sensitivity study should be performed to ensure that random fluctuations in the outlet pressure do not significantly alter the slugging. The artificial boundary conditions of a constant pressure tend to dampen slugging and should be avoided whenever possible for slugging simulations. True boundary conditions on the flowline/riser can be obtained using a transient pipeline model coupled to a dynamic process model. Integrated dynamic pipeline and process simulation is rarely necessary for the design of wells, flowlines, and risers. However, in certain instances, integrated modeling is advisable for the design and control of the process facilities.

13.6.8 Parameters for Slug Characteristics

The key parameters required for the assessment of the performance of separation and associated downstream facilities are the average slug volume and frequency, and the greatest likely slug volume and its frequency of occurrence. The combination of slug velocity, frequency and liquid holdup is important for the design of pipeline supports. These are the control parameters: V_sg (gas superficial velocity), V_sl (liquid superficial velocity), D (pipeline diameter), L (pipeline length), ϕ (pipeline inclination), ρ_g (gas density), ρ_l (liquid density), μ_g (gas viscosity), μ_l (liquid viscosity), and σ_l (surface tension).

13.6.9 Slug Detection and Control Systems

Although slug control is very important to avoid facility damage and upsets, control options are limited. The potential impact of slugging on the topside system operation must be addressed and then analyses of the subsea system carried out to assess the effects. Usually a trade-off results between the design of slug catchers and the optimization of the flowline to reduce slugging.

Slugs have been successfully detected using gamma densitometers located on the riser, acoustical measurements, and measurements of pressure at the base of the riser. Slug detection systems should be considered when predicted slugging is expected to give operational difficulties and/or when an advanced control system is to be used for slug mitigation. In this case, the slugging simulations should include advanced controls to test the control algorithm. Flow assurance and process disciplines should demonstrate advanced controls to critically dampen predicted slug volumes and frequency.

13.6.10 Equipment Design for Slug Flow

The design of slug catchers, separators, and control systems in downstream pipelines is based on the presence and severity of slug flow in the system. The parameter estimations of slug volumes, liquid and gas rates exiting the pipeline as a function of time, etc., must be factored into the equipment design for slug flow. These parameters should be calculated for steady-state operation and for a series of transient operation cases: turndown rates, pigging, shutdown and start-up, rate changes, etc.

Transient modeling gives the best estimations of transient slug flow behavior. Some of the transient simulators include separators and control systems as part of the run. In general, the size of slug catching equipment for gas/condensate pipelines will be governed by pigging considerations. For oil-dominated systems, the size of the slug catcher is usually governed by the maximum slug length due to either hydrodynamic or terrain slugs.

13.6.11 Slug Catcher Sizing

Slug catchers should be sized to dampen surges to a level that can be handled by downstream processing equipment. Before dynamic models of the topside facilities are available, the level of acceptable surging is unknown and designers are often forced to make assumptions vis-à-vis surge volumes, such as designing for the “one-in-a-thousand” slug.

The surge volume for gas/condensate requirements is determined from the outlet liquid rates predicted in the ramp-up, start-up, and pigging cases. The required slug catcher size is dependent on liquid handling rate, pigging frequency, and ramp-up rates. An iterative process may be required to identify optimum slug catcher size, pigging frequency, liquid handling rate, and acceptable ramp-up rates. For this optimization, the results of the simulations should be presented as surge volume requirements as a function of liquid handling rate for representative ramp-up rates and pigging frequencies. Separator volumes for black-oil systems are typically set by separation requirements rather than liquid slug handling capacity. Consequently, the ability of the separator to accommodate slugs from all operations should be confirmed based on the results of the slugging simulations.

13.7.1 Slug Catcher Design Process

The goal of slug catcher design is to properly size the slug catcher for the appropriate conditions. The process consists of the following steps:

13.7.2 Slug Catcher Functions

The slug catcher functions may be summarized as follows:

Each slug catcher can serve one or more functions, and each function is detailed in the following sections.

13.8.1 Fundamentals of Pressure Surge

An important consideration in the design of single-liquid-phase pipelines is pressure surge, also known as water hammer. Typical surge events in a pipeline or piping system are generally caused by a pump shutdown or a valve closure [37]. The kinetic energy of flow is converted to pressure energy. The velocity of the pressure wave propagation is determined by fluid and pipeline characteristics. Typical propagation velocities range from 1100 ft/s for propane/butane pipelines to 3300 ft/s for crude oil pipelines and up to 4200 ft/s for heavy-wall steel water pipelines.

A rough estimate of the total transient pressure in a pipeline/piping system following a surge event can be obtained from the following equations:

(13-27)

where

The surge pressure wave travels upstream and is reflected downstream, oscillating back and forth until its energy is dissipated in pipe wall friction. The amplitude of the surge wave, or the magnitude of the pressure surge P_surge, is a function of the change in velocity and the steepness of the wave front and is the inverse of the time it took to generate the wave:

(13-28)

where

This is only a rough estimation of the surge pressure magnitude for cases limited by the stated time of closure criteria. An accurate analysis of the maximum surge pressure and location of critical points in the system may be obtained from a dynamic simulation by using software such as Pipe Simulator. The surges are attenuated by friction, and the surge arriving at any point on the line is less than at the origin of the surge wave. Nevertheless, when the flow velocity is high and stoppage is complete, or when a pump station is bypassed suddenly, the surge energy generated can produce pressures high enough to burst pipe. Based on the ASME codes, the level of pressure rise due to surges and other variations from normal operations shall not exceed the internal design pressure at any point in the piping system and equipment by more than 10%.

13.8.2 Pressure Surge Analysis

Surge analysis should be performed during a project’s early design and planning phases. This analysis will help to ensure the achievement of an integrated and economical design. Surge analysis provides assurance that the selected pumps/compressors, drivers, control valves, sensors, and piping can function as an integrated system in accordance with a suitable control system philosophy. A surge analysis becomes mandatory when one repairs a ruptured pipeline/piping system to determine the source of the problem.

All long pipelines/piping systems designed for high flow velocities should be checked for possible surge pressures that could exceed the maximum allowable surge pressure (MASP) of the system piping or components. A long pipeline/piping system is defined as one that can experience significant changes in flow velocity within the critical period. By this definition, a long pipeline/piping system may be 1500 ft long or 800 miles long.

A steady-state design cannot properly reflect system operation during a surge event. On one project in which a loading hose at a tanker loading system in the North Sea normally operated at 25 to 50 psig, a motor-operated valve at the tanker manifold malfunctioned and closed during tanker loading. The loading hose, which had a pressure rating of 225 psig, ruptured. Surge pressure simulation showed that the hydraulic transient pressure exceeded 550 psig. Severe surge problems can be mitigated through the use of quick-acting relief valves, tanks, and gas-filled surge bottles, but these facilities are expensive single-purpose devices.

13.9.1 Hydraulic Calculations

Hydraulic calculations should be carried out from wells, pipelines, and risers to the surface facilities in line sizing. The control factors in the calculation are fluids (oil, gas, or condensate fluids), flowline size, flow pattern, and application region. Unlike single-phase pipelines, multiphase pipelines are sized taking into account the limitations imposed by production rates, erosion, slugging, and ramp-up speed. Artificial lift is also considered during line sizing to improve the operational range of the system.

Design conditions are as follows:

Usually either of these conditions requires a trial-and-error approach based on assumed pipe sizes to meet the stated conditions. Some design problems may require determination of maximum flow for a given line size and length, but this is a reverse of the conditions above.

During the FEED stage of a project, the capability of a given line size to deliver the production rate throughout the life of the field is determined by steady-state simulation from the reservoir to the surface facilities throughout the field life. Sensitivities to the important variables such as GOR, water cut, viscosity, and separator pressure should be examined.

The technical criteria for line sizing of pipelines are stated in next section; however, the optimum economic line size is seldom realized in the engineering design. Unknown factors such as future flow rate allowances, actual pressure drops through certain process equipment, etc., can easily overcompensate for what the design predicted when selecting the optimum. In sizing a pipeline, one is always faced with a compromise between two factors. For a given flow rate of a given fluid, piping cost increases with diameter. But pressure loss decreases, which reduces potential pumping or compression costs. There is an economic balance between material costs and pumping costs in downstream flowlines. The optimum pipe size is found by calculating the smallest capitalization/operating cost or using the entire pressure drop available; or increasing velocity to highest allowable.

13.9.2 Criteria

The line sizing of the pipeline is governed by the following technical criteria:

Other criteria considered in the selection of the optimum line size include:

13.9.3 Maximum Operating Velocities

Liquid velocity is usually limited because of erosion effects at fittings. Erosion damage can occur in flowlines with multiphase flow because of the continuous impingement of high-velocity liquid droplets. The damage is almost always confined to the place where the flow direction is changed, such as elbows, tees, manifolds, valves, and risers. The erosional velocity is defined as the bulk fluid velocity that will result in the removal of corrosion product scales, corrosion inhibitors, or other protective scales present on the intersurface of a pipeline [38].

The velocity of pipeline fluids should be constrained as follows:

(13-29)

where

The mixture density, ρ_m is defined as:

(13-30)

where

Assuming on erosion rate of 10 mils per year, the following maximum allowable velocity is recommended by Salama and Venkatesh [40], when sand appears in an oil/gas mixture flow:

(13-31)

where

Erosion limits or maximum velocities for flexible pipes should be specified by the manufacturer. The API erosion velocity limit is normally very conservative and the practice of this limit could vary depend on different operators.

13.9.4 Minimum Operating Velocities

The following items may effectively impose minimum velocity constraints:

13.9.5 Wells

During FEED analysis, the production system should be modeled starting at the reservoir using inflow performance relationships. OLGA 2000 may be used to evaluate the hydraulic stability of the entire system including the wells, flowlines, and risers subject to artificial lift to get minimum stable rates as a function of the operating requirements and artificial lift.

In the hydraulic and thermal analyses of subsea flowline systems, the wellhead pressure and temperature are generally used as the inlet pressure and temperature of the system. The wellhead pressure and temperature are functions of reservoir pressure, temperature, productivity index, and production rate and can be obtained from steady and transient hydraulic and thermal analyses of the well bore. The hydraulic and thermal models may be simulated using the commercial software PIPESIM or OLGA 2000.

The hydraulic and thermal processes of a wellbore are simulated for the following purposes:

The transient analyses determine how hydrates can be controlled during startup and shut-in processes. Vacuum insulated tubing (VIT) is one method to help a wellbore out of the hydrate region. Use of 3000 to 4000 ft of VIT results in warm-up of the wellbore to temperatures above the hydrate temperature in typically 2 hr or less. These warm-up rates are rapid enough to ensure that little or no hydrates form in the wellbore. This makes it possible to eliminate the downhole injection of methanol above the subsurface safety valve during warm-up. Instead, methanol is injected upstream of the choke to prevent any hydrates that may be formed during warm-up from plugging the choke. One drawback of VIT is that it cools more quickly than bare tubing since the insulation prevents the tubing from gaining heat from the earth, which also warms up during production. As a result of this quick cooling, it is necessary to quickly treat the wellbore with methanol in the event that a production shut-in occurs during the warm-up period. For a higher reservoir temperature, VIT may not be used in the wellbore, and the cooldown speed of a wellbore is very slow after a process interruption from steady-state production conditions.

13.9.6 Gas Lift

Artificial lift can be used to increase production rates or reduce line sizes. Line sizing may include the affects of gas lift.

For systems employing gas lift, models should extend from the reservoir to the surface facilities incorporating the gas lift gas in addition to the produced fluids. The composition of the gas-lift gas is provided by the topside process but should meet the dew-point control requirement. The preferred location (downhole, flowline inlet, or riser base) is determined based on effectiveness and cost.

The effect of gas lift on the severe slugging boundary may be simulated with OLGA 2000 for the situations, when it is required. Normally, it is set at the riser base. The simulation may decide the minimum flow rate without severe slugging as a function of gas-lift rates and water cuts. When the gas-lift injection point does not include a restriction or valve to choke the gas flows, OLGA 2000 transient simulations may model the transient flow in the gas-lift line by including the gas-lift line as a separate branch to determine maximum variation in mass rate and pressure and still maintain system stability. Maximum production rates throughout the field life are determined as a function of gas-lift rate. In FEED designs, the gas-lift rate and pressure requirements should be recommended, including a measurement and control strategy for the gas-lift system.