Large Probe Arrays for Measuring Mean and Time Dependent Local Oil Volume Fraction and Local Oil Velocity Component Distributions in Inclined Oil-in-water Flows

Arrays of dual-sensor and four-sensor needle conductance probes have been used to measure the mean and time dependent local properties of upward inclined, bubbly oil-in-water flows (also known as dispersed oil-in-water flows) in a 153mm diameter pipe. The flow properties that were measured were (i) the local in-situ oil volume fraction α ; (ii) the local oil velocity o u in the axial direction of the pipe (the Z direction); and (iii) the local oil velocity Y u in the direction from the lower side of the inclined pipe to its upper side (the Y direction). Oil velocities in the X direction (orthogonal to the Y and Z directions) were found to be negligible. For all of the flow conditions investigated it was found that the mean value of α varied from a maximum value at the upper side of the inclined pipe to a minimum value at the lower side, and that the rate of decrease of this mean value of α with distance in the Y − direction became greater as the pipe inclination angle θ from the vertical was increased. It was also found that the mean value of o u was greatest at the upper side of the inclined pipe and decreased towards the lower side of the inclined pipe, the rate of decrease with distance in the Y − direction again becoming greater as θ was increased. For o 45 = θ , a water volumetric flow rate 1 hr m 38 . 16 − = w Q , an oil volumetric flow rate 1 hr m 0 . 6 − = o Q and using a sampling period s 05 . 0 = T over a total time interval of s 60 , it was found that at the upper side of the inclined pipe the standard deviation in o u was 31.6% of the mean value of o u . Furthermore for s 05 . 0 = T , o 30 = θ , 1 hr m 38 . 16 − = w Q and 1 hr m 0 . 6 − = o Q it was found that the standard deviation in the cross-pipe oil velocity component Y u was approximately equal to the standard deviation in the axial velocity component o u . These large temporal variations in the local flow properties have been attributed to the presence of large scale Kelvin-Helmholtz waves which intermittently appear in the flow. It is believed that the techniques outlined in this paper for measuring the standard deviation of local flow properties as a function of the sampling period T will be of considerable value in validating mathematical models of time dependent oilwater flows. It should be noted that the principal focus of this paper is on the measurement techniques that were used and the methods of data analysis rather than the presentation of exhaustive experimental results at numerous different flow conditions.