الفهرس 
SummaryOnly 14 pages are availabe for public view
 |  | from | 146 |  |  |
| | | from | 146 | | |
Abstract
Heavy oils cannot easily flow through pipes due to their high viscosity. Technologies to enhance the mobility of heavy oils are required in oil production and transportation. Transportation of heavy oil by lubricated technique looks an attractive method for saving pumping power. In the present study, the effect of varying mean injection velocities of water and oil on the flow structure and the pressure gradient of the two-phase flow was studied experimentally by using heavy oil (viscosity = 750 mPa.s at 40 C) flow in a horizontal pipe of 29 mm inner diameter and 3.66 m long test section. With the help of photography, six different flow patterns; dispersion oil in water flow, plug flow, slug flow, smooth stratified flow, wavy stratified flow, core-annular flow and wavy core-annular flow were identified and the flow pattern maps were constructed for each case. The pressure DROP reduction factor was introduced to characterize the oil-water two-phase flow. from the experimental results, the pressure DROP reduction factor increases with increasing the amount of oil transported. The water input ratio decreases with increasing the oil flow rate (in the direction of annular flow observation). The experimental data for pressure DROP was compared with the analytical solution of Arney model. from the comparison, the deviation between the experimental data and the predicted results of the analytical model is less than 20%. Also, the two-phase frictional multiplier (ф_o) as a function of Lockhart-Martinelli parameter (X) was used to compare the experimental data with the Brauner model and modified Brauner model and the comparison shows a good agreement for the Brauner model with e_RMS=15.4% and with e_RMS=13.5% for the modified Brauner model.Two homogenous models were developed to compare the experimental two-phase pressure DROP with predicted pressure gradient obtained from the two models. In addition, the experimental pressure drops obtained from previous works were compared with the predicted correlation showing a good agreement.The RMS error was minimized by introducing the empirical factor Z to modify the predicted correlations.