Much of engineering in the petroleum industry is about transportation of fluids. Petroleum engineers are concerned with how fluids in the reservoir move through the geological formation, up man-made sub-surface facilities, to the surface. In general, we say that the Petroleum engineer's role for fluid transport is in the sub-surface
The well fluid comprises, not just the hydrocarbon we desire, but quite a few unwanted materials. We may intend to produce crude oil. But, what we get out of the "oil well" is a cocktail of oil, water, gas, and dirt. The (unwanted) water and gas are known respectively as associated water and associated gas - they are associated with the needed oil. The facilities engineer is responsible for treatment of the well fluid to separate out the oil we desire. He is responsible for transportation of the well fluid from the well to wherever he plans to install the processing facilities, provision of the processing facilities, and the export system. Equipment installed by the facilities engineer, from the wellhead to and including the export terminal, are referred to as surface facilities.
Surface facilities must be designed on the basis of the petroleum engineer's estimated potential quantities and characteristics of the well fluid. These include the expected flow rates, gas-liquid ratios, oil and gas specific gravities and viscosities, and the anticipated rate of increase in the water content of the produced fluid. A key issue is the fact that the quantities and characteristics of the produced fluid do not remain the same throughout the life of the well. In many cases, the water-cut increases progressively as the well is produced.
Well fluid is normally transported in small-diameter (80 - 200mm) pipelines (known as flowlines) to a gathering station. A gathering station collects fluid from several wells. A flowline is considered a 2-phase line, transporting gas and liquid. Several wells (usually, but not necessarily in the same field) produce into a gathering station. At the gathering station, gas is separated from the liquid in a 2-phase separation process. Part of the gas is used to power prime- movers for pumps and generate electricity for electrically-powered station equipment, including lighting. Gas that is surplus to these requirements may be collected in a gas-gathering system if such exists, or just flared.
Liquid from the 2-stage separation process in the gathering station comprises oil and water. This is pumped into pipelines (known as delivery lines) that lead to some dehydration centre. This fluid is known as wet oil, that is, oil from which associated water has not been removed. Delivery lines are of a larger diameter than flowlines since they must transport liquid produced from several wells. Although delivery lines are single-phase lines in the sense that they transport only liquid, it must be recognised that the liquid actually comprises two sub-phases, viz, oil and water. The pipeline engineer is required to design and specify the pipeline configuration (diameter, wall thickness, material, etc) required for each application. Selection of the appropriate diameter for the pipeline depends not only on the expected flow rates, but also on the composite properties of the oil/water mixture.
At the dehydration centre, the wet oil (i.e. oil with water) is received into a settling tank for gravity separation. Settled water drained from the tank is treated to meet statutory requirement for disposal into the environment before disposal. How quickly the entrained water separates from the oil in the tank depends, not only on the fluid properties, but on the history of the fluid's journey from the gathering station. An under-sized pipeline creates extreme fluid turbulence, higher emulsion viscosities and a slower gravity separation of the oil and water at the dehydration centre. On the other hand, an over-sized pipeline, while facilitating the eventual gravity separation, costs more to provide. Furthermore, an over-sized line tends to corrode faster as the water remains at the bottom of the pipe during flow.
Prediction of pressure loss in pipe flow is crucial to selection of the specification of the pipe, and associated equipment such as pumps and terminal facilities. The three main components of the pressure loss are those due to:
In most cases, that due to friction dominates. Pressure loss due to friction has received the most attention in research and publications because the parameters from which it derives are not always easily determined. One thing that is agreed is that friction pressure loss is calculated from Darcy's equation:
| Eqn 1/1 |
Using volume flow as input instead of velocity, this equation transforms to:
| Eqn 1/2 |
Central to Darcy's equation is the friction factor which generally derives from the fluid properties, flow rate and the characteristics of the internal surface of the pipe. The pipe internal surface characteristics is not of any consequence at very low flow rates where laminar regime obtains. In laminar flow, the boundary layer effectively isolates the bulk fluid flow from the pipe surface. However, regardless of the flow regime, the fluid properties - viscosity and specific gravity, feature prominently in determination of the friction factor. Of these, the most contentious is the viscosity, when the fluid is some admixture of oil and water such as we have in wet crude oil.
Reynold's dimensionless parameter, known as Reynold's number (NRE) is universally accepted as the basis of assessing the turbulence with which fluid flows in a pipe. It is given by:
| Eqn 1/3 |
In general, if NRE is less than 2000, the flow is considered laminar. Above that, we assume it is turbulent. (Yes, there is a transition region somewhere between NRE = 4000 where there is uncertainty in quantifying what happens.) For laminar flow, the friction factor is calculated from Poiseuille's equation:
| Eqn 1/4 |
For turbulent flow, Colebrook-White's implicit equation applies:
![]() |
Eqn 1/5 |
For very high values of NRE such as are found in gas flow, fn is hardly affected by NRE. However, practical liquid flow cases generally fall in the region NRE < 100,000 where the effect of NRE on fn is most significant. It is therefore important that the basis of calculating NRE be reasonably accurate. For values of Reynolds number in the transition zone, we interpolate linearly between the friction factor values obtained by Poisseuille's and Colebrook-White equations.
We should note that what we refer to as friction pressure drop actually comprises two components: pipe friction pressure drop, and fluid friction pressure drop. These two components can be identified in the Colebrook-White equation (1/5). The pipe friction relates to the interaction between the fluid and the rough surface of the pipe. The first term in braces accounts for this. Fluid friction pressure drop is due to friction between the fluid molecules. In laminar flow, pipe friction is absent because the rough pipe surface is protected by the boundary layer. Thus, in laminar flow, the total friction pressure drop is equal to the internal fluid friction pressure drop.
By inspection of equation 1/3, the impact of an error in the viscosity µ is evident. The density of wet oil is the total mass of the constituent oil and water, divided by the total volume. This is a simple weighted mean of the densities of the constituent oil and water weighted in the proportion of their volume contributions to the mixture, since neither mass nor volume is lost by mixing the oil and water, regardless of the mixing vigour. However, it is not the case with viscosity, an interactive physical quantity.